Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special...

83
NIST Special Publication 260-189 Certification of Standard Reference Material ® 2372a Human DNA Quantitation Standard Erica L. Romsos Margaret C. Kline David L. Duewer Blaza Toman Natalia Farkas This publication is available free of charge from: https://doi.org/10.6028/NIST.SP.260-189

Transcript of Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special...

Page 1: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .

NIST Special Publication 260-189

Certification of Standard Reference Materialreg 2372a

Human DNA Quantitation Standard

Erica L Romsos Margaret C Kline David L Duewer

Blaza Toman Natalia Farkas

This publication is available free of charge from httpsdoiorg106028NISTSP260-189

NIST Special Publication 260-189

Certification of Standard Reference Materialreg 2372a

Human DNA Quantitation Standard Erica L Romsos

Margaret C Kline Biomolecular Measurement Division

Material Measurement Laboratory

David L Duewer Chemical Sciences Division

Material Measurement Laboratory

Blaza Toman Statistical Engineering Division

Information Technology Laboratory

Natalia Farkas Engineering Physics Division

Physical Measurement Laboratory

This publication is available free of charge from httpsdoiorg106028NISTSP260-189

March 2018

US Department of Commerce Wilbur L Ross Jr Secretary

National Institute of Standards and Technology Walter Copan Under Secretary of Commerce for Standards and Technology and Director

Certain commercial entities equipment or materials may be identified in this document to describe an experimental procedure or concept adequately

Such identification is not intended to imply recommendation or endorsement by the National Institute of Standards and Technology nor is it intended to imply that the entities materials or equipment are necessarily the best available for the purpose

National Institute of Standards and Technology Special Publication 260-189 Natl Inst Stand Technol Spec Publ 260-189 83 pages (March 2018)

CODEN NSPUE2

This publication is available free of charge from httpsdoiorg106028NISTSP260-189

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

i

Abstract Standard Reference Material (SRM) 2372a is intended for use in the value assignment of human genomic deoxyribonucleic acid (DNA) forensic quantitation materials A unit of SRM 2372a consists of three well-characterized human genomic DNA materials solubilized in 10 mmolL 2-amino-2-(hydroxymethyl)-13 propanediol hydrochloride (Tris HCl) and 01 mmolL ethylenediaminetetraacetic acid disodium salt (disodium EDTA) pH 80 buffer (TE-4) The three component genomic DNA materials labeled A B and C are respectively derived from a single male donor a single female donor and 13 mixture of a male and a female donor A unit of the SRM consists of one 05 mL tube of each component each tube containing approximately 55 microL of DNA solution Each of these tubes is labeled and is sealed with a color-coded screw cap This publication documents the production analytical methods and statistical evaluations involved in production of this SRM

Keywords Human mitochondrial DNA (mtDNA)

Human nuclear DNA (nDNA) Standard Reference Material (SRM)

Technical Information Contact for this SRM Please address technical questions about this SRM to srmsnistgov where they will be assigned to the appropriate Technical Project Leader responsible for support of this material For sales and customer service inquiries please contact srminfonistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

ii

Table of Contents Purpose and Description 1 Warning SRM 2372a is a Human Source Material 1 Storage and Use 1 History and Background 2 Experimental Methods 4

Polymerase Chain Reaction Assays 4 Droplet Digital Polymerase Chain Reaction Measurements 7

Nuclear DNA (nDNA) 7 Mitochondrial DNA (mtDNA) 8

Sample Preparation 9 Absorbance Measurements 10 SRM Unit Production 10 Component Homogeneity 10 Component Stability 11 Component Certification Measurements 12 Mitochondrial DNA Measurements 13 Performance in Commercial qPCR Chemistries 13

Measurement Results 15 Absorbance Spectrophotometry 15 Droplet Volume 16 Certification Measurements 16

Homogeneity Results 16 Stability Results 18 Ten-Assay Certification Results 20 Estimation of Copy Number Concentration 22 Conversion of Copy Number to Mass Concentration 22

Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA) 23 Performance in Commercial qPCR Chemistries 25

Qualification of ddPCR as Fit-For-Purpose 27 Digital Polymerase Chain Reaction Assays 28 Heat Treatments (Melting of the DNA into Single Strands) 28 Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a 29 Quantitative Evaluation of ssDNA Content 30

ddPCR Measurement of ssDNA Proportion 30 Proof of Principle 31 Time at Denaturation Temperature 33 Temperature 35 DNA Concentration in the ddPCR Solution 36 Equivalence of Assays 38

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials 40 Conclusions and Recommendations 43

Recommended Certification Values 43 Use of SRM 2372a for Value-Assigning In-House Reference Materials 44

Certificate of Analysis 44 References 44

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

iii

Appendix A DNA Extraction Method 46 Procedure 46 Reagents 46 Reference 47

Appendix B Spectrophotometric Measurement Protocol 48 Spectrophotometer Calibration 48

Establish Baseline 48 Enable Verification of Wavelength Calibration 48 Enable Verification of Absorbance Calibration 48

For ldquonativerdquo dsDNA Samples 48 Verify Optical Balance of Cuvettes 48

For NaOH denatured ssDNA Samples 49 Prepare a 04 molL NaOH Stock Solution 49 Prepare the Sample Solutions 49 Measure Sample Solutions 50

After All Sample Measurements are Complete 50 Enable Verification of Spectrophotometer Stability 50 Clean Up 50

Evaluation 50 Appendix C SRM 2372a Statisticianrsquos report 51

1 Introduction 51 11 OpenBUGS Analysis 51 12 OpenBUGS Exemplar Output 51 2 Homogeneity assessment 52 21 Data 52 22 Model 52 23 OpenBUGS code and data 52 24 Results 55 3 Certification measurements 57 31 Data 57 32 Model 57 33 OpenBUGS code inits and data 58 34 Results 64 4 Mito-to-nuclear DNA ratios (mtDNAnDNA) 64 41 Data 64 42 Model 64 43 OpenBUGS code inits and data 65 5 References 68

Appendix D Performance in Commercial qPCR Chemistries 69

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

iv

Table of Tables Table 1 NIST-Developed Human Nuclear DNA Assays 5 Table 2 NIST-Optimized Human Mitochondrial DNA Assays 6 Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays 8 Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays 8 Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays 8 Table 6 Tubes used for Certification Measurements 12 Table 7 Run Parameters for Mitochondrial DNA 13 Table 8 Commercial qPCR Chemistries Tested with SRM 2372a 14 Table 9 Absorbance at Selected Wavelengths 15 Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component 17 Table 11 Within- and Between-Tube Relative Standard Deviations 18 Table 12 Stability Data as λacute Copy Number per Nanoliter of Component 19 Table 13 Certification Data as λacute Copy Number per Nanoliter of Component 21 Table 14 Measured Copy Number Concentrations 22 Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution 23 Table 16 mtDNAnDNA Ratios 24 Table 17 mtDNAnDNA Summary Results 25 Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a 29 Table 19 Comparison of Snap-Cooling to Slow-Cooling 32 Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures 34 Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 36 Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution 37 Table 23 Comparison of Assays 39 Table 24 Evaluation of Limiting Ratio 41 Table 25 Recommended Values and 95 Uncertainties for Certification 43 Table C-1 Example OpenBUGS Homogeneity Summary 55 Table C-2 Example OpenBUGS Certification Summary for Component A 60 Table C-3 Example OpenBUGS Certification Summary for Component B 61 Table C-4 Example OpenBUGS Certification Summary for Component C 63 Table C-5 Parameter Values for Components A B and C 64 Table C-6 Certified Values for Components A B and C 64 Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A 66 Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B 66 Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C 67 Table C-10 Results for mtDNAnDNA 68 Table D-1 Summary [DNA] Results for the Commercial qPCR Assays 72

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

i

Table of Figures Figure 1 Sales History of SRM 2372 3 Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays 6 Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays 7 Figure 4 Plate Layout Design for Homogeneity Measurements 11 Figure 5 Plate Layout Design for Stability Measurements 11 Figure 6 Basic Plate Layout Design for Certification Measurements 12 Figure 7 Plate Layout Design for Performance Assessments 14 Figure 8 Absorbance Spectra of SRM 2372a Components A B and C 16 Figure 9 ddPCR Stability Measurements as Function of Time 18 Figure 10 Comparison of Assay Performance 20 Figure 11 mtDNAnDNA for Components A B and C 24 Figure 12 qPCR Autosomal Quantitation Performance Summary 26 Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary 26 Figure 14 Absorbance Spectra of SRM 2372 Components A and B 27 Figure 15 Comparison of Snap-Cooling to Slow-Cooling 32 Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures 33 Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 35 Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution 37 Figure 19 Comparison of Assays 38 Figure 20 Evaluation of Limiting Ratio 40 Figure 21 Degrees of Equivalence 42 Figure C-1 Homogeneity Assessment 56 Figure D-1 Results for the Innogenomics Commercial qPCR assays 69 Figure D-2 Results for the Thermo Fisher Commercial qPCR assays 70 Figure D-3 Results for the Qiagen Commercial qPCR assays 70 Figure D-4 Results for the Promega Commercial qPCR assays 71

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

1

Purpose and Description Standard Reference Material (SRM) 2372a is intended for use in the value assignment of human genomic deoxyribonucleic acid (DNA) forensic quantitation materials A unit of SRM 2372a consists of three well-characterized human genomic DNA materials solubilized in 10 mmolL 2-amino-2-(hydroxymethyl)-13 propanediol hydrochloride (Tris HCl) and 01 mmolL ethylenediaminetetraacetic acid disodium salt (disodium EDTA) pH 80 buffer (TE-4) The three component genomic DNA materials labeled A B and C are respectively derived from a single male donor a single female donor and 13 mixture of a male and a female donor A unit of the SRM consists of one 05 mL tube of each component each tube containing approximately 55 microL of DNA solution Each of these tubes is labeled and is sealed with a color-coded screw cap NIST is guided by and adheres to the ethical principles set forth in the Belmont Report1 SRM 2372a was developed after an appropriate human subject research determination by the NIST Human Subjects Protections Office

Warning SRM 2372a is a Human Source Material The three SRM 2372a components are human genomic DNA extracts prepared at NIST from buffy coat white blood cells from single-source anonymous donors All buffy coats were purchased from Interstate Blood Bank (Memphis TN) The supplier reported that each donor unit used in the preparation of this product was tested by FDA-licensed tests and found to be negative for human immunodeficiency virus (HIV-1 RNA and Anti-HIV 12) hepatitis B (HBV DNA and HBsAg) hepatitis C (HCV RNA and Anti-HCV) West Nile virus (WNV RNA) and syphilis Additionally each donor was tested according to FDA guidelines for Trypanosoma cruzi (Chagas) and found to be negativenon-reactive However no known test method can offer complete assurance that infectious agents are absent from this material Accordingly this human blood-based product should be handled at the Biosafety Level 12 SRM 2372a components and derived solutions should be disposed of in accordance with local state and federal regulations

Storage and Use Until required for use SRM 2372a should be stored in the dark between 2 degC and 8 degC The SRM 2372a component tubes should be mixed briefly and centrifuged (without opening the tube cap) prior to removing sample aliquots for analysis For the certified values to be applicable materials should be withdrawn immediately after opening the tubes and processed without delay Certified values do not apply to any material remaining in recapped tubes The certification only applies to the initial use and the same results are not guaranteed if the remaining material is used later

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

2

History and Background SRM 2372a is the second member of the SRM 2372 Human DNA Quantitation Standard series The supply of the initial material SRM 2372 was exhausted in May 2017 SRM 2372a is derived from commercially obtained buffy coat cells from different donors than those used in SRM 2372 A series of interlaboratory studies coordinated by what was then the Biotechnology Division of the National Institute of Standards and Technology (NIST) during the 1990s demonstrated that the DNA quantitation technologies then in use by the forensic human identification community adequately met the needs of the DNA typing tools then available34 However our 2004 study demonstrated that the increasingly sensitive and informative polymerase chain reaction (PCR)-based multiplex typing tools required more reliable determinations of DNA input quantity for best performance5 Quantitative PCR (qPCR) methods for DNA quantitation require calibration to an external standard of defined DNA concentration ([DNA]) expressed in units of nanograms of DNA per microliter of extract solution The true [DNA] of a commercial secondary standard may vary from its nominal concentration depending on supplier and lot number To enable providers of secondary standards to more reliably evaluate their materials and forensic testing laboratories to validate their use of commercial quantitation methods and supplies in 2007 we delivered the human DNA calibration standard SRM 2372 Human DNA Quantitation Standard6 The quantity of DNA in each of the three components of SRM 2372 was originally assigned based on the absorbance of double-stranded DNA (dsDNA) at 260 nm7 However by 2012 the absorbance properties of the materials had increased beyond the certified uncertainty limits due to slow but progressive changes in DNA tertiary structure Sales of the SRM were restricted until the DNA quantities were reassigned based on measurements of single-stranded DNA (ssDNA) in early 20138-10 Figure 1 displays the sales history of this relatively high-selling SRM While SRM 2372 was fit for the practical purpose of harmonizing DNA quantitation measurements to an accessible and reproducible reference material the conversion of both dsDNA and ssDNA absorbance to mass [DNA] is through conventional proportionality constants that are not metrologically traceable to the modern reference system for measurements the International System of Units (SI) Responding to requests from the forensic and clinical communities for calibration standards that provide traceability to the SI we and others have established droplet digital polymerase chain reaction (ddPCR) as a potentially primary measurement process that can provide traceable results11-18 The copy numbers of human nuclear DNA (nDNA) in the reaction mixture per partition symbolized as λ were determined by ddPCR using ten PCR assays optimized for use with our ddPCR platform These assays target loci on eight chromosomes with three assays on widely separated locations on one chromosome (see Table 1)

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

3

Figure 1 Sales History of SRM 2372

Each small black dot represents the sale of one unit of SRM 2372 The red lines summarize the average rate of sales as originally certified (2007 to 2012) and after re-certification (2013 to 2017) The magenta labels state the average yearly sales rate during these two periods

The nDNA copy number concentration human haploid genome equivalents (HHGE) per nanoliter component solution and symbolized here as λacute is calculated as

copy number [DNA] = λ = λ(VF) HHGEnL

where V is the average droplet volume and F is the total dilution factor the volume fraction of the component solution in the ddPCR reaction mixture The manufacturer of the ddPCR platform used a Bio-Rad QX200 Droplet Digital PCR System (Hercules CA) system does not provide metrologically traceable droplet volume information This lack of information prevented metrologically traceable conversion of the number per partition measurements to the desired number per reaction volume units In consequence NIST developed a measurement method that provides metrologically traceable droplet volumes19 We show elsewhere that the mass [DNA] nanograms of human DNA per microliter component can be traceably estimated as

mass [DNA] = 3031timesλacute ngmicroL

where the constant is the product of the number of nucleotide basepairs (bp) in the human reference genome the average bp molecular weight Avogadrorsquos number of bp per mole and

2372

2372(Recertif ied)

050

010

00

1713 unitsyear

1416 unitsyear

2007 2009 2011 2013 2015 2017

Uni

ts S

old

Calendar Year

Human DNA Quantitation StandardSRM 2372

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

4

unit scaling factors18 The three components of SRM 2372a are certified for both copy number and mass [DNA] Mitochondrial DNA (mtDNA) analysis within the forensic community is useful when the nDNA may be degraded or is in low-to-undetectable quantity The challenges of mtDNA testing include contamination of the standard due to routine laboratory use degradation of plasmids or synthetic oligomers or improper commutability of the standard relative to casework samples Commonly cell line materials are used as standards for qPCR with human genomic DNA which may lead to an incorrect mtDNA-to-nDNA ratio (mtDNAnDNA) for unknown samples Non-certified values of mtDNAnDNA for the three components are provided to bridge the current gap between well characterized mitochondrial DNA quantification assays and the availability of commercial standards With the understanding that the forensic DNA community primarily uses qPCR as a method for DNA quantification the performance of the three SRM 2372a materials was evaluated in eleven commercial qPCR chemistries currently used employed within the forensic DNA community

Experimental Methods Production and value-assignment of any certified reference material (CRM) is a complex process In addition to acquiring processing and packaging the materials included in the CRM the materials must be sufficiently homogenous across the units stable over time within its packaging and well-characterized to deliver results that are fit-for-purpose within the intended measurement community20 The process has been particularly complicated for SRM 2372a because we needed to determine the metrological properties of the required ddPCR measurement systems before we could trust the results they provide The following sections describe the methods and processes used to produce and characterize the SRM 2372a materials including sample preparation SRM unit production PCR assays ddPCR measurement processes spectrophotometric absorbance measurements homogeneity testing stability testing ddPCR nDNA and mtDNAnDNA measurements and performance in commercial qPCR assessments Many of these processes were mutually dependent therefore the sections are presented in a ldquodefinitionalrdquo order that sequentially introduces terminology Polymerase Chain Reaction Assays

Ten PCR assays were developed for characterization of SRM 2372a Each assay was confirmed to target one locus per HHGE using the National Center for Biotechnology Informationrsquos BLASTblastn system21 Primers were purchased from Eurofins Operon (Huntsville AL) TaqMan probes labeled with 6-carboxyfluorescein (FAM) were purchased from Thermo Fisher (Waltham MA) FAM-labeled Blackhole Quencher and Blackhole Quencher+ probes were purchased from LGC Biosearch Technologies (Novato CA) Table 1 details the ten nDNA assays used to certify copy number and mass [DNA] of the SRM 2372a components FAM-labeled TaqMan probes were purchased from Thermo Fisher

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

5

and used in monoplex reactions VIC (a proprietary dye)-labeled probes were purchased from the same source and were used when nDNA and mtDNA were analyzed in the same reaction mixture Figure 2 displays exemplar droplet patterns for the assays All assays were developed at NIST and have been optimized for ddPCR18 Table 2 details three mitochondrial DNA (mtDNA) assays used to estimate mtDNAnDNA in the SRM 2372a components Figure 3 displays exemplar droplet patterns for the assays These assays were optimized for ddPCR from real-time quantitative PCR (qPCR) assays from the literature All primers probes were diluted with TE-4 buffer purchased from Affymetrix-USB Products (Affymetrix Inc Cleveland OH) to a nominal 5 micromol working solution for all single-plex assays (individual nDNA and mtDNA assays) For duplex assays (combined nDNA and mtDNA) primer and probe working solutions were diluted to a nominal 10 micromol

Table 1 NIST-Developed Human Nuclear DNA Assays

Assay Target

Chromosome Band Accession Primers and Probe a

Amplicon Length bp

NEIF Gene EIF5B

Chr 2 p111-q111 NC_00000212

F gccaaacttcagccttctcttc R ctctggcaacatttcacactaca PB+ tcatgcagttgtcagaagctg

67

2PR4 Gene RPS27A

Chr 2 p16 NC_00000212

F cgggtttgggttcaggtctt R tgctacaatgaaaacattcagaagtct PB tttgtctaccacttgcaaagctggccttt

97

POTP STR TPOX

Chr 2 p253 NC_00000212

F ccaccttcctctgcttcacttt R acatgggtttttgcctttgg PT caccaactgaaatatg

60

NR4Q Gene DCK

Chr 4 q133-q211 NC_00000412

F tggtgggaatgttcttcagatga R tcgactgagacaggcatatgtt PB+ tgtatgagaaacctgaacgatggt

83

D5 STR D5S2500

Chr 5 q112 NC_00000510

F ttcatacaggcaagcaatgcat R cttaaagggtaaatgtttgcagtaatagat PT ataatatcagggtaaacaggg

75

ND6 STR D6S474

Chr 6 q21-22 NC_00000612

F gcatggctgagtctaaagttcaaag R gcagcctcagggttctcaa PB+ cccagaaccaaggaagatggt

82

D9 STR D9S2157

Chr 9 q342 NC_00000912

F ggctttgctgggtactgctt R ggaccacagcacatcagtcact PT cagggcacatgaat

60

HBB1 Gene HBB

Chr 11 p155 NC_00001110

F gctgagggtttgaagtccaactc R ggtctaagtgatgacagccgtacct PT agccagtgccagaagagccaagga

76

ND14 STR D14S1434

Chr 14 q3213C NC_0000149

F tccaccactgggttctatagttc R ggctgggaagtcccacaatc PB+ tcagactgaatcacaccatcag

109

22C3 Gene PMM1

Chr 22 q132 NC_00002210

F cccctaagaggtctgttgtgttg R aggtctggtggcttctccaat PB caaatcacctgaggtcaaggccagaaca

78

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

6

a F Forward primer R Reverse primer PB Blackhole quencher probe (FAM labeled) PB+ Blackhole Plus quencher probe (FAM labeled) PT TaqMan MGB probe (FAM labeled)

Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays

Table 2 NIST-Optimized Human Mitochondrial DNA Assays

Assay Location Primers and Probe a Amplicon Length bp Ref

mtND1 34853504 F ccctaaaacccgccacatct

69 22 35333553 R gagcgatggtgagagctaaggt 35063522 PT ccatcaccctctacatc

mtBatz 84468473 F aatattaaacacaaactaccacctacct

79 23 85038524 R tggttctcagggtttgttataa 84758491 PT cctcaccaaagcccata

mtKav 1318813308 F ggcatcaaccaaccacaccta

105 24 1336913392 R attgttaaggttgtggatgatgga 1331013325 PT cattcctgcacatctg

a F Forward primer R Reverse primer PT TaqMan MGB probe (FAM labeled)

2PR4

POTP

NEI

F

R4Q

5

D5

ND6

D9

HBB1

ND1

4

22C3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

7

Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays

Droplet Digital Polymerase Chain Reaction Measurements

Nuclear DNA (nDNA) All nDNA ddPCR measurements used in this study followed the same measurement protocol

bull Droplets were generated on the Auto Droplet Generator using the Bio-Rad ldquoddPCR ldquoSupermix for Probes (No dUTP)rdquo (Bio-Rad cat 186-3024 Control 64079080) and ldquoDroplet Generation Oil for Probesrdquo (Bio-Rad cat 186-4110 Control 64046051) Note desoxyuridine 5-triphosphatase (dUTP) is an enzyme sometimes included in PCR reaction mixtures to minimize PCR carryover contamination

bull Table 3 details the composition of the ddPCR mastermix setup for a given assay and sample where ldquomastermix refers to reaction mixture of manufacturerrsquos proprietary PCR reagent ldquoSupermixrdquo forward and reverse PCR primers for a given assay probe for that assay the sample DNA and sometimes additional water to produce the desired value dilution factor F This setup was used for all assays and samples

bull Non-Template Controls (NTCs) for each assay were included in every analysis bull Droplets were thermal cycled on a ProFlex thermal cycler (Applied Biosystems) for

95 degC for 10 min followed by 60 cycles of 94 degC for 05 min and 61 degC for 1 min then 98 degC for 10 min before a 4 degC hold until the plate was removed

bull Droplets were read on the QX200 Droplet Reader and analysis was performed using the QuantaSoft Analysis Software version 1740917

mtND1 mtBatz mtKav

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

8

bull The numbers of positive and negative droplets at the end of 60 cycles were determined and were exported into a spreadsheet for further analysis Assay-specific intensity thresholds were determined by visual inspection for each assay for each measurement session

Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays

Concentration nDNA Mastermix Microliter

per Reaction 2X ldquoSupermix for Probes (no dUTPs)rdquo 125

5 μmolL Forward Primer 1875 5 μmolL Reverse Primer 1875 5 μmolL Probe (FAM) 125

PCR Grade Water 50 14 Diluted Sample 25 Total Volume 250

Mitochondrial DNA (mtDNA) All the mtDNA measurements used in this study followed the nDNA protocol described above but with different mastermix setups The duplex mtDNA measurements used the setup described in Table 4 The monoplex mtDNA measurements used the setup described in Table 5

Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 125

10 μmolL Forward Primer 09375 10 μmolL Reverse Primer 09375 10 μmolL Probe (VIC) 0625

PCR Grade Water 95 nDNA Mastermix from Table 3 05 Total Volume 250

Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 120

10 μmolL Forward Primer 09 10 μmolL Reverse Primer 09 10 μmolL Probe (FAM) 06

PCR Grade Water 72 Diluted Sample (See Table 7) 24

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

9

Total Volume 240 Sample Preparation

The buffy coat white blood cells used to prepare SRM 2372a materials were purchased from Interstate Blood Bank These materials were from four single-source anonymous individuals supplied to us labeled as Female1 Female2 Male1 and Male2 In our laboratories each donor buffy coat bag was aliquoted (5 mL per aliquot) into sterile 50 mL conical tubes and stored at 4 degC prior to extraction The modified salt-out manual extraction protocol described in Appendix A was performed for each aliquot with rehydration of the DNA in TE-4 buffer From 150 mL to 250 mL of solubilized DNA for each component was accumulated through the extraction of the 5 mL aliquots The DNA was stored in perfluoroalkoxy polymer (PFA) containers that had been bleach sterilized thoroughly rinsed with deionized water rinsed with ethanol and air-dried The minimum volume for production of 2000 units was 120 mL Preliminary estimates of the mass [DNA] of the re-solubilized components were obtained using a Nanodrop microvolume spectrophotometer Dilutions of the component materials were made until the absorbance at 260 nm was approximately 1 corresponding to a nominal mass [DNA] of 50 ngμL of dsDNA7 After addition of the TE-4 buffer the containers were placed on an orbital shaker and gently rotated for several hours prior to refrigerated storage at 4 ordmC The four single-source components were initially screened with the nDNA ddPCR measurement process using four nDNA assays 2PR4 NR4Q ND6 and 22C3 Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate The gravimetric 1rarr4 dilution was prepared by weighing an empty PFA vial and recording the weight then pipetting 30 microL TE-4 into the vial and recording the weight then pipetting 10 microL DNA and recording the final weight The dilution factor was calculated in a spreadsheet These screening measurements verified that the individual component materials were fit-for-purpose The four single-source components were evaluated with the nDNA ddPCR measurement process using the ten nDNA assays Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate Quantitative information was derived from the average of the ten nDNA assays to prepare the component C mixture Thirty milliliters of sample Male2 (estimated at 452 ngμL mass [DNA]) were combined with 80 mL of sample Female2 (estimated at 507 ngμL mass [DNA]) and 10 mL of TE-4 buffer to create a 1 male nDNA to 3 female nDNA producing a 14 malehuman mixture After addition of the TE-4 buffer the component C container was gently rotated on the orbital shaker for several hours prior to refrigerated storage at 4 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

10

Absorbance Measurements

In February 2017 absorbance spectra of the recently prepared fully diluted nominally dsDNA and ssDNA solutions of the three components were acquired with a BioCary 100 spectrophotometer using the measurement protocol described in Appendix B In December 2017 additional spectra were acquired for the remaining bulk solution of components A and B and for SRM 2372 components A and B These solutions had been stored at 4 degC in well-sealed PFA containers There was no remaining bulk component C solution SRM Unit Production

In the morning of the day a component was transferred into the sterile component tubes (Sarstedt Biosphere SC micro tube-05 mL Newton NC) used to package the SRM the container holding the diluted component was removed from the refrigerator and placed on a slow orbital shaker for 2 h The tubes were opened and placed under the laminar flow hood in rows of 5 within 80-hole tube racks for a total of 30 tubes per rack Tubes were placed in every second row of the racks to facilitate filling with an Eppendorf Repeater Xstream pipette (Eppendorf North America Inc Hauppauge NY) fitted with a 1-mL positive displacement tip set to dispense 55 microL per tube The component container was removed from the shaker a magnetic stir-bar was added the container was placed on a magnetic stir plate in the laminar flow hood and the material was stirred gently Filling proceeded until 2200 vials were filled with pipet tip replacement after filling every 200 tubes The filling process consisted of two individuals manually filling each vial within a biosafety cabinet (one physically pipetting each sample the other verifying that each tube was filled) and four individuals tightly closing the lids to each tube on a sterilized and covered bench When all tubes in a rack were filled the rack was moved out of the hood the tubes checked for proper filling their lids closed and the tubes transferred to 100-unit pre-labeled storage boxes As the 100-unit boxes were filled they were transferred to the labeling room where SRM labels were applied by hand and colored inserts were added to the lids red for component A white for component B and blue for component C After all tubes for a given component were labeled the 22 boxes were placed in a locked refrigerator Filling of each of the components took 60 min to 90 min All units were transferred to the refrigerator within 4 hours of beginning the filling process All units of the three materials were allowed to equilibrate for a minimum of two weeks prior to beginning homogeneity analysis Component Homogeneity

The ddPCR measurement protocol described above was used to assess the nDNA homogeneity of the three components One tube from the approximate center of each of the 22 storage boxes was assayed using the NEIF and ND6 assays (see Table 1) A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Three technical replicates were obtained for every tube A total

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

11

of six plates were used two plates per component one with the NEIF assay and one with the ND6 assay Figure 4 details the layout of the samples

Figure 4 Plate Layout Design for Homogeneity Measurements

Component Stability

Stability of the materials was evaluated at 4 degC room temperature (RT asymp 22 degC) and 37 degC Tubes from the homogeneity study were divided among the three temperatures for the stability study Additionally two never-opened tubes (one from box 8 and one from box 12 of each component) were held at each temperature Two tubes at each temperature for each component were opened sampled and resealed 4 times over the course of 8 weeks Measurements were made at week 1 week 3 week 5 and week 8 The measurement on week 8 was from an unopened tube (from box 8) for each of the components at all temperatures The data from the homogeneity study was used as timepoint 0 in the stability study anchored to 4 degC by the storage recommendations A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used for both the NEIF and ND6 assays All temperatures components and both assays fit onto one plate for timepoints 1 to 5 Twelve NTCs placed in rows G and H were included for each assay Figure 5 details the sample layout

Figure 5 Plate Layout Design for Stability Measurements

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

ND6 ND6 ND6

NEIFNEIFNEIFComponent A

Component A

Component B

Component B

Component C

Component C

1 2 3 4 5 6 7 8 9 10 11 12A A-7 A-7 B-7 B-7 C-7 C-7 A-7 A-7 B-7 B-7 C-7 C-7B A-10 A-10 B-10 B-10 C-10 C-10 A-10 A-10 B-10 B-10 C-10 C-10C A-9 A-9 B-9 B-9 C-9 C-9 A-9 A-9 B-9 B-9 C-9 C-9D A-12 A-12 B-12 B-12 C-12 C-12 A-12 A-12 B-12 B-12 C-12 C-12E A-8 A-8 B-8 B-8 C-8 C-8 A-8 A-8 B-8 B-8 C-8 C-8F A-11 A-11 B-11 B-11 C-11 C-11 A-11 A-11 B-11 B-11 C-11 C-11G NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTCH NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

NEIF

4C

RT

37C

ND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

12

Component Certification Measurements

After establishing homogeneity and stability measurements intended to be used to certify nDNA values and inform mtDNA values were started using randomly selected SRM tubes from the boxes listed in Table 7 The ten nDNA assays described in Table 1 and two of the three mtDNA assays described in Table 2 were run simultaneously on the same sample plate A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used in each of the six measurement campaigns

Table 6 Tubes used for Certification Measurements

Component Plate A B C

dd20170711 Box 1 Box 1 Box 1 dd20170712a Box 10 tube a Box 10 tube a Box 10 tube a dd20170712b Box 10 tube b Box 10 tube b Box 10 tube b dd20170713 Box 20 Box 20 Box 20 dd20170714 Box 5 Box 5 Box 5 dd20170801 Box 5 Box 5 Box 5

Every sample was analyzed in duplicate with every assay in all six plates Component A was always located in rows B and C component B in rows D and E component C in rows F and G and NTCs in rows A and H Figure 6 displays the basic layout of the ten nDNA and two mtDNA assays In addition to the plate column assignment scheme shown the three assay groupings were also ordered as Group 2 Group 3 Group 1 and Group 3 Group 1 Group 2 to address potential positional bias

Group 1 Group 2 Group 3 2PR4 POTP NEIF NR4Q D5 ND6 D9 HBB1 ND14 22C3 mtND1 mtBatz 1 2 3 4 5 6 7 8 9 10 11 12 A NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Comp A B Comp A C Comp B D Comp B E Comp C F Comp C G

H NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Figure 6 Basic Plate Layout Design for Certification Measurements

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

13

Mitochondrial DNA Measurements

A ldquoduplexed reactionrdquo was used to determine mtDNAnDNA where 1 μL of one of the nDNA mastermix was added to 49 μL of the mtDNA mastermix resulting in a 1rarr50 volumetric dilution of the nDNA The nDNA was targeted with a FAM-labeled probe while the mtDNA was targeted with a VIC labeled probe The primer and probe concentrations were increased from 5 micromol to 10 micromol in the duplexed reactions Additional testing of the mtDNA assays were performed in monoplex with FAM-labeled probes to determine the additional sources of bias between sampling assay setup and dilution factors Table 7 outlines the different methods examined to determine mtDNAnDNA for the three components

Table 7 Run Parameters for Mitochondrial DNA Droplet

Generation nDNA mtDNA

Plate Component Assay Dilution Reps Assay Dilution Reps dd20170824 A manualDG HBB1 1rarr4 3 mtND1 1rarr200 9 dd20170825 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170829 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170831 AB autoDG HBB1 1rarr3 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr6 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr12 2 mtND1 1rarr10rarr100 3 dd20170907 ABC manualDG HBB1 1rarr3 2 mtBatz 1rarr10rarr100 3 dd20170913 AB manualDG NEIF 1rarr4 3 mtND1 mtBatz mtKav 1rarr10rarr100 3 dd20170913 C manualDG NEIF 1rarr4 3 mtBatz mtKav 1rarr10rarr100 3 Performance in Commercial qPCR Chemistries

Table 8 lists the 11 qPCR chemistries that were commercially available in late 2017 All three of the SRM 2372a components were evaluated with these chemistries using component A of SRM 2372 as the reference standard For each component of SRM 2372a four tubes were pulled from box 22 and combined into one tube A 1rarr5 dilution of all samples was prepared followed by three serial 1rarr2 dilutions A five-point serial dilution of each SRM 2372a component was amplified in triplicate for each chemistry This consisted of the neat material the 1rarr5 dilution then the three serial 1rarr2 dilutions A six-point serial dilution of SRM 2372 Component A was also created for the standard curve for the commercial qPCR chemistries This consisted of the neat material followed by a 1rarr5 dilution then followed by four 1rarr2 dilutions serially The six-point serial dilution series of SRM 2372 Component A was also amplified in triplicate These commercial chemistries were amplified in individual plates on an Applied Biosystems 7500 HID Real-Time PCR Instrument (Foster City CA) following the manufacturersrsquo recommended protocols for reaction mix setup and cycling conditions Figure 7 details the plate layout for the qPCR plates Data was analyzed with the HID Real-Time PCR Analysis Software v12 and further analyzed in a spreadsheet

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

14

Table 8 Commercial qPCR Chemistries Tested with SRM 2372a

Manufacturer Commercial Kit Target Codea

Innogenomics

InnoQuant Long IQh Short IQd

InnoQuantHY Long IQHYh Short IQHYd Y IQHYy

Thermo Fisher Quantifiler Duo Human QFDh Male QFDy

Quantifiler HP Large QFHPh Small QFHPd

Quantifiler Human Human QFHh

Quantifiler Trio Large QFTh Small QFTd Y QFTy

Qiagen

Quantiplex Autosomal QPh

Quantiplex Hyres Autosomal QPHh Y QPHy

Quantiplex Pro Human QPPh Degradation QPPd Y QPPy

Promega

Plexor Autosomal Ph Y Py

PowerQuant Autosomal PQh Degradation PQd Y PQy

a Assays are coded with a short acronym derived from the kit name followed by the following codes for the type of target ldquohrdquo autosomal non-degradation described as ldquoAutosomalrdquo ldquoHumanrdquo or ldquoLongrdquo ldquodrdquo autosomal degradation described as ldquoDegradationrdquo or ldquoShortrdquo ldquoyrdquo Y-chromosome described as ldquoMalerdquo or ldquoYrdquo

1 2 3 4 5 6 7 8 9 10 11 12

A

B 2372-A_57 2372-A_57 2372-A_57 2372a-neat_A

2372a-neat_A

2372a-neat_A

2372a-neat_B

2372a-neat_B

2372a-neat_B

2372a-neat_C

2372a-neat_C

2372a-neat_C

C 2372-A_11

4 2372-A_11

4 2372-A_11

4 2372a-15_A

2372a-15_A

2372a-15_A

2372a-15_B

2372a-15_B

2372a-15_B

2372a-15_C

2372a-15_C

2372a-15_C

D 2372-A_5

7 2372-A_5

7 2372-A_5

7 2372a-110_A

2372a-110_A

2372a-110_A

2372a-110_B

2372a-110_B

2372a-110_B

2372a-110_C

2372a-110_C

2372a-110_C

E 2372-A_2

85 2372-A_2

85 2372-A_2

85 2372a-120_A

2372a-120_A

2372a-120_A

2372a-120_B

2372a-120_B

2372a-120_B

2372a-120_C

2372a-120_C

2372a-120_C

F 2372-A_1

425 2372-A_1

425 2372-A_1

425 2372a-140_A

2372a-140_A

2372a-140_A

2372a-140_B

2372a-140_B

2372a-140_B

2372a-140_C

2372a-140_C

2372a-140_C

G 2372-A_0

7125 2372-A_0

7125 2372-A_0

7125

H NTC NTC NTC

Figure 7 Plate Layout Design for Performance Assessments

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

15

Measurement Results After preparing the SRM 2372a bulk solutions their approximate 50 ngmicroL mass [DNA] were confirmed spectrophotometrically The bulk solutions were then packaged into individual tubes The materials contained in the tubes were characterized for homogeneity stability nDNA copy number [DNA] and mtDNAnDNA ratio The following sections detail the results of these measurements Absorbance Spectrophotometry

By widely accepted convention an aqueous DNA solution having an absorbance of 10 at 260 nm corresponds to a mass [DNA] of 50 ngmicroL for dsDNA and (37 to 40) ngmicroL for ssDNA7 Absorbance at (230 270 280 and 330) nm are traditionally used in the assessment of DNA quality725 Due to minor baseline shifts during acquisition of some spectra it was necessary to bias-adjusted all spectra to have absorbance of 00 at 330 nm Table 9 lists the resulting bias-adjusted values for selected wavelengths and the conventional mass [DNA] conversion of the absorbance at 260 nm The close agreement between the mass concentration values for the nominally dsDNA and the converted-to-ssDNA spectra confirm that the bulk solutions were correctly prepared to be asymp 50 ngμL dsDNA The slight increase in absorbance over the 10 months between acquiring the spectra suggests that like SRM 2372 the tertiary structure of the dsDNA in the SRM 2372a materials may be changing slowly with time Figure 8 displays the spectra

Table 9 Absorbance at Selected Wavelengths

Wavelength nm Form Date Component 230 260 270 280 ngμL dsDNA 2142017 A 0394 0931 0758 0488 466

B 0480 1112 0905 0583 556 C 0451 1006 0815 0524 503

ssDNA 2142017 A 0390 0614 0540 0316 454 B 0457 0723 0634 0372 535 C 0411 0641 0563 0330 474

dsDNA 12142017 A 0402 0956 0775 0503 478 B 0484 1128 0914 0589 564

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

16

Figure 8 Absorbance Spectra of SRM 2372a Components A B and C

dsDNA is denatured to ssDNA by diluting with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to DNA mass concentration

Droplet Volume

Droplet volumes for the two lots of ldquoSupermixrdquo used in the studies described in this report were determined using NISTrsquos Special Test Method 11050S-D as described in Dagata et al19 The lot used for most of the ddPCR analyses produced (07349 plusmn 00085) nL droplets where the ldquoplusmnrdquo value is the combined standard uncertainty The lot used for analyses performed after October 11 2017 produced (07376 plusmn 00085) nL droplets NIST measurements have demonstrated that droplet volumes are not influenced by the presence or absence of DNA in the sample solution Measurements repeated over time on an earlier lot of the Supermix product used in these studies suggest that droplet volumes remain constant over the shelf-life of the lot However droplet volumes for different lots of this product as well as for different Supermix products have been observed to differ by several percent We intend to document these measurements in a future report Other researchers have demonstrated that non-constant droplet volume distributions may contribute to ddPCR measurement imprecision and bias2627 However these effects become significant only with distributions wider than observed using Special Test Method 11050S-D19 and at copydroplet values larger than the λ le 05 copydroplet typically used in our measurements Certification Measurements

Homogeneity Results Table 10 lists the λacute nDNA copy number values adjusted for both droplet volume and sample dilution from the homogeneity measurements The data was analyzed using the three-level Gaussian hierarchical ANOVA model described in Appendix C Table 11 lists the values of the within- and between-tube variance components expressed as percent relative standard deviations At the 95 confidence level all components are homogeneous

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372a-A2141721417-NaOH121417

220 260 300 340

Wavelength nm

2372a-B2141721417-NaOH121517

220 260 300 340

Wavelength nm

2372a-C21417

21417-NaOH

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

17

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 NEIF A-1-01 1632 1529 1511 B-1-01 1790 1909 1885 C-1-01 1493 1385 1510 ND6 1616 1640 1582 NAe 1870 1860 1565 1562 1466

NEIF A-1-02 1531 1525 1549 B-1-02 1712 1719 1704 C-1-02 1516 1513 1577 ND6 1615 1589 1601 1775 1718 1756 1490 1547 1570

NEIF A-1-03 1524 1540 1533 B-1-03 1693 1755 1749 C-1-03 1451 1528 1506 ND6 1582 1152 1555 1824 1483 1896 1485 1505 1471

NEIF A-1-04 1454 1458 1427 B-1-04 1754 1745 1739 C-1-04 1440 1486 1520 ND6 1627 1605 1607 1670 1632 1814 1529 1508 1506

NEIF A-1-05 1465 1540 1504 B-1-05 1757 1687 1735 C-1-05 1523 1472 1483 ND6 1607 1633 1586 1683 1759 1703 1513 1520 1491

NEIF A-1-06 1512 1491 1491 B-1-06 1697 1709 1714 C-1-06 1437 1493 1522 ND6 1511 1483 1524 1693 1723 1722 1492 1481 1461

NEIF A-1-07 1418 1480 1455 B-1-07 1722 1625 1703 C-1-07 1485 1477 1463 ND6 1479 1557 1656 1690 1684 1688 1475 1547 1485

NEIF A-1-08 1496 1560 1501 B-1-08 1884 1802 1772 C-1-08 1433 1405 1478 ND6 1514 1615 1540 1864 1887 1821 1497 1501 1495

NEIF A-1-09 1588 1581 1637 B-1-09 1713 1734 1727 C-1-09 1463 1499 1505 ND6 1474 1438 1454 1709 1756 1758 1503 1491 1516

NEIF A-1-10 1574 1513 1570 B-1-10 1759 1809 1738 C-1-10 1534 1458 1454 ND6 1594 1713 1669 1764 1816 1787 1448 1465 1474

NEIF A-1-11 1557 1508 1633 B-1-11 1749 1782 1740 C-1-11 1533 1514 1555 ND6 1571 1611 1631 1741 1792 1784 1488 1470 1483

NEIF A-1-12 1500 1476 1515 B-1-12 1748 1712 1718 C-1-12 1450 1525 1497 ND6 1626 1586 1608 1797 1614 1790 1496 1483 1519

NEIF A-1-13 1672 1634 1804 B-1-13 1632 1668 1701 C-1-13 1537 1527 1515 ND6 1490 1516 1497 1626 1566 1681 1499 1479 1475

NEIF A-1-14 1551 1568 1538 B-1-14 1752 1535 1673 C-1-14 1515 1490 1494 ND6 1617 1614 1639 1724 1741 1685 1469 1480 1466

NEIF A-1-15b 1793 1722 1733 B-1-15 1871 1835 1887 C-1-15 1511 1568 1594 ND6 1818 1753 1762 1770 1752 1798 1515 1543 1565

NEIF A-1-16 1573 1520 1617 B-1-16 1688 1725 1743 C-1-16 1449 1452 1471 ND6 1634 1627 1575 1744 1697 1710 1543 1448 1491

NEIF A-1-17 1516 1516 1628 B-1-17 1789 1769 1825 C-1-17 1388 1421 1394 ND6 1613 1567 1567 1716 1762 1638 1452 NAe 1435

NEIF A-1-18 1612 1535 1672 B-1-18 1826 1780 1849 C-1-18 1475 1455 1445 ND6 1529 1495 1528 1805 1801 1931 1458 1531 1459

NEIF A-1-19 1598 1514 1596 B-1-19 1751 1759 1880 C-1-19 1468 1525 1446 ND6 1571 1535 1511 1827 1766 1786 1435 1478 1421

NEIF A-1-20 1575 1568 1586 B-1-20 1738 1678 1768 C-1-20 1549 1531 1499 ND6 1652 1545 1627 1752 1747 1768 1499 1465 1468

NEIF A-1-21 1658 1611 1578 B-1-21 1742 1836 1713 C-1-21 1530 1440 1484 ND6 1647 1607 1612 1746 1744 1764 1499 1454 NAe

NEIF A-1-22 1547 1528 1532 B-1-22 1761 1743 1660 C-1-22 1442 1487 1436 ND6 1537 1513 1536 1666 1692 1819 1589 1591 1556

NEIF A-1-15c 1456 1448 1523 ND6 1525 1402 1436

NEIF A-2-15d 1679 1599 1674 ND6 1581 1550 1584

a Tube coded as Component-Tube-Box b These results were identified as a probable error in the preparation or volumetric delivery of the sample

solution into the microfluidic device and were not used in the homogeneity evaluations c Repeat assessment of the solution in sample A-1-15 d Results for a second tube from box 15 analyzed at the same time as the repeat assessment of A-1-15 e Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

18

Table 11 Within- and Between-Tube Relative Standard Deviations

Within-Tube Component NEIF ND6 Between-Tube

A 28 40 33 B 27 38 25 C 23 18 19

Stability Results Table 12 lists the λacute volume- and dilution-adjusted nDNA copy number values from the stability measurements Figure 9 summarizes these results The results for the tubes held at 4 degC room temperature (RT asymp 22 degC) and 37 degC are very similar for all three components The absence of between-temperature differences identifies the between-week differences as attributable to variation in the measurement processes rather than sample instability These results indicate that the SRM 2372a genomic copy number is thermally stable from 4 degC to 37 degC over an extended period

Figure 9 ddPCR Stability Measurements as Function of Time

Each dot-and-bar symbol represents the mean λacute and the approximate 95 expanded uncertainty for all results for a given component at a given storage temperature The thick black horizontal lines represent an approximate 95 confidence interval on the λacute values estimated from the homogeneity data

13

14

15

16

17

18

19

0 1 2 3 4 5 6 7 8

λacute H

aplo

id n

DNA

Copi

es μL

Weeks

A

0 1 2 3 4 5 6 7 8

Weeks

B

4 degC 22 degC 37 degC

0 1 2 3 4 5 6 7 8

Weeks

C

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

19

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Week Temp Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3

1 4 degC NEIF A-10 1568 1441 B-10 1666 1653 C-10 1424 1394 ND6 A-10 1474 1494 B-10 1655 1705 C-10 1461 1394

NEIF A-07 1500 1463 B-07 1718 1640 C-07 1446 1452 ND6 A-07 1479 1479 B-07 1708 1686 C-07 1522 1482

RT (22 degC) NEIF A-12 1674 1608 B-12 1805 1777 C-12 1506 1495 ND6 A-12 1522 1507 B-12 1759 1788 C-12 1478 1461 NEIF A-09 1635 1539 B-09 1791 1793 C-09 1411 NAb ND6 A-09 1543 1519 B-09 1661 1682 C-09 1482 1496

37 degC NEIF A-11 1614 1637 B-11 1748 1744 C-11 1515 1464 ND6 A-11 1494 1502 B-11 1740 1697 C-11 1495 1535 NEIF A-08 1592 1614 B-08 1691 1725 C-08 1528 1500 ND6 A-08 1581 1595 B-08 1727 1699 C-08 1458 1483 3 4 degC NEIF A-10 1520 1706 B-10 NAb 1733 C-10 1427 1473 ND6 A-10 1543 1624 B-10 1670 1755 C-10 1443 1364 NEIF A-07 1537 1570 B-07 1549 1565 C-07 1424 1512 ND6 A-07 1517 1479 B-07 1705 1723 C-07 1483 1439

RT (22 degC) NEIF A-12 1521 1607 B-12 1817 1842 C-12 1566 1471 ND6 A-12 1644 1597 B-12 1845 1786 C-12 1504 1511 NEIF A-09 1432 1626 B-09 1825 1856 C-09 1392 1387 ND6 A-09 1539 1528 B-09 1734 1737 C-09 1535 1465

37 degC NEIF A-11 1651 1685 B-11 1759 1779 C-11 1437 1440 ND6 A-11 1680 1594 B-11 1806 1806 C-11 1471 1509 NEIF A-08 1580 1613 B-08 1740 1791 C-08 1582 1590 ND6 A-08 1592 1524 B-08 1769 1694 C-08 1468 1520 5 4 degC NEIF A-10 1486 1440 B-10 1626 1551 C-10 1381 1330 ND6 A-10 1555 1577 B-10 1673 1625 C-10 1547 1395 NEIF A-07 1387 1429 B-07 1723 1787 C-07 1330 1314 ND6 A-07 1538 1522 B-07 1862 1909 C-07 1452 1458

RT (22 degC) NEIF A-12 1601 1631 B-12 1718 1752 C-12 1546 1541 ND6 A-12 1699 1677 B-12 1772 1721 C-12 1490 1471 NEIF A-09 1381 1484 B-09 1675 1694 C-09 1371 1376 ND6 A-09 1625 1615 B-09 1904 1815 C-09 1468 1472

37 degC NEIF A-11 1456 1517 B-11 1720 1704 C-11 1505 1493 ND6 A-11 1650 1658 B-11 1148 1152 C-11 1527 1600 NEIF A-08 1590 1666 B-08 1745 1723 C-08 1469 1540 ND6 A-08 1235 1244 B-08 1809 1749 C-08 1508 1529 8 4 degC NEIF A-08a 1667 1632 1589 B-08a 1880 1884 1940 C-08a 1525 1475 1542 ND6 A-08a 1599 1593 1590 B-08a 1859 1897 1972 C-08a 1550 1502 1520

RT (22 degC) NEIF A-08b 1693 1666 1634 B-08b 1769 1884 1771 C-08b 1595 1557 1545 ND6 A-08b 1644 1620 1647 B-08b 1862 1738 1811 C-08b 1611 1585 1535

37 degC NEIF A-08c 1620 1576 1608 B-08c 1934 1952 1956 C-08c 1595 1600 1569 ND6 A-08c 1576 1553 1532 B-08c 1889 1844 1889 C-08c 1569 1514 1581

a Tube coded as Component-Tube-Box b Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

20

Ten-Assay Certification Results The original intention was to certify the copy number [DNA] of the three components with results from five independently prepared sets of samples analyzed over four days Each sample of the three components was from a new tube from different boxes and was evaluated with 10 nDNA assays The assay positions were alternated across these five ldquo10-assay certificationrdquo plates The λ results for these five plates are systematically somewhat larger than those for the homogeneity and stability data On review we realized that a different workspace and its suite of pipettes had been used to prepare the 1rarr10 dilution of the 1rarr4 diluted components into the reaction mastermix The ldquohighrdquo workstation was used for stability and homogeneity measurements while the ldquolowrdquo workstation was used for certification measurements Gravimetric testing of the pipettes established that the volumes delivered by these alternate pipettes was sufficiently biased to account for the observed λ differences A sixth ldquocertificationrdquo plate was prepared using the original (and now re-verified) pipettes The λ results or this plate agreed well with the homogeneity and stability data Table 13 lists the λacute volume- and dilution-adjusted nDNA copy number values for all six of the ten-assay datasets Figure 10 displays the extent of agreement among the ten assays While there are small-but-consistent differences between the assays there is only a 4 difference between the assays providing smallest and largest λ

Figure 10 Comparison of Assay Performance

The blue dot-and-bars represent the mean and its combined standard uncertainty for the A B and C components for each assay relative to the over-all grand mean for each component The black dot-and-bars represent combination of three component results for each assay The solid horizontal line represents the median difference of the assay means from the grand mean the dotted horizontal lines bound an approximate 95 confidence region about the median

2PR4POTPNEIFR4Q5 D5ND6 D9 HBB1ND14 22C3-10

-8

-6

-4

-2

0

2

4

6

8

10

Rela

tive

Diffe

renc

e fr

om M

ean

Assays

Component MeansAssay MeansMedianU95

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

21

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

7112017 2PR4 1667 1698 1869 1960 1630 1706 POTP 1655 1685 1850 1781 1596 1576 NEIF 1647 1665 1709 1707 1574 1538 R4Q5 1594 1658 1856 1787 1590 1555 D5 1655 1683 1848 1829 1610 1607 ND6 1629 1647 1828 1826 1589 1548 D9 1655 1652 1861 1858 1569 1572 HBB1 1644 1611 1825 1869 1565 1681 ND14 1645 1707 1871 1773 1564 1492 22C3 1545 1603 1776 1902 1548 1655

7122017 2PR4 1668 1739 2174 2231 1499 1497 POTP 1587 1702 2202 2292 1576 1611 NEIF 1678 1614 2172 2216 1545 1560 R4Q5 1628 1762 NA 2078 1584 1546 D5 1738 1649 2230 2192 1577 1529 ND6 1667 1678 2094 2034 1599 1478 D9 1695 1653 1949 2110 1587 1504 HBB1 1757 1750 1971 2123 1520 1551 ND14 1771 1764 2123 2066 1479 1438 22C3 1762 1759 2107 2046 1512 1494

7122017 2PR4 1638 1667 1810 1806 1452 1424 POTP 1676 1638 1865 1801 1347 1397 NEIF 1616 1624 1806 1786 1395 1350 R4Q5 1559 1587 1878 1768 1442 1459 D5 1608 1652 1824 1790 1448 1602 ND6 1611 1589 1768 1670 1463 1435 D9 1517 1607 1811 1749 1402 1433 HBB1 1557 1582 1763 1741 1474 1481 ND14 1604 1625 1865 1781 1466 1482 22C3 1675 1616 1815 1866 1439 1568

7132017 2PR4 1888 1939 1784 1821 1470 1444 POTP 1868 1815 1735 1813 1411 1427 NEIF 1953 1902 1778 1813 1512 1439 R4Q5 1779 1871 1835 1749 1452 1468 D5 1924 2124 1794 1848 1567 1524 ND6 1861 1787 1700 1711 1480 1424 D9 1824 1844 1799 1779 1442 1459 HBB1 1809 1799 1765 1769 1430 1467 ND14 1864 1857 1754 1725 1413 1411 22C3 1930 1942 1792 1814 1486 1480

7142017 2PR4 1597 1590 1718 1768 1485 1458 POTP 1533 1404 1753 1698 1422 1459 NEIF 1462 1533 1773 1792 1438 1465 R4Q5 1493 1519 1777 1740 1439 1468 D5 1538 1586 1803 1795 1515 1457 ND6 1492 1516 1718 1693 1409 1412 D9 1528 1538 1802 1743 1469 1474 HBB1 1548 1581 1834 1745 1453 1428 ND14 1513 1514 1731 1714 1465 1425 22C3 1562 NA 1842 NA 1363 1393

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

22

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

101217 2PR4 1431 1411 1791 1770 1357 1364 POTP 1407 1395 1746 1820 1447 1361 NEIF 1487 1398 1803 1803 1448 1404 R4Q5 1450 1439 1767 1739 1409 1411 D5 1394 1422 1745 1773 1389 1429 ND6 1396 1393 1824 1729 1415 1417 D9 1414 1390 1752 1744 1431 1380 HBB1 1401 1382 1813 1754 1381 1336 ND14 1412 1460 1648 1671 1403 1387 22C3 1435 1410 1818 1741 1435 1372

Estimation of Copy Number Concentration It was decided that all three sets of results (homogeneity stability and certification) would be used to derive the certified values with the λ of the first five ldquoten-assayrdquo plates adjusted by a data-driven estimate of the volumetric bias This bias was incorporated into the statistical model providing better estimates of the λ means while capturing all of the variance components (homogeneity stability between-assay and between data sets) as random effects using Bayesian MCMC programmed in OpenBUGS28 Table 14 lists the estimated values and their uncertainties See Appendix C for details

Table 14 Measured Copy Number Concentrations

Genomic Copies per Nanoliter of Solution Component Mean a u(Mean) b U95(Mean) c CV95

d A 151 07 14 93 B 175 03 06 34 C 145 05 10 69

a Mean the experimentally determined most probable estimate of the true value b u(Mean) the standard uncertainty of the Mean c Mean plusmnU95(Mean) the interval which is believed with about 95 confidence to contain the true value d CV95 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the counting unit one29 2) the validity of the Poisson endpoint transformation for digital PCR endpoint assays when applied to samples providing le 05 copies per droplet 3) the determination that the copies are dispersed as dsDNA and 4) calibrated mean droplet volume measurements made at NIST during sample dilution and mastermix preparation Conversion of Copy Number to Mass Concentration As described in Reference 18 metrologically valid conversion from copies per nanoliter to nanograms per microliter requires quantitative estimates for the number of nucleotide bases per reference human genome and the average nucleotide mass Using the values estimated in that report and asserting that each target copy corresponds to one HHGE the required transformation is

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

23

[DNA] ngmicroL

= [DNA] HHGE

nLtimes

3301 ngHHGE

The combined standard uncertainty is

119906119906 [DNA] ng

microL = 1199061199062 [DNA] HHGE

nL + 0531100

times[DNA] HHGE

nL 2

Both the measured and the conversion factor uncertainties are associated with ldquolargerdquo degrees of freedom Therefor the approximate 95 confidence uncertainty is estimated

11988011988095 [DNA] ng

microL = 2 times 119906119906

[DNA] ngmicroL

Table 15 lists the transformation of the Table 14 copy number values These values agree well with the conventional 260 nm absorbance values averaging (4 plusmn 5) larger

Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution

ngmicroL Component Value a u(Measurement)b u(Conversion)c U95(Value)d CV95

e A 498 23 03 47 93 B 578 10 03 21 36 C 479 17 03 33 70

a Estimated true value b Standard uncertainty component from the ddPCR measurements c Standard uncertainty estimated from the relative uncertainty of the conversion factor d Value plusmnU95(Value) is believed with about 95 confidence to contain the true value e 100U95(Value)Value the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the measured estimates of copies per nanoliter and 2) determinations made by us from internationally recognized literature resources of the average number of nucleotide bases per HHGE and the mean molecular weights of the bases18 Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)

While not intended to provide certified values mtDNAnDNA was measured as part of 13 experiments over 5 months Measurements were made using dilution-adjusted λ for the mtND1 mtBatz andor mtKav mitochondrial assays referenced against the dilution-adjusted λ of either the NEIF or HBB1 nDNA assays By chance both DNAs in the component C mixture have a binding site mutation that silences the mtND1 probe Figure 11 graphically summarizes the results by assay and component Table 16 lists the data Table 17 summarizes results from a random effects model that accounts for between-assay variability See Appendix C for details

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

24

Due to the limited number of mtDNA assays used in these measurements these results are metrologically traceable only to the assays used

Figure 11 mtDNAnDNA for Components A B and C

Each dot-and-bar symbol represents a mean mtDNAnDNA and its approximate 95 expanded uncertainty The dotted horizontal lines represent approximate 95 confidence intervals on the ratios

Table 16 mtDNAnDNA Ratios

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

20170711a 1 mtND1 1931 1870 2461 2401 1 mtBatz 1845 1764 2555 2608 3201 3299

20170711b 1 mtND1 1810 1906 2229 2200 1 mtBatz 1785 1837 2108 2192 3347 3287

20170712 1 mtND1 1743 1761 2127 2198 1 mtBatz 1829 1847 2455 2506 3217 3384

20170713 1 mtND1 1716 1800 2211 2199 1 mtBatz 1740 1813 2380 2249 3200 3363

20170714 1 mtND1 1923 1835 2105 1943 1 mtBatz 1935 1851 2136 2119 3294 3013

20170814 1 mtND1 1780 1776 1706 2046 2057 2024 1 mtBatz 1804 1751 1738 1998 2043 1952 2927 2776 2762 1 mtKav 1769 1786 1761 2077 2117 2041 3044 2830 2825

20170824 1 mtND1 1947 1928 1922 2 mtND1 1829 1895 1851 3 mtND1 1674 1617

20170825 1 mtBatz 1667 1813 1708 1598 1928 1916 1915 2832 2840 2788 2725 2 mtBatz 1566 1687 1663 1589 1947 1946 1961 1909 2835 2790 2736 2679 3 mtBatz 1643 1668 1604 1573 1953 1952 1941 1881 2803 2773 2757 2752

20170829 1 mtBatz 1719 1716 1670 1666 1865 1863 1828 1828 2599 2508 2551 2609 2 mtBatz 1581 1573 1601 1612 1944 1964 1947 1895 2617 2483 2626 2602 3 mtBatz 1546 1542 1538 1587 1908 1909 1940 1839 2665 2535 2570 2650

20170831 1 mtND1 1617 1599 1688 1783

ND1

ND1

ND1Ka

v

Kav

Kav

Batz

Batz

Batz

A

B

C

160

210

260

mtD

NAn

DNA

Ratio

Assay

Bind

ing s

ite m

utat

ion

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

25

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

2 mtND1 1622 1579 1823 1631 3 mtND1 1575 1528 1920 1887

20170907 1 mtBatz 1584 1544 1890 1886 2501 2522 20170913 1 mtND1 1757 1724 1768 1990 1921 1982

1 mtBatz 1743 1671 1766 1990 1924 1958 2805 2825 2829 1 mtKav 1692 1662 1767 1903 1831 1919 2709 2738 2727

20171016 1 mtND1 1924 1944 1893 2304 2341 2332 1 mtBatz 1786 1752 1743 2048 2078 2067 2797 2642 2792 1 mtKav 1530 1542 1523 2399 2479 2493 2804 2789 2842 2 mtBatz 1580 1557 1592 2715 2606 2687 2 mtKav 2103 2100 2049 2799 2628 2559 3 mtND1 1749 1769 1777 2065 2166 2164 3 mtBatz 1965 1889 1947 2314 2280 2258 2639 2535 2657 3 mtKav 1813 1864 1864 1867 1987 1971 2953 2757 2802

Table 17 mtDNAnDNA Summary Results

mtDNAnDNA Component Value a u(Value)b U95(Value)c CV95

d A 1737 16 38 17 B 2058 22 44 21 C 2791 23 47 17

a Estimated true value b Standard uncertainty c Value plusmnU95(Value) is believed with about 95 confidence to contain the true value d 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) Performance in Commercial qPCR Chemistries

Traditional qPCR assays only quantify relative to a reference standard Thus qPCR assays do not provide absolute [DNA] for the individual materials but they can provide [DNA] of one component relative to another Figure 12 summarizes the BA and CA ratios for the commercial autosomal (Human) targets from the eleven assays evaluated (see Table 8) The Y-chromosome (Male) assays cannot be summarized in this manner as the (all female) component B does not include a Y chromosome Figure 13 summaries the mass [DNA] results for the (all male) component A and (male-female mixture) component C Since this material was prepared as a nominal 13 malefemale mixture the expected malehuman (total) ratio is 14 sjc As displayed in these figures there is intrinsic variability between the SRM 2372a components and between the commercial qPCR chemistries examined This reflects the difference in the targets and amplification chemistries for each of the commercial kits Many of the commercial chemistries on the market to date contain multi-copy targets for both autosomal and Y-chromosomal quantification to increase the sensitivity of the assays

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

26

Figure 12 qPCR Autosomal Quantitation Performance Summary

Open circles denote the BA CA ratios the 17 commercial qPCR assays evaluated each labeled with the code listed in Table 8 The solid black circle represents ratios as estimated by ddPCR The red lines bound the approximate 95 confidence intervals on the ddPCR-estimated ratios

Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary

Open squares denote the ngμL values for SRM 2372a Components A (male) and C (mixture of male and female) for the seven commercial Y-chromosome qPCR assays evaluated each labeled with the code listed in Table 8 The dotted red lines represent malehuman ratios of 13 14 and 15

Appendix D provides a graphical analysis of all the data generated in the performance assessments for each of the targets amplified in the eleven commercial qPCR chemistries and the [DNA] values for all of the assays

dd

IQ_h

IQ_d

IQHY_h

IQHY_d

QFD_h

QFH_h QFHP_h

QFHP_dQFT_h

QFT_d

QPH_h

QPP_hQPP_d

P_h

PQ_h

PQ_d

06

08

10

12

14

07 09 11 13

CA

BA

IQHY_y

QFD_y

QFT_y

QPH_y

QPP_y

P_y

PQ_y

1413

15

10

12

14

16

18

30 40 50 60 70

[DNA

] C n

gμL

[DNA]A nguL

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

27

Qualification of ddPCR as Fit-For-Purpose We have documented the requirements for establishing metrological traceability of ddPCR human genomic assay results expressed as copies per nanoliter of sample in a series of three peer-reviewed papers151617 Through the experimental evidence they are based upon these papers demonstrate that ddPCR can accurately enumerate the number of independently migrating genomic copies in an aqueous extract Our most recent paper establishes the metrological traceability of these ddPCR results when transformed into the nanogram per microliter units used by the forensic and clinical communities18 However our previous studies have not addressed a fundamental requirement for correctly interpreting digital PCR measurements In broad simplification end-point assay systems count the number of independently partitioning entities regardless of the number of copies the entities contain With assays that target just one site per HHGE entities that migrate as ssDNA will yield counts that are twice the value had they migrated as dsDNA30 Should some originally dsDNA be converted over time to ssDNA ddPCR measurements of the number of independently migrating entities would increase although the number of HHGE would be unchanged Measurements of mass [DNA] provided by absorbance spectrophotometry and ddPCR reflect different properties absorbance at 260 nm is sensitive to the spatial configuration of the nucleotides as well as their number while ddPCR counts independently migrating entities The SRM 2372a materials have absorbance spectra shown in Figure 8 consistent with being predominantly dsDNA The SRM 2372 materials now have the absorbance spectra shown in Figure 14 which are consistent with them being entirely ssDNA

Figure 14 Absorbance Spectra of SRM 2372 Components A and B dsDNA is denatured to ssDNA by diluting the source material with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to mass [DNA]

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372-A200720122012-NaOH121417

0

4

6

0

4

6

220 260 300 340

Wavelength nm

2372-B200720122012-NaOH121517

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

28

To ensure the validity of ddPCR-based value assignments it is thus necessary to evaluate the nature of the partitioning entities Should a solution of what is thought to be dsDNA contain an appreciable fraction of ssDNA then the ddPCR results would require adjustment31 To avoid further delay in making SRM 2372a available to the forensic community the following sections detail a series of experiments that have enabled us to conclude with high confidence that the SRM 2372a components have a dsDNA conformation and that for this material ldquoentitiesrdquo can be considered dsDNA ldquocopiesrdquo The remaining stock of the SRM 2372 components also have a dsDNA conformation however the strands in the old materials are much less robustly connected than they are in the new materials Digital Polymerase Chain Reaction Assays

Three nDNA assays were used HBB1 NEIF and ND6 The HBB1 and NEIF assays reproducibly yield very similar results for DNA extracts from many different donors The HBB1 target is at the tip of the short (p) arm of chromosome 11 The NEIF target is in the centromere region of chromosome 2 The ND6 assay was selected for use because it typically yields the lowest results of the ten nDNA assays used to certify the SRM 2372a components (see Figure 10 below) The ND6 target is about halfway down the long (q) arm of chromosome 6 Heat Treatments (Melting of the DNA into Single Strands)

In all experiments the non-heated control and the heat-treated diluted solutions were stored in the same type of PCR reaction tube The starting point for the heat treatment was to determine if there was a difference between snap-cooling on ice and slower cooling at 25 degC Two tubes of the diluted material were placed in a GeneAmp 9700 thermal cycler (Applied Biosystems) and cycled up to 98 degC and held for 15 min After 15 min one tube was removed and placed on ice while the remaining tube was cooled at 1 degCs in the thermal cycler to 25 degC then placed on ice until prepared for ddPCR Additional heat treatments included cycling samples on a GeneAmp 9700 thermal cycler

a) from 98 degC for 5 min then cooling to 25 degC for 2 min During the 25 degC hold a randomly selected tube was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 45 min at 98 degC

b) from 94 degC for 1 min then cooling to 25 degC for 1 min During the 1 min hold at 25 degC a random sample was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 9 min at 94 degC

c) from 90 degC for 1 min then cooling to 25 degC for 1 min for a total of 20 cycles Samples were pulled at 25 degC after 1 min 3 min 5 min 10 min and 20 min and placed on ice

After the above experiments heat treatments were performed using a ProFlex thermal cycler This thermal cycler has six temperature zones that can be le 5 degC apart This enables heating

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

29

tubes at the same time but at different temperatures After treatment tubes were cooled at 5 degCs to 4 degC in the thermal cycler While designed for use with PCR plates a MicroAmp 96-well tray for the VeriFlex block enabled its use with single PCR reaction tubes Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a

As a first step in evaluating whether the DNA in the SRM materials disperse as dsDNA or ssDNA a direct comparison of the five materials was made using the HBB1 and NEIF assays Table 18 reports summary statistics for the ddPCR measurements reported as λ copiesdroplet and the estimated mass [DNA] values from the ddPCR results18 from the conventional dsDNA relationship7 and from the conventional ssDNA relationship after denaturing with 04 molL NaOH89 While the mass [DNA] values differ somewhat between the three methods there is no evidence for the SRM materials segregating exclusively or even mostly as ssDNA The extent of the hypochromic interactions among the nucleotides does not accurately reflect the extent to which the two strands of dsDNA have separated

Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a

λ copies per droplet λ copies per droplet ngmicroL Sample Assay Nrep

a Mean CVb ncmbc Mean CVb ddPCRd dsDNAe ssDNAf

2372-A HBB1 4 03645 19 8 03644 20 566g 524 57 2372-A NEIF 4 03644 23 2372-B HBB1 4 03513 23 8 03487 35 580h 536 61 2372-B NEIF 4 03461 47 2372a-A HBB1 4 02603 27 8 02649 30 474 465 454 2372a-A NEIF 4 02694 24 2372a-B HBB1 4 02908 34 8 02891 37 517 556 534 2372a-B NEIF 4 02873 45 2372a-C HBB1 4 02757 28 8 02711 28 485 502 474 2372a-C NEIF 4 02665 16

a Number of technical replicates for the given assay b CV = 100(standard deviation)mean c Combined number of technical replicates ignoring potential differences between the assays d Estimated from the ddPCR λ measurements listed in this Table18 e Estimated from the conventional relationship between the mass [DNA] of dsDNA and absorbance at

260 nm7 using absorbance spectra acquired shortly after solutions were produced f Estimated from the relationship between the mass [DNA] of ssDNA and absorbance at 260 nm after 11

dilution with 04 molL NaOH89 g Values adjusted to measured (132 plusmn19) volume loss of archived units due to evaporation h Values adjusted to measured (71 plusmn25) volume loss of archived due to evaporation

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

30

Quantitative Evaluation of ssDNA Content

Absorbance spectrophotometry is routinely used to quantitatively assess DNA tertiary structure3233 On storage ldquoat physiological pH the intrastrand repulsion between the negatively charged phosphate groups forces the double helix into more rigid rodlike conformationrdquo34 suggesting a mechanism for the observed slow changes in the SRM 2372 materials As demonstrated in Table 18 absorbance at 260 nm does not necessarily correlate with ddPCR behavior Yet ldquofurthermore the repulsion between phosphate groups on opposite strands tends to separate the complementary strandsrdquo34 Therefore a method that is independent of hypochromism is needed to adequately evaluate the proportion (if any) of ldquotruerdquo (completely disjoint) ssDNA in the SRM materials Circular dichroism spectroscopy is sensitive to the conformational structures of DNA in solution35 and different DNA configurations migrate differentially in agarose gel electrophoresis7 Both techniques have been used to qualitatively discriminate ssDNA from dsDNA in ldquopurerdquo single-sequence materials but neither technique appears to be able to quantify small proportions of ssDNA in a complex multi-chromosome genomic material Wilson and Ellison have proposed a real-time chamber digital PCR (cdPCR) technique that can estimate excess proportions of ssDNA copies from the crossing-threshold distribution31 However their method is insufficiently sensitive for low ssDNA proportions It is well known that heating DNA beyond the melting temperature separates mammalian dsDNA into ssDNA36 albeit we have shown using cdPCR that boiling dsDNA for some minutes also renders some target sites non-amplifiable15 In the following section we document a ddPCR-based approach for estimating the proportion if any of ssDNA in a sample that involves less destructive heating While the method is not yet optimized the results from these preliminary studies demonstrate that the SRM 2372 and 2372a components disperse into the ddPCR droplets as dsDNA ddPCR Measurement of ssDNA Proportion Given two otherwise identical aliquots of a given DNA material one subjected to some treatment and the other untreated the ratio between the average number of copies per droplet of treated material and that of its untreated sibling λtλu quantitatively estimates the change in the number of independently dispersing entities Should the treatment reduce the number of accessible amplifiable targets (AAT) λtλu will be less than 1 Reduction in AAT may result from damage to the target itself andor conformational changes to the fragment containing the target that render the target inaccessible or non-amplifiable16 Assuming no denaturation of dsDNA to ssDNA either because of an ineffective treatment or the material having started as ldquopurerdquo ssDNA if the treatment does not itself damage targets then λtλu will equal 1 Should treatment convert at least some dsDNA to ssDNA without damaging targets the number of AAT will increase and λtλu will be greater than 1 Assuming complete conversion of ldquopurerdquo dsDNA to ldquopurerdquo ssDNA without other damage then λtλu will equal 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

31

Values greater than 1 but less than 2 indicate some conversion of dsDNA to ssDNA but also some combination of (a) incomplete denaturation of dsDNA to ssDNA (b) partial renaturation of ssDNA to dsDNA after treatment but before ddPCR analysis (c) damage to targets during the process or ndash of greatest interest ndash (d) the original sample material was an impure mixture of dsDNA and ssDNA The following discussions graphically summarize sets of λtλu followed by tabular listings of the relevant experimental conditions ddPCR measurements and the numerical λtλu values The information in these tables includes

Plate Index Experiment identifier indicates date performed in YYYYMMDD format Material Sample material SRM 2372-A 2372-B 2372a-A 2372a-B or 2372a-C Assay ddPCR assay HBB1 NEIF or ND6 degC Denaturation temperature in degC Min Length of time the treated material was held at denaturation temperature in

minutes Cool Method used to cool treated material from denaturation temperature Snap

Slow R1 or R5 where ldquoSnaprdquo indicates cooling on ice ldquoSlowrdquo cooling at room temperature to 25 degC ldquoR1rdquo cooling in a thermal cycler to 25 degC at 1 degCs and ldquoR5rdquo cooling in a thermal cycler to 4 degC at 5 degCs

F Dilution factor expressed as (volume Material) per (final volume of diluted solution) 12 14 or 18

n Number of technical replicates Mean Mean of technical replicates CV Coefficient of Variation expressed as a percentage 100 Standard Deviation

Mean

λtλu the copies per droplet result of the treated solution λt divided by the copies per droplet result of its untreated sibling λu

u(λtλu) Standard uncertainty of λtλu 119906119906(120582120582t 120582120582ufrasl ) = 120582120582t 120582120582ufrasl CV(120582120582t)100

2

+ CV(120582120582u)100

2

Proof of Principle Figure 15 summarizes two replicate sets of measurements performed to assess whether the SRM 2372 and 2372a materials respond similarly whether the denatured materials renature if allowed to cool slowly and whether results are (reasonably) repeatable Table 19 details the experimental conditions The λtλu values for the four materials are similar all much greater than 1 but not approaching 2 Most of the ratios are on or close to the ldquoslow cooling is the same as snap coolingrdquo line suggesting that renaturation is not a significant concern While the replicate results for the SRM 2372-A and 2372a-A materials do not overlap as closely as those for SRM 2372-B and 2372a-B similar treatment conditions produce similar results

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

32

Figure 15 Comparison of Snap-Cooling to Slow-Cooling

Symbols mark the means of replicate measurements made using the same dilutions on successive days with bars representing plusmn one standard combined uncertainty Results are labeled by material ldquoArdquo and ldquoBrdquo for SRM 2372 components and ldquoaArdquo ldquoaBrdquo and ldquoaCrdquo for SRM 2372a components The diagonal black line represents equality between the snap- and slow-cooling treatments

Table 19 Comparison of Snap-Cooling to Slow-Cooling

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171218 2372-A HBB1 14 4 03830 29

dd20171218 2372-A HBB1 98 15 Slow 14 2 05644 18 1474 0051 dd20171218 2372-A HBB1 98 15 Snap 14 4 05852 15 1528 0051

dd20171218 2372a-A HBB1 14 4 03029 18

dd20171218 2372a-A HBB1 98 15 Slow 14 4 04430 21 1463 0040 dd20171218 2372a-A HBB1 98 15 Snap 14 4 04454 33 1471 0055

dd20171218 2372-B HBB1 14 4 03594 31

dd20171218 2372-B HBB1 98 15 Slow 14 4 05806 18 1615 0058 dd20171218 2372-B HBB1 98 15 Snap 14 4 05825 13 1620 0055

dd20171218 2372a-B HBB1 14 4 03383 15

dd20171218 2372a-B HBB1 98 15 Slow 14 4 05043 38 1491 0061 dd20171218 2372a-B HBB1 98 15 Snap 14 4 05254 20 1553 0039

dd20171219 2372-A HBB1 14 4 03357 23

dd20171219 2372-A HBB1 98 15 Slow 14 4 05340 16 1591 0046 dd20171219 2372-A HBB1 98 15 Snap 14 4 05340 12 1591 0042

A

aA

B

aB

A

aA

B

aB

14

15

16

17

14 15 16 17

λ tλ

u Sn

ap C

oole

d

λtλu Slow Cooled

Rep 1Rep 2Snap = Slow

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

33

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171219 2372a-A HBB1 14 4 02735 24

dd20171219 2372a-A HBB1 98 15 Slow 14 4 04151 03 1518 0037 dd20171219 2372a-A HBB1 98 15 Snap 14 4 03918 21 1433 0045 dd20171219 2372-B HBB1 14 4 03346 22

dd20171219 2372-B HBB1 98 15 Slow 14 4 05476 17 1637 0045 dd20171219 2372-B HBB1 98 15 Snap 14 4 05480 29 1638 0059 dd20171219 2372a-B HBB1 14 4 03158 19

dd20171219 2372a-B HBB1 98 15 Slow 14 4 04798 16 1519 0038 dd20171219 2372a-B HBB1 98 15 Snap 14 4 04856 17 1538 0039

Time at Denaturation Temperature Figure 16 summarizes λtλu for the SRM 2372-B material as a function of denaturation temperature and the cumulative time the solution is held at that temperature Table 20 details the experimental conditions

Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures

The symbols represent the ratio of the mean result of three technical replicates of the untreated and each of the treated solutions the bars represent combined standard uncertainties The lines denote regression fits to the function λtλu = atimese-bt where t is the cumulative time Values and the standard uncertainties of the coefficients are listed in the legend

12

14

16

18

20

0 5 10 15 20

λ t λ

u

Cummulative Treatment Time min

degC a u (a ) b u (b )90 1804 0039 00198 0000494 1788 0029 00099 0002998 1855 0019 00067 00019

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

34

For all three of the denaturation temperatures studied the relationship between λtλu and cumulative time t can be modeled as

λtλu = ae-bt

where a is the extrapolated value of λtλu at t = 0 and b is the rate of change in AAT per minute As expected the rate of loss due to the treatment increases with increasing treatment temperature and cumulative time at that temperature37 Given the much longer treatment time for the 98 degC series the agreement among the three t = 0 parameter (a) values is remarkable This suggests that the first 1 min of treatment efficiently separates dsDNA into ssDNA The pooled relative standard uncertainty of these regression parameter is 17 somewhat smaller than the 25 pooled relative standard uncertainties of the individual λtλu However the plusmnstandard uncertainty interval of the 1-min λtλu results for the 94 degC and 90 degC both overlap the extrapolated t = 0 value suggesting that reliable results can be obtained using a single treatment of short duration

Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171228 2372-B HBB1 14 3 03650 17 dd20171228 2372-B HBB1 98 5 R1 14 3 06157 12 1687 0035 dd20171228 2372-B HBB1 98 10 R1 14 3 05493 08 1505 0028 dd20171228 2372-B HBB1 98 15 R1 14 3 05019 13 1375 0029 dd20171228 2372-B HBB1 98 20 R1 14 3 04583 21 1256 0034 dd20171228 2372-B HBB1 98 25 R1 14 3 04154 32 1138 0041 dd20171228 2372-B HBB1 98 30 R1 14 3 03784 14 1037 0023 dd20171228 2372-B HBB1 98 35 R1 14 3 03408 10 0934 0019 dd20171228 2372-B HBB1 98 40 R1 14 3 02980 01 0816 0014 dd20171228 2372-B HBB1 98 40 R1 14 3 02815 08 0771 0015

dd20171229 2372-B HBB1 14 3 03708 27

dd20171229 2372-B HBB1 94 1 R1 14 3 06598 26 1779 0056 dd20171229 2372-B HBB1 94 2 R1 14 3 06553 32 1767 0063 dd20171229 2372-B HBB1 94 3 R1 14 3 06301 20 1699 0045 dd20171229 2372-B HBB1 94 4 R1 14 3 06252 14 1686 0037 dd20171229 2372-B HBB1 94 5 R1 14 3 06577 14 1773 0039 dd20171229 2372-B HBB1 94 6 R1 14 3 06156 07 1660 0030 dd20171229 2372-B HBB1 94 7 R1 14 3 06118 10 1650 0033 dd20171229 2372-B HBB1 94 8 R1 14 3 06256 12 1687 0035 dd20171229 2372-B HBB1 94 9 R1 14 3 05999 10 1618 0032

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

35

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180104 2372-B HBB1 14 4 03713 25 dd20180104 2372-B HBB1 90 1 R1 14 3 06916 22 1863 0052 dd20180104 2372-B HBB1 90 3 R1 14 3 06465 06 1741 0032 dd20180104 2372-B HBB1 90 5 R1 14 3 06431 20 1732 0045 dd20180104 2372-B HBB1 90 10 R1 14 3 06014 25 1620 0049 dd20180104 2372-B HBB1 90 15 R1 14 3 06024 10 1623 0032 dd20180104 2372-B HBB1 90 20 R1 14 3 06003 14 1617 0035

Temperature Figure 17 summarizes the effect of denaturation temperature using a single treatment of 05 min on the 2372-B and 2372a-B materials Table 21 details the experimental conditions

Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

The symbols represent the ratio of the mean result of four technical replicates of the untreated and three replicates of the treated solutions the bars represent combined standard uncertainties In all cases the duration of the treatment was 05 min The lines denote regression fits to the linear function λtλu = a +bT where Tis the denaturation temperature

The SRM 2372 and 2372a materials respond very differently to changes in denaturation temperature The older spectrophotometrically ssDNA material suffers somewhat greater loss in λtλu as temperature increases The new spectrophotometrically dsDNA material suffers much greater loss in λtλu as temperature decreases

14

15

16

17

18

19

20

82 86 90 94 98

λ tλ

u

Temperature degC

2372-B

2372a-B

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

36

Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180109 2372-B HBB1 14 4 03577 25 dd20180109 2372-B HBB1 100 05 R5 14 3 06314 16 1765 0051

dd20180104a 2372-B HBB1 14 4 03577 05

dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032 dd20180104a 2372-B HBB1 94 05 R5 14 3 06459 25 1806 0046 dd20180104a 2372-B HBB1 90 05 R5 14 2 06591 08 1842 0017 dd20180104a 2372-B HBB1 88 05 R5 14 3 06589 05 1842 0013 dd20180104a 2372-B HBB1 84 05 R5 14 3 06729 10 1881 0022 dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

We hypothesize that in the older SRM 2372 material the between-strand bond is relatively fragile and can be completely severed at relatively low temperature while avoiding much temperature-dependent AAT loss In contrast the between-strand bond in the SRM 2372a material requires higher temperature to completely disassociate the strands This suggests that there is not a sample-independent ldquooptimumrdquo combination of denaturation temperature and treatment duration However 98 degC for 05 min appears to provide comparable results for these two materials DNA Concentration in the ddPCR Solution Figure 18 summarizes the effect of [DNA] in the ddPCR solution with SRM 2372-B at a denaturation temperature of 98 degC and treatment duration of 05 min Table 22 details the experimental conditions

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

37

Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution

The solid black symbols represent the ratio of the mean result of four technical replicates of the untreated and treated solutions with the λtλu values shown on the left-hand vertical axis the bars represent combined standard uncertainties The solid blue circles represent λu and the open red circles λt with values shown on the right-hand vertical axis at this graphical scale the standard deviations of the technical replicates are covered by the symbol The diagonal blue and red lines represent the log-linear relationships between just the 18 and 14 λ values

The primary effect of [DNA] appears to be the loss of linearity in λt with copiesdroplet values greater than about 1 copiesdroplet There may be some advantage to use somewhat greater dilutions than the routine 14 however the small increase in λtλu may be offset by the increased variability with small λu This suggests the use of dilutions that give λu values of (01 to 04) copiesdroplet

Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180117 2372-B HBB1 12 4 07127 11 dd20180117 2372-B HBB1 98 05 R5 12 4 12331 14 1730 0031

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 3 06151 15 1843 0040

dd20180117 2372-B HBB1 18 4 01549 18 dd20180117 2372-B HBB1 98 05 R5 18 4 02897 20 1870 0051

12 14 18

170

175

180

185

190

02

04

06

08

101214

λ T

arge

t D

ropl

et

λ tλ

u

Dilution Factor

λ₊λᵤλᵤλ₊

λtλu

λu

λt

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

38

Equivalence of Assays Figure 19 summarizes the agreement between the HBB1 NEIF and ND6 assays at a denaturation temperature of 98 degC and treatment duration of 05 min Table 23 details the experimental conditions

Figure 19 Comparison of Assays

The symbols compared the λtλu results from the HBB1 and either the NEIF or ND6 assays Each λtλu value is the ratio of the mean result of four technical replicates of an untreated and three or four replicates of the treated solutions The bars represent combined standard uncertainties The diagonal line represents equality between the assays

There is little or no difference between the λtλu results from the HBB1 NEIF and ND6 assays suggesting that the chromosomal location of the assay target has little influence Since the NEIFHBB1 comparison includes results for both the SRM 2372 and 2372a materials this equivalence is not material-specific

14

16

18

20

14 16 18 20

λ tλ

u NE

IF o

r ND6

λtλu HBB1

NEIFHBB1

ND6HBB1

2nd Assay = HBB1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

39

Table 23 Comparison of Assays

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu)

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045 dd20180105 2372a-B ND6 94 05 R5 14 3 05526 36 1711 0074 dd20180105 2372a-B ND6 90 05 R5 14 3 05326 18 1649 0051 dd20180105 2372a-B ND6 88 05 R5 14 3 05077 22 1572 0052 dd20180105 2372a-B ND6 86 05 R5 14 3 05038 23 1560 0053 dd20180105 2372a-B ND6 84 05 R5 14 3 04906 10 1519 0041

dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

40

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials

Figure 20 summarizes the currently available λtλu results obtained by denaturing at 98 degC for 05 min rapidly cooling in the thermal cycler with dilutions that provide asymp 03 copiesdroplet The figure also displays the consensus result and its 95 confidence region estimated using the Hierarchical Bayes method provided in the NIST Consensus Builder (NICOB) system3839 Table 24 details the experimental conditions

Figure 20 Evaluation of Limiting Ratio

The symbols represent the ratio of the mean result of three or four technical replicates of the untreated and treated solutions the bars represent combined standard uncertainties The thick horizontal line represents a consensus value covering all displayed results the dotted horizontal lines bracket the 95 confidence interval about the consensus value

The Hierarchical Bayes consensus result for these values is 183 with a 95 confidence interval of plusmn003 All but three of the 16 λtλu values are within one standard uncertainty of this consensus value with most of the λtλu means lying within the 95 confidence interval This suggests that there are no statistically significant differences between the SRM 2372 and 2372a materials with regard to ssDNA proportion

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

17

18

19

20

λ tλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

41

Table 24 Evaluation of Limiting Ratio

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180104a 2372-B HBB1 14 4 03577 05 dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032

dd20180108 2372-B HBB1 14 3 03463 09 dd20180108 2372-B HBB1 98 05 R5 14 3 06562 14 1895 0031

dd20180109 2372-B HBB1 14 3 03577 25 dd20180109 2372-B HBB1 98 05 R5 14 3 06495 21 1816 0059

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 4 06151 15 1843 0040

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

42

The agreement among the measured λtλu is more directly visualized using estimated unilateral ldquoDegrees of Equivalencerdquo d Each d is the absolute difference between the measured and consensus values

119889119889 = |120582120582119905119905 120582120582119906119906frasl minus consensus|

and is associated with a 95 confidence interval U95(d) that combines the uncertainty estimates of the measured and consensus values The U95(d) calculation details are specific to the consensus estimation method39 Values of d plusmn U95(d) that include 0 indicate that the λtλu result does not differ from the majority at about a 95 level of confidence Figure 21 displays the d for the λtλu results for the Hierarchical Bayes consensus value39 By this figure of merit all the λtλu can be considered as members of a single population

Figure 21 Degrees of Equivalence

The symbols represent the Degree of Equivalence d for the λtλu results relative to the Hierarchical Bayes consensus value The bars represent 95 level of confidence uncertainty intervals d plusmn U95(d) The horizontal line marks the d = 0 ldquono differencerdquo value Results with d plusmn U95(d) that include this line are not significantly different from the majority

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

-024

-018

-012

-006

000

006

012

018

024

Degr

ees o

f Equ

ival

ence

λtλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

43

Conclusions and Recommendations With high confidence all tubes of the SRM 2372a solution have the same nDNA copy number content within measurement repeatability Additionally within measurement repeatability all tubes of the SRM 2372a have the same mtDNAnDNA With high confidence the nDNA copy number content of the SRM 2372a solution is thermally stable from 4 degC to 37 degC over an extended period The solution should not be shipped or stored below 4 degC With high confidence the ddPCR measurements used to establish the entity and mass concentrations and the mtDNAnDNA ratios are adequately precise and unbiased With high confidence the DNA in the SRM 2372a solutions is now dsDNA and can be expected to remain dsDNA for at least ten years when stored at 4 degC Recommended Certification Values

Since the component materials are believed to be homogeneous are of very similar composition and were prepared and processed similarly the observed differences in the measured entity concentration uncertainties are likely artifacts of the measurement process To limit over-interpretation of these differences we recommend that the copy per nanoliter values of the three components be assigned same 95 relative uncertainty interval (CV95) based on the largest of the measured CV95 values (93 ) rounded up to 10 Since the ngμL transformation does not appreciably inflate this relative uncertainty we recommend that the mass concentration values also be assigned CV95 of 10 Similarly we recommend that the mtDNAnDNA ratios be rounded to integer values all three components be assigned the same CV95 and that this uncertainty be based on the largest (21 ) of the measured CV95 rounded up to 25 Table 25 lists the values that we recommend be used in the Certificate of Analysis

Table 25 Recommended Values and 95 Uncertainties for Certification

Component Units Value U95(Value) a A Copies per nanoliter 151 15 B Copies per nanoliter 175 18 C Copies per nanoliter 145 15 A ngμL 498 50

B ngμL 578 58 C ngμL 479 48 A mtDNAnDNA 174 4

B mtDNAnDNA 206 5 C mtDNAnDNA 279 7

a The interval Value plusmn U95(Value) is believed with about 95 confidence to contain the true value of the measurand

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

44

Use of SRM 2372a for Value-Assigning In-House Reference Materials

The within-component and between-assay variability observed for the commercial qPCR assays is not unexpected since different qPCR methods exploit qualitatively different molecular targets Given the intrinsic variability of the human genome the prospect of developing a human DNA source material that responds equally to all current and future qPCR assays is improbable All three components should be used to elucidate potential bias when using these materials to value assign in-house DNA reference and control materials for qPCR assays

Certificate of Analysis In accordance with ISO Guide 31 2000 a NIST SRM certificate is a document containing the name description and intended purpose of the material the logo of the US Department of Commerce the name of NIST as a certifying body instructions for proper use and storage of the material certified property value(s) with associated uncertainty(ies) method(s) used to obtain property values the period of validity if appropriate and any other technical information deemed necessary for its proper use A Certificate is issued for an SRM certified for one or more specific physical or engineering performance properties and may contain NIST reference information or both values in addition to certified values A Certificate of Analysis is issued for an SRM certified for one or more specific chemical properties Note ISO Guide 31 is updated periodically check with ISO for the latest version [httpswwwnistgovsrmsrm-definitions] For the most current version of the Certificate of Analysis for NIST SRM 2372a Human DNA Quantification Standard please visit httpswww-snistgovsrmorsview_detailcfmsrm=2372a

References

1 National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research Department of Health Education and Welfare (1978) Belmont Report Ethical Principles and Guidelines for the Protection of Human Subjects of Research DHEW Publication No (OS) 78-0012 httpvideocastnihgovpdfohrp_belmont_reportpdf

2 CDCNIH Biosafety in Microbiological and Biomedical Laboratories 5th ed HHS publication No (CDC) 21-1112 Chosewood LC Wilson DE Eds US Government Printing Office Washington DC (2009) httpwwwcdcgovbiosafetypublicationsbmbl5indexhtm

3 Duewer DL Kline MC Redman JW Newall PJ Reeder DJ NIST Mixed Stain Studies 1 and 2 Interlaboratory Comparison of DNA Quantification Practice and Short Tandem Repeat Multiplex Performance with Multiple-Source Samples J Forensic Sci 200146(5)1199-1210 PMID 11569565

4 Kline MC Duewer DL Redman JW Butler JM NIST Mixed Stain Study 3 DNA quantitation accuracy and its influence on short tandem repeat multiplex signal intensity Anal Chem 200375(10)2463-2469 httpsdoiorg101021ac026410i

5 Kline MC Duewer DL Redman JW Butler JM Results from the NIST 2004 DNA Quantitation Study J Forensic Sci 200550(3)571-578 PMID 15932088

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

45

6 Kline MC Duewer DL Travis JC Smith MV Redman JW Vallone PM Decker AE Butler JM Production and Certification of Standard Reference Material 2372 Human DNA Quantitation Standard Anal Bioanal Chem 2009394(4)1183-1192 httpsdoiorg101007s00216-009-2782-0

7 Sambrook J and Russell DW (2001) Molecular Cloning a Laboratory Manual Cold Spring Harbor Laboratory Press Cold Spring Harbor New York

8 EU-JRC Protocol NK603 ndash Community Reference Laboratory Event-specific method for the quantification of maize line NK603 using real-time PCR Section 33 Spectrophotometric Measurement of DNA Concentration httpgmo-crljrceceuropaeusummariesNK603-WEB-ProtocolValidationpdf

9 ISO 215712005(E) Foodstuffs ndash Methods of analysis for the detection of genetically modified organisms and derived products ndash Nucleic acid extraction Annex B Methods for quantitation of extracted DNA pp 34-36 International Organization for Standardization Geneva (2005)

10 Certificate of Analysis Standard Reference Material 2372 Human DNA Quantitation Standard (2013) Gaithersburg MD USA Standard Reference Materials Program National Institute of Standards and Technology httpswww-snistgovsrmorsview_certcfmsrm=2372

11 Vogelstein B Kinzler KW Digital PCR Proc Natl Acad Sci USA 1999969236-9241 12 Hindson BJ Ness KD Masquelier DA Belgrader P Heredia NJ Makarewicz AJ Bright IJ Lucero

MY Hiddessen AL Legler TC Kitano TK Hodel MR Petersen JF Wyatt PW Steenblock ER Shah PH Bousse LJ Troup CB Mellen JC Wittmann DK Erndt NG Cauley TH Koehler RT So AP Dube S Rose KA Montesclaros L Wang S Stumbo DP Hodges SP Romine S Milanovich FP White HE Regan JF Karlin-Neumann GA Hindson CM Saxonov S Colston BW High-Throughput Droplet Digital PCR System for Absolute Quantitation of DNA Copy Number Anal Chem 2011838604-8610 httpsdoiorg101021ac202028g

13 Pinheiro LB Coleman VA Hindson CM Herrmann J Hindson BJ Bhat S Emslie KR Evaluation of a Droplet Digital Polymerase Chain Reaction Format for DNA Copy Number Quantification AnalChem 2012841003-1011 httpsdoiorg101021ac202578x

14 Corbisier P Pinheiro L Mazoua S Kortekaas AM Chung PY Gerganova T Roebben G Emons H Emslie K DNA copy number concentration measured by digital and droplet digital quantitative PCR using certified reference materials Anal Bioanal Chem 2015407(7)1831-1840 httpsdoiorg101007s00216-015-8458-z

15 Duewer DL Kline MC Romsos EL Real-time cdPCR opens a window into events occurring in the first few PCR amplification cycles Anal Bioanal Chem 2015407(30)9061-9069 httpsdoiorg101007s00216-015-9073-8

16 Kline MC Romsos EL Duewer DL Evaluating Digital PCR for the Quantification of Human Genomic DNA Accessible Amplifiable Targets Anal Chem 201688(4)2132-2139 httpsdoiorg101021acsanalchem5b03692

17 Kline MC Duewer DL Evaluating Droplet Digital Polymerase Chain Reaction for the Quantification of Human Genomic DNA Lifting the Traceability Fog Anal Chem 201789(8)4648-4654 httpsdoiorg101021acsanalchem7b00240

18 Duewer DL Kline MC Romsos EL Toman B Evaluating Droplet Digital PCR for the quantification of human genomic DNA Conversion from Copies per Nanoliter to Nanogram Genomic DNA per Microliter Anal Bioanal Chem 2018in press httpsdoiorg101007s00216-018-0982-1

19 Dagata JA Farkas N Kramer JA Method for measuring the volume of nominally 100 μm diameter spherical water in oil emulsion droplets NIST Special Publication 260-184 2016 httpsdoiorg106028NISTSP260-184

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

20 ISO 170342016 General requirements for the competence of reference material producers International Organization for Standardization Geneva (2016)

21 NCBI Standard Nucleotide BLAST httpsblastncbinlmnihgovBlastcgiPAGE_TYPE=BlastSearch

22 Timken MD Swango KL Orrego C Buoncristiani MR A Duplex Real-Time qPCR Assay for the Quantification of Human Nuclear and Mitochondrial DNA in Forensic Samples Implications for Quantifying DNA in Degraded Samples J Forensic Sci 200550(5)1044-60 PMID 16225209

23 Walker JA Hedges DJ Perodeau BP Landry KE Stoilova N Laborde ME Shewale J Sinha SK Batzer MA Multiplex polymerase chain reaction for simultaneous quantitation of human nuclear mitochondrial and male Y-chromosome DNA application in human identification Anal Biochem 200533789ndash97 httpsdoiorg101016jab200409036

24 Kavlick MF Lawrence HS Merritt RT Fisher C Isenberg A Robertson JM Budowle B Quantification of Human Mitochondrial DNA Using Synthesized DNA Standards J Forensic Sci 201156(6) 1457ndash1463 httpsdoiorg101111j1556-4029201101871x

25 Stulnig TM Amberger A Exposing contaminating phenol in nucleic acid preparations BioTechniques 199416(3) 403-404

26 Majumdar N Wessel T Marks J Digital PCR Modeling for Maximal Sensitivity Dynamic Range and Measurement Precision PLoS ONE 201510(3)e0118833 httpsdoi101371journalpone0118833

27 Tellinghuisen J Partition Volume Variability in Digital Polymerase Chain Reaction Methods Polydispersity Causes Bias but Can Improve Precision Anal Chem 2016881283-12187 httpsdoiorg101021acsanalchem6b03139

28 Lunn DJ Spiegelhalter D Thomas A Best N (2009) Statistics in Medicine 283049ndash3082 29 De Biegravevre P Dybkaer R Fajgelj A Hibbert DB Metrological traceability of measurement results in

chemistry Concepts and implementation Pure Appl Chem 201183(10)1873-1935 httpsdoiorg101007s00769-011-0805-y

30 Huggett JF Foy CA Benes V Emslie K Garson JA Haynes R Hellemans J Kubista M Mueller RD Nolan T Pfaffl MW Shipley GL Vandesompele J Wittwer CT Bustin SA The Digital MIQE Guidelines Minimum Information for Publication of Quantitative Digital PCR Experiments Clin Chem 201359(6)892-902 httpsdoiorg101373clinchem2013206375

31 Wilson PJ Ellison SLR Extending digital PCR analysis by modelling quantification cycle data BMC Bioinfo 201617421 httpsdoiorg101186s12859-016-1275-3

32 Wang X Son A Effects of pretreatment on the denaturation and fragmentation of genomic DNA for DNA hybridization Environ Sci Processes Impacts 2013152204 httpsdoiorg101039c3em00457k

33 Nwokeoji AO Kilby PM Portwood DE Dickman MJ Accurate Quantification of Nucleic Acids Using Hypochromicity Measurements in Conjunction with UV Spectrophotometry Anal Chem 201789(24)13567-13574 https101021acsanalchem7b04000

34 Farkas DH Kaul KL Wiedbrauk DL Kiechle FL Specimen collection and storage for diagnostic molecular pathology investigation Arch Pathol Lab Med 1996120(6)591-6

35 Kypr J Kejnovskaacute I Renčiuk D Vorličkovaacute M Circular dichroism and conformational polymorphism of DNA Nucleic Acids Res 200937(6)1713-1725 httpsdoiorg101093nargkp026

36 Strachan T Read AP (1999) Human Molecular Genetics 2nd Ed John Wiley amp Sons NY

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

37 Bhat S McLaughlin JL Emslie KR Effect of sustained elevated temperature prior to amplification on template copy number estimation using digital polymerase chain reaction Analyst 2011136724ndash32 httpsdoiorg101039c0an00484g

38 NIST Consensus Builder (NICOB) httpsconsensusnistgov 39 Koepke A Lafarge T Toman B Possolo A NIST Consensus Builder Userrsquos Manual

httpsconsensusnistgovNISTConsensusBuilder-UserManualpdf

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

Appendix A DNA Extraction Method The following method has been adapted from Miller et al (1988)A1 Procedure

1 Transfer approximately 5 mL of buffy coat into a 50 mL sterile polypropylene centrifuge tube and add 30 mL Red Blood Cell Lysis Buffer

2 Mix by gentle inversion and let stand at room temperature for 15 min

3 Centrifuge at approximately 209 rads (2000 rpm) for 20 min

4 Discard supernatant

5 Add 9 mL of Nucleic Lysis Buffer and re-suspend the pelleted cells by agitating with a plastic transfer pipette

6 Add 300 microL of 20 sodium dodecyl sulfate (SDS) and 18 mL of Protease K Solution (freshly prepared 2 mg to 5 mg Protease K per 1 mL Protease K Buffer)

7 Mix by inversion followed by incubation overnight or longer at 37 ordmC to 40 ordmC in a shaking water bath The length of incubation is determined by the time required to solubilize the cells present in the tube

8 Add 4 mL saturated ammonium acetate solution (in deionized water stored at 2 ordmC to 8 ordmC) Shake vigorously for 15 s

9 Centrifuge at approximately 471 rads (4500 rpm) for 10 min

10 Pipette supernatant into a 50 mL sterile polypropylene centrifuge tube and add two volumes of room temperature absolute ethanol Mix by gentle inversion until the DNA aggregates and floats to the top of the tube

11 Spool the aggregated DNA from the current tube into a fresh tube of 80 ethanol using a sterile inoculation loop At this step multiple tubes of precipitated DNA may be pooled to make a large lot of the purified material Mix by gentle inversion

12 Transfer the aggregated DNA to a fresh dry 50 mL polypropylene centrifuge tube Remove as much ethanol as possible by squeezing the DNA against the wall of the tube with a sterile inoculation loop Air-dry the material in a laminar flow hood

13 Resolubilize DNA in TE-4 buffer and store material at 4 ordmC Reagents

Red Blood Cell Lysis Buffer 144 mmolL ammonium chloride 1 mmolL sodium bicarbonate in deionized (DI) water Store at 2 ordmC to 8 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

Nucleic Lysis Buffer 10 mmolL Tris-HCl 400 mmolL sodium chloride 20 mmolL disodium ethylenediaminetetraacetic acid in DI Water adjust pH to 82 Store at 2 ordmC to 8 ordmC

Protease K Buffer

20 mmolL disodium ethylenediaminetetraacetic acid 10 (mass per volume) Sodium Dodecyl Sulfate (SDS) in DI water Store at 2 ordmC to 8 ordmC

Protease K

Sigma-Aldrich cat P6556 Tritirachium album lyophilized powder Store at ndash20 ordmC

Ammonium Acetate

192 molL ammonium acetate (saturated) in DI water Store at 2 ordmC to 8 ordmC

Absolute Ethanol (200 proof)

Store at room temperature Reference

A1 Miller SA Dykes DD Polesky HF A simple salting out procedure for extracting DNA from human nucleated cells Nucleic Acids Res 198816(3)1215 httpsdoiorg101093nar1631215

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

48

Appendix B Spectrophotometric Measurement Protocol The following is for use with a scanning recording spectrophotometer The procedures include the spectrophotometer calibration measurement of the ldquonativerdquo dsDNA and measurement of the NaOH denatured ssDNA The procedure for evaluating solutions after treatment with NaOH is adapted from references89 Spectrophotometer Calibration

Establish Baseline Set the spectrophotometer to acquire data from 220 nm to 345 nm in steps of 01 nm Zero the instrument with nothing (air) in the sample and reference cells Acquire an air-air baseline of the instrument with nothing (air) in the sample and reference cells If the trace has significant structure or excessive noise take corrective action If the trace is acceptable proceed with data collection Enable Verification of Wavelength Calibration SRM 2034 is a solution of holmium oxide in 10 perchloric acid contained in a sealed quartz cuvette delivering an intrinsic UVVis wavelength standard for absorption spectrophotometryB1 Evaluation of the location of peak maxima in the spectra of this solution enables verification of the spectrophotometers wavelength calibration Place SRM 2034 in the sample cell Create and name a file to hold the verification spectra Perform three replicate SRM 2034-to-air scans saving each scan to the file Remove SRM 2034 from the spectrophotometer and store appropriately Enable Verification of Absorbance Calibration SRM 2031 consists of three metal-on-fused silica filters held in cuvette-sized filter holders plus an empty filter holder The three filters are certified for transmittance and transmittance density (-log10(transmittance) effectively absorbance) at several wavelengths with nominal wavelength-independent transmittances of 10 30 and 90 Evaluation of the absorbance of each of these three filters at the certified 250 nm 280 nm and 340 nm wavelengths enables verification of the spectrophotometers absorbance calibration Place the empty SRM 2031 filter holder in the reference cell Remove the protective covers from the 10 transmittance filter and place in the sample cell Perform three replicate scans saving each scan to file Remove the filter from the spectrophotometer replace the protective covers and store appropriately Repeat for the 30 and 90 filters Close the file containing the verification spectra For ldquonativerdquo dsDNA Samples

Verify Optical Balance of Cuvettes The measurements of the DNA-containing samples are made using a 100 microL quartz cuvette in the sample cell compared to an optically matched cuvette in the reference cell The optical equivalence of these two cuvettes can be verified by filling both cuvettes with appropriate volumes of a sample solution matrix (TE-4 buffer) that does not contain DNA Fill two dry

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

49

clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes followed by a methanol rinse in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquonativerdquo dsDNA remove and clean both the reference cuvette and the sample cuvette For NaOH denatured ssDNA Samples

Prepare a 04 molL NaOH Stock Solution To a 15 mL disposable centrifuge tube add 016 g NaOH pellets add deionized (DI) water to the 10 mL line and mix until all the NaOH pellets have dissolved Centrifuge at 524 rads (5000 rpm) for 5 min For every 9 samples to be analyzed aliquot 1 mL of this 04 molL NaOH stock solution into a 15 mL screw top tube Use these tubes to prepare the NaOH blanks and DNA samples In a 15 mL screw cap vial mix 200 microL TE-4 buffer with 200 microL 04 molL sodium hydroxide solution to achieve a final NaOH concentration of 02 molL Vortex 5 s then centrifuge at 1257 rads (12 000 rpm) for 1 min Fill two dry clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Prepare the Sample Solutions For each unit of material to be measured (Tube or vial)

a) Vortex the unit for 5 s b) Centrifuge at 1257 rads (12 000 rpm) for 1 min c) Remove a 100 microL aliquot and place in a new labeled vial d) Add 100 microL of the freshly prepared 04 molL NaOH solution (with the same pipet) e) Vortex 5 s f) Centrifuge at 1257 rads (12 000 rpm) for 1 min

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

50

Measure Sample Solutions Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes rinse with methanol in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the NaOH denatured DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquodenaturedrdquo ssDNA close the file holding the spectra for this series Remove and clean both the reference cuvette and the sample cuvette After All Sample Measurements are Complete

Enable Verification of Spectrophotometer Stability The stability of the spectrophotometer over the course of the data collection can be evaluated by comparing spectra for SRM 2034 and 2031 taken at the start of data collection with spectra taken at the end of data collection Repeat the wavelength and absorbance verification measurements described above collecting the spectra in a suitably named file Clean Up Transfer all relevant data files from the spectrophotometers storage to a transfer device Clean the spectrophotometer work area as necessary and restore the SRMs and cuvettes to their appropriate storage locations Evaluation

Export the files to a spreadsheet Average the replicate spectra Subtract the average absorbance at A320 from the absorbance at A260 resulting in the corrected absorbance at A260 The conventional spectrophotometric estimate of the mass concentration of the dsDNA in the sample is achieved by multiplying the corrected absorbance by 50 The conventional estimate for the NaOH-treated ssDNA is achieved by multiplying the corrected absorbance by 37times2 where the ldquotimes2rdquo adjusts for the 1rarr2 volumetric dilution of the sample Note reliable spectrophotometric measurements require corrected A260 to be greater than 005 Reference

B1 Travis JC Acosta JC Andor G Bastie J Blattner P Chunnilall CJ Crosson SC Duewer DL Early EA Hengstberger F Kim C-S Liedquist L Manoocheri F Mercader F Metzdorf J Mito A Monardh LAG Nevask S Nilsson M Noeumll M Rodriguez AC Ruiacutez A Schirmacher A Smith MV Valencia G van Tonder N Zwinkels J Intrinsic Wavelength Standard Absorption Bands in Holmium Oxide Solution for UVvisible Molecular Absorption Spectrophotometry J Phys Chem Ref Data 200534(1)41-56

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

51

Appendix C SRM 2372a Statisticianrsquos report Extracted from Blaza Toman

November 30 2017 1 Introduction Candidate material for SRM 2372a is composed of three components labeled A B and C All components are human genomic DNA and prepared at NIST from buffy coat white blood cells from single-source anonymous donors Component A and B are single source (malefemale respectively) and Component C is a 1 male to 3 female mixture of two single-source donors Droplet digital PCR (ddPCR) assays were used to obtain measurements of the concentration of copies per microliter of sample Ten NIST-developed human nuclear DNA and three literature-derived human mitochondrial DNA PCR assays were used in this study

11 OpenBUGS Analysis All parameters in the following sections were estimated using Markov Chain Monte Carlo (MCMC) implemented in the OpenBUGS freewareC1 OpenBUGS is a software application for the Bayesian analysis of complex statistical models

12 OpenBUGS Exemplar Output All following sections list the OpenBUGs model (ldquocoderdquo) initialization values (ldquoinitsrdquo) if they are needed and the data used to evaluate the various data sets discussed in the body of this report homogeneity stability nDNA certification and mtDNAnDNA value assignment The provided information in each of the following sections is sufficient to produce estimates but the estimates may not exactly reproduce the exemplar results provided The MCMC paradigm does not produce ldquoanalyticalrdquo values but approaches them asymptotically as the number o model updates increases All exemplar results presented here are based upon the following bull 10-fold ldquothinningrdquo to minimize autocorrelation there were 10 MCMC estimates for every

model update used in parameter estimation bull a ldquoburn inrdquo of 5 000 updates before beginning parameter estimation and bull 10 000 model updates

The following statistics are provided in the tables listing the exemplar results

mean arithmetic mean of the parameterrsquos model update values sd standard deviation of the parameterrsquos model update values MC_error Monte Carlo standard deviation of the mean meanradic(number of model updates) val25pc 25th percentile of the model update values median 50th percentile of the model update values val975pc 975th percentile of the model update values startnumber of ldquoburn inrdquo updates sample number of model update values

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

52

2 Homogeneity assessment 21 Data

Samples of the material of all three components were drawn from 22 different boxes Two assays (NEIF ND6) were used to obtain 3 replicates of the measurement in copies per nanoliter for each sample and component For the complete homogeneity data see

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

22 Model The data were analyzed using a three-level Gaussian hierarchical ANOVA model

119910119910119894119894119894119894119894119894~119873119873120572120572119894119894119894119894119894119894 120590120590119908119908119894119894119905119905ℎ119894119894119894119894 119887119887119887119887119887119887 1198941198941198941198942

where ~N() denotes ldquodistributed as a normal distribution of some mean μ and variancerdquo 119894119894 denotes component (1 for A 2 for B 3 for C) 119895119895 denotes assay (1 for NEIF 2 for ND6) 119896119896 denotes box (1hellip22) and 120572120572119894119894119894119894119894119894~119873119873micro119894119894119894119894120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119887119887119887119887119887119887119894119894

2

All variance components were given a Half Cauchy prior as described in Reference C2 Note The Half Cauchy prior is specified by ldquo~dnorm(0 00016)I(0001)rdquo The ldquoIrdquo notation is for censoring It cuts off any random draws from the Normal distribution that are lower than 0001 An informative prior is needed when there are less than four data per distribution and the half Cauchy is a good general-purpose prior

23 OpenBUGS code and data Homogeneity code ModelBegin for(i in 13)xiNsb[i] ~ dnorm(0 00016)I(0001) chSqNsb[i] ~ dgamma(0505) sigb[i] lt- xiNsb[i]sqrt(chSqNsb[i]) for(i in 13)for(j in 12)mu[ij]~dnorm(010E-5) for(i in 13) muA[i]lt-mean(mu[i])sigbP[i]lt-100(muA[i]sqrt(sigb[i])) for(i in 13)for(j in 12)xiNs[ij] ~ dnorm(0 00016)I(0001) chSqNs[ij]~dgamma(0505) sigt[ij]lt-xiNs[ij]sqrt(chSqNs[ij]) sigtP[ij]lt-100(mu[ij]sqrt(sigt[ij])) for(i in 13)for(j in 12)for(k in 122)alpha[ijk]~dnorm(mu[ij]sigb[i]) for(idx in 1393)lamb[idx]~dnorm(alpha[cmp[idx]asy[idx]box[idx]]sigt[cmp[idx]asy[idx]]) ModelEnd Inits None required

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

53

Homogeneity Data lamb[] asy[] box[] cmp[] rep[] 1632 1 1 1 1 1529 1 1 1 2 1511 1 1 1 3 1616 2 1 1 1 1640 2 1 1 2 1582 2 1 1 3 1531 1 2 1 1 1525 1 2 1 2 1549 1 2 1 3 1615 2 2 1 1 1589 2 2 1 2 1601 2 2 1 3 1524 1 3 1 1 1540 1 3 1 2 1533 1 3 1 3 1582 2 3 1 1 1152 2 3 1 2 1555 2 3 1 3 1454 1 4 1 1 1458 1 4 1 2 1427 1 4 1 3 1627 2 4 1 1 1605 2 4 1 2 1607 2 4 1 3 1465 1 5 1 1 1540 1 5 1 2 1504 1 5 1 3 1607 2 5 1 1 1633 2 5 1 2 1586 2 5 1 3 1512 1 6 1 1 1491 1 6 1 2 1491 1 6 1 3 1511 2 6 1 1 1483 2 6 1 2 1524 2 6 1 3 1418 1 7 1 1 1480 1 7 1 2 1455 1 7 1 3 1479 2 7 1 1 1557 2 7 1 2 1656 2 7 1 3 1496 1 8 1 1 1560 1 8 1 2 1501 1 8 1 3 1514 2 8 1 1 1615 2 8 1 2 1540 2 8 1 3 1588 1 9 1 1 1581 1 9 1 2 1637 1 9 1 3 1474 2 9 1 1 1438 2 9 1 2 1454 2 9 1 3 1574 1 10 1 1 1513 1 10 1 2 1570 1 10 1 3 1594 2 10 1 1 1713 2 10 1 2 1669 2 10 1 3

1557 1 11 1 1 1508 1 11 1 2 1633 1 11 1 3 1571 2 11 1 1 1611 2 11 1 2 1631 2 11 1 3 1500 1 12 1 1 1476 1 12 1 2 1515 1 12 1 3 1626 2 12 1 1 1586 2 12 1 2 1608 2 12 1 3 1672 1 13 1 1 1634 1 13 1 2 1804 1 13 1 3 1490 2 13 1 1 1516 2 13 1 2 1497 2 13 1 3 1551 1 14 1 1 1568 1 14 1 2 1538 1 14 1 3 1617 2 14 1 1 1614 2 14 1 2 1639 2 14 1 3 1456 1 15 1 1 1448 1 15 1 2 1523 1 15 1 3 1525 2 15 1 1 1402 2 15 1 2 1436 2 15 1 3 1573 1 16 1 1 1520 1 16 1 2 1617 1 16 1 3 1634 2 16 1 1 1627 2 16 1 2 1575 2 16 1 3 1516 1 17 1 1 1516 1 17 1 2 1628 1 17 1 3 1613 2 17 1 1 1567 2 17 1 2 1567 2 17 1 3 1612 1 18 1 1 1535 1 18 1 2 1672 1 18 1 3 1529 2 18 1 1 1495 2 18 1 2 1528 2 18 1 3 1598 1 19 1 1 1514 1 19 1 2 1596 1 19 1 3 1571 2 19 1 1 1535 2 19 1 2 1511 2 19 1 3 1575 1 20 1 1 1568 1 20 1 2 1586 1 20 1 3 1652 2 20 1 1 1545 2 20 1 2 1627 2 20 1 3 1658 1 21 1 1

1611 1 21 1 2 1578 1 21 1 3 1647 2 21 1 1 1607 2 21 1 2 1612 2 21 1 3 1547 1 22 1 1 1528 1 22 1 2 1532 1 22 1 3 1537 2 22 1 1 1513 2 22 1 2 1536 2 22 1 3 1790 1 1 2 1 1909 1 1 2 2 1885 1 1 2 3 1870 2 1 2 2 1860 2 1 2 3 1712 1 2 2 1 1719 1 2 2 2 1704 1 2 2 3 1775 2 2 2 1 1718 2 2 2 2 1756 2 2 2 3 1693 1 3 2 1 1755 1 3 2 2 1749 1 3 2 3 1824 2 3 2 1 1483 2 3 2 2 1896 2 3 2 3 1754 1 4 2 1 1745 1 4 2 2 1739 1 4 2 3 1670 2 4 2 1 1632 2 4 2 2 1814 2 4 2 3 1757 1 5 2 1 1687 1 5 2 2 1735 1 5 2 3 1683 2 5 2 1 1759 2 5 2 2 1703 2 5 2 3 1697 1 6 2 1 1709 1 6 2 2 1714 1 6 2 3 1693 2 6 2 1 1723 2 6 2 2 1722 2 6 2 3 1722 1 7 2 1 1625 1 7 2 2 1703 1 7 2 3 1690 2 7 2 1 1684 2 7 2 2 1688 2 7 2 3 1884 1 8 2 1 1802 1 8 2 2 1772 1 8 2 3 1864 2 8 2 1 1887 2 8 2 2 1821 2 8 2 3 1713 1 9 2 1 1734 1 9 2 2 1727 1 9 2 3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

54

1709 2 9 2 1 1756 2 9 2 2 1758 2 9 2 3 1759 1 10 2 1 1809 1 10 2 2 1738 1 10 2 3 1764 2 10 2 1 1816 2 10 2 2 1787 2 10 2 3 1749 1 11 2 1 1782 1 11 2 2 1740 1 11 2 3 1741 2 11 2 1 1792 2 11 2 2 1784 2 11 2 3 1748 1 12 2 1 1712 1 12 2 2 1718 1 12 2 3 1797 2 12 2 1 1614 2 12 2 2 1790 2 12 2 3 1632 1 13 2 1 1668 1 13 2 2 1701 1 13 2 3 1626 2 13 2 1 1566 2 13 2 2 1681 2 13 2 3 1752 1 14 2 1 1535 1 14 2 2 1673 1 14 2 3 1724 2 14 2 1 1741 2 14 2 2 1685 2 14 2 3 1871 1 15 2 1 1835 1 15 2 2 1887 1 15 2 3 1770 2 15 2 1 1752 2 15 2 2 1798 2 15 2 3 1688 1 16 2 1 1725 1 16 2 2 1743 1 16 2 3 1744 2 16 2 1 1697 2 16 2 2 1710 2 16 2 3 1789 1 17 2 1 1769 1 17 2 2 1825 1 17 2 3 1716 2 17 2 1 1762 2 17 2 2 1638 2 17 2 3 1826 1 18 2 1 1780 1 18 2 2 1849 1 18 2 3 1805 2 18 2 1 1801 2 18 2 2 1931 2 18 2 3 1751 1 19 2 1 1759 1 19 2 2 1880 1 19 2 3 1827 2 19 2 1 1766 2 19 2 2

1786 2 19 2 3 1738 1 20 2 1 1678 1 20 2 2 1768 1 20 2 3 1752 2 20 2 1 1747 2 20 2 2 1768 2 20 2 3 1742 1 21 2 1 1836 1 21 2 2 1713 1 21 2 3 1746 2 21 2 1 1744 2 21 2 2 1764 2 21 2 3 1761 1 22 2 1 1743 1 22 2 2 1660 1 22 2 3 1666 2 22 2 1 1692 2 22 2 2 1819 2 22 2 3 1493 1 1 3 1 1385 1 1 3 2 1510 1 1 3 3 1565 2 1 3 1 1562 2 1 3 2 1466 2 1 3 3 1516 1 2 3 1 1513 1 2 3 2 1577 1 2 3 3 1490 2 2 3 1 1547 2 2 3 2 1570 2 2 3 3 1451 1 3 3 1 1528 1 3 3 2 1506 1 3 3 3 1485 2 3 3 1 1505 2 3 3 2 1471 2 3 3 3 1440 1 4 3 1 1486 1 4 3 2 1520 1 4 3 3 1529 2 4 3 1 1508 2 4 3 2 1506 2 4 3 3 1523 1 5 3 1 1472 1 5 3 2 1483 1 5 3 3 1513 2 5 3 1 1520 2 5 3 2 1491 2 5 3 3 1437 1 6 3 1 1493 1 6 3 2 1522 1 6 3 3 1492 2 6 3 1 1481 2 6 3 2 1461 2 6 3 3 1485 1 7 3 1 1477 1 7 3 2 1463 1 7 3 3 1475 2 7 3 1 1547 2 7 3 2 1485 2 7 3 3 1433 1 8 3 1

1405 1 8 3 2 1478 1 8 3 3 1497 2 8 3 1 1501 2 8 3 2 1495 2 8 3 3 1463 1 9 3 1 1499 1 9 3 2 1505 1 9 3 3 1503 2 9 3 1 1491 2 9 3 2 1516 2 9 3 3 1534 1 10 3 1 1458 1 10 3 2 1454 1 10 3 3 1448 2 10 3 1 1465 2 10 3 2 1474 2 10 3 3 1533 1 11 3 1 1514 1 11 3 2 1555 1 11 3 3 1488 2 11 3 1 1470 2 11 3 2 1483 2 11 3 3 1450 1 12 3 1 1525 1 12 3 2 1497 1 12 3 3 1496 2 12 3 1 1483 2 12 3 2 1519 2 12 3 3 1537 1 13 3 1 1527 1 13 3 2 1515 1 13 3 3 1499 2 13 3 1 1479 2 13 3 2 1475 2 13 3 3 1515 1 14 3 1 1490 1 14 3 2 1494 1 14 3 3 1469 2 14 3 1 1480 2 14 3 2 1466 2 14 3 3 1511 1 15 3 1 1568 1 15 3 2 1594 1 15 3 3 1515 2 15 3 1 1543 2 15 3 2 1565 2 15 3 3 1449 1 16 3 1 1452 1 16 3 2 1471 1 16 3 3 1543 2 16 3 1 1448 2 16 3 2 1491 2 16 3 3 1388 1 17 3 1 1421 1 17 3 2 1394 1 17 3 3 1452 2 17 3 1 1435 2 17 3 3 1475 1 18 3 1 1455 1 18 3 2 1445 1 18 3 3 1458 2 18 3 1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

55

1531 2 18 3 2 1459 2 18 3 3 1468 1 19 3 1 1525 1 19 3 2 1446 1 19 3 3 1435 2 19 3 1 1478 2 19 3 2 1421 2 19 3 3 1549 1 20 3 1

1531 1 20 3 2 1499 1 20 3 3 1499 2 20 3 1 1465 2 20 3 2 1468 2 20 3 3 1530 1 21 3 1 1440 1 21 3 2 1484 1 21 3 3 1499 2 21 3 1

1454 2 21 3 2 1442 1 22 3 1 1487 1 22 3 2 1436 1 22 3 3 1589 2 22 3 1 1591 2 22 3 2 1556 2 22 3 3 END

Table C-1 Example OpenBUGS Homogeneity Summary

mean sd MC_error val25pc median val975pc start sample mu[11] 15460 0122 000123 1522 1546 1570 5001 10000 mu[12] 15640 0137 000141 1537 1564 1591 5001 10000 mu[21] 17490 0113 000118 1727 1749 1772 5001 10000 mu[22] 17480 0129 000125 1722 1748 1774 5001 10000 mu[31] 14870 0073 000073 1473 1487 1501 5001 10000 mu[32] 14960 0069 000063 1482 1496 1509 5001 10000 muA[1] 15550 0092 000092 1536 1555 1573 5001 10000 muA[2] 17490 0085 000086 1732 1749 1765 5001 10000 muA[3] 14910 0050 000046 1481 1491 1501 5001 10000 sigbP[1] 3269 04985 000514 2376 3244 4324 5001 10000 sigbP[2] 2548 04385 000461 1733 2533 3472 5001 10000 sigbP[3] 1850 0292 000260 1329 1832 2477 5001 10000

sigtP[11] 2774 0298 000260 2268 2750 3442 5001 10000 sigtP[12] 4036 0429 000438 3290 4005 4986 5001 10000 sigtP[21] 2664 0301 000321 2155 2640 3313 5001 10000 sigtP[22] 3807 0395 000377 3116 3772 4662 5001 10000 sigtP[31] 2313 0250 000209 1880 2291 2873 5001 10000 sigtP[32] 1812 0195 000170 1473 1797 2243 5001 10000

24 Results The material appears to be homogeneous as can be seen in Figure C-1 Based on these results the between-tube variability is of the same order as within-tube variability

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

56

Component A Assay NEIF Component A Assay ND6 box plot gamma[11]

[111][112][113]

[114]

[115][116]

[117]

[118]

[119]

[1110][1111]

[1112]

[1113]

[1114]

[1115]

[1116][1117]

[1118]

[1119][1120]

[1121]

[1122]

114

box plot gamma[12]

[121][122]

[123]

[124][125]

[126]

[127][128]

[129]

[1210]

[1211][1212]

[1213]

[1214]

[1215]

[1216]

[1217]

[1218][1219]

[1220][1221]

[1222]

115

Component B Assay NEIF Component B Assay ND6

box plot gamma[21]

[211]

[212]

[213][214]

[215][216]

[217]

[218]

[219]

[2110][2111]

[2112]

[2113][2114]

[2115]

[2116]

[2117]

[2118]

[2119]

[2120]

[2121]

[2122]

129

box plot gamma[22]

[221]

[222][223]

[224][225][226][227]

[228]

[229]

[2210][2211]

[2212]

[2213]

[2214]

[2215]

[2216][2217]

[2218]

[2219]

[2220][2221][2222]

128

Component C Assay NEIF Component C Assay ND6

box plot gamma[31]

[311]

[312]

[313][314]

[315][316]

[317]

[318]

[319][3110]

[3111]

[3112]

[3113]

[3114]

[3115]

[3116]

[3117]

[3118]

[3119]

[3120]

[3121]

[3122]109

box plot gamma[32]

[321][322]

[323]

[324][325]

[326]

[327][328][329]

[3210]

[3211]

[3212]

[3213][3214]

[3215]

[3216]

[3217]

[3218]

[3219]

[3220][3221]

[3222]

11

Figure C-1 Homogeneity Assessment

Note These results were obtained using the 1-4 gravimetric dilution-adjusted λ but uncorrected for droplet volume

or the 1-10 volumetric dilution That is the vertical axis in each panel should be multiplied by 07349times10 = 7349 This constant difference in scale does not impact the homogeneity analysis Error bars span approximate 95 confidence intervals

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

57

3 Certification measurements 31 Data

Originally the experiment consisted of 5 plates containing the 3 components (A B C) and 12 assays Ten of the assays were for genomic DNA identification (for the certified value) and two of the assays were for mitochondrial detection which is used for informational values (and are not included in this data set) Each of the plates contained samples from a new vial from different boxes (noted in the ldquovial usedrdquo portion the number references the box) The assay positions were alternated across the five plates as well For the plate design and which tubes were used see

Figure 6 Basic Plate Layout Design for Certification Measurements It was later determined that the homogeneity and stability data sets were obtained using a different set of pipettes than the certification data set and since all data was to be combined for certification another plate was obtained using the original pipette For the complete certification data see

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component Stability data was measured at different temperatures over several weeks using assays NEIF and ND6 Only the 4degC data was also used for certification For the complete stability data see

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

32 Model It was decided that all three sets of data should be used to derive the certified values In this way the uncertainty of the certified values would better reflect any potential heterogeneity in the samples as well as any additional unknown factors that may have effected the results Because two different pipettes were used to obtain different sets of data there was a significant bias in some of the measurements Specifically the ldquocertificationrdquo data from the first five plates was obtained using pipette set ldquo1rdquo and the remaining measurements were obtained using pipette set ldquo2rdquo This bias was introduced into the statistical model as part of the mean Possible effects of different assays and type of data (heterogeneity certification and stability) were accounted for in the statistical model in terms of a random effect Specifically the response for each component (A B C) was represented as

119910119910119894119894119894119894119894119894~1198731198731205741205741198941198941198941198941198941198941205901205901198941198942 where ~N() denotes ldquodistributed as a normal distribution of some mean and variancerdquo i = 123 denotes type of measurement heterogeneity certification and stability j = 1hellip10 denotes assay 2PR4 POTP NEIF R4Q5 D5 ND6 D9 HBB1 ND14 22C3 k = 1 2 denotes pipette set 120574120574119894119894119894119894119894119894 = 120572120572119894119894119894119894 + 120573120573119894119894 120572120572119894119894119894119894~119873119873(micro1198941198941205901205901205721205722) 120573120573119894119894 = 120573120573 for data type 1 and 120573120573119894119894 = 0 for data types 2 and 3 For a given 2372a component the variance 1205901205901205721205722 measures the between assay variability and 1205901205901198941198942 measures the repeatability variability for each data type The data was analyzed using Bayesian MCMC programmed in OpenBUGSC2 with non-informative Gaussian priors for the means and Gamma priors for the variance components

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

58

33 OpenBUGS code inits and data Certification code ModelBegin for(i in 13)mu[i]~dnorm(010E-5) sig[i]~dgamma(10E-510E-5) sigalph~dgamma(10E-510E-5) for(i in 110)alpha[1i]~dnorm(mu[1] sigalph) for(i in 12)alpha[2i]~dnorm(mu[2] sigalph) alpha[3i]~dnorm(mu[3]sigalph) b~dnorm(010E-5) beta[1]lt-b beta[2]lt-00 for(i in 1Ntot)mean[i]lt-alpha[typ[i]asy[i]]+beta[pip[i]] for(i in 1Ntot)lamb[i]~dnorm(mean[i]sig[typ[i]]) ModelEnd Certification Inits list(sig=c(111) sigalph=1) Certification Data for Component A list(Ntot=275) lamb[] asy[] cmp[] typ[] pip[] 1667 1 1 1 1 1655 2 1 1 1 1647 3 1 1 1 1594 4 1 1 1 1655 5 1 1 1 1629 6 1 1 1 1655 7 1 1 1 1644 8 1 1 1 1645 9 1 1 1 1545 10 1 1 1 1698 1 1 1 1 1685 2 1 1 1 1665 3 1 1 1 1658 4 1 1 1 1683 5 1 1 1 1647 6 1 1 1 1652 7 1 1 1 1611 8 1 1 1 1707 9 1 1 1 1603 10 1 1 1 1668 1 1 1 1 1587 2 1 1 1 1678 3 1 1 1 1628 4 1 1 1 1738 5 1 1 1 1667 6 1 1 1 1695 7 1 1 1 1757 8 1 1 1 1771 9 1 1 1 1762 10 1 1 1 1739 1 1 1 1 1702 2 1 1 1 1614 3 1 1 1 1762 4 1 1 1 1649 5 1 1 1 1678 6 1 1 1 1653 7 1 1 1

175 8 1 1 1 1764 9 1 1 1 1759 10 1 1 1 1638 1 1 1 1 1676 2 1 1 1 1616 3 1 1 1 1559 4 1 1 1 1608 5 1 1 1 1611 6 1 1 1 1517 7 1 1 1 1557 8 1 1 1 1604 9 1 1 1 1675 10 1 1 1 1667 1 1 1 1 1638 2 1 1 1 1624 3 1 1 1 1587 4 1 1 1 1652 5 1 1 1 1589 6 1 1 1 1607 7 1 1 1 1582 8 1 1 1 1625 9 1 1 1 1616 10 1 1 1 1888 1 1 1 1 1868 2 1 1 1 1953 3 1 1 1 1779 4 1 1 1 1924 5 1 1 1 1861 6 1 1 1 1824 7 1 1 1 1809 8 1 1 1 1864 9 1 1 1 193 10 1 1 1 1939 1 1 1 1 1815 2 1 1 1 1902 3 1 1 1 1871 4 1 1 1 2124 5 1 1 1

1787 6 1 1 1 1844 7 1 1 1 1799 8 1 1 1 1857 9 1 1 1 1942 10 1 1 1 1597 1 1 1 1 1533 2 1 1 1 1462 3 1 1 1 1493 4 1 1 1 1538 5 1 1 1 1492 6 1 1 1 1528 7 1 1 1 1548 8 1 1 1 1513 9 1 1 1 1562 10 1 1 1 159 1 1 1 1 1404 2 1 1 1 1533 3 1 1 1 1519 4 1 1 1 1586 5 1 1 1 1516 6 1 1 1 1538 7 1 1 1 1581 8 1 1 1 1514 9 1 1 1 1431 1 1 1 2 1407 2 1 1 2 1487 3 1 1 2 145 4 1 1 2 1394 5 1 1 2 1396 6 1 1 2 1414 7 1 1 2 1401 8 1 1 2 1412 9 1 1 2 1435 10 1 1 2 1411 1 1 1 2 1395 2 1 1 2 1398 3 1 1 2 1439 4 1 1 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

59

1422 5 1 1 2 1393 6 1 1 2 139 7 1 1 2 1382 8 1 1 2 146 9 1 1 2 141 10 1 1 2 1632 1 1 2 2 1531 1 1 2 2 1524 1 1 2 2 1454 1 1 2 2 1465 1 1 2 2 1512 1 1 2 2 1418 1 1 2 2 1496 1 1 2 2 1588 1 1 2 2 1574 1 1 2 2 1557 1 1 2 2 15 1 1 2 2 1672 1 1 2 2 1551 1 1 2 2 1573 1 1 2 2 1516 1 1 2 2 1612 1 1 2 2 1598 1 1 2 2 1575 1 1 2 2 1658 1 1 2 2 1547 1 1 2 2 1529 1 1 2 2 1525 1 1 2 2 154 1 1 2 2 1458 1 1 2 2 154 1 1 2 2 1491 1 1 2 2 148 1 1 2 2 156 1 1 2 2 1581 1 1 2 2 1513 1 1 2 2 1508 1 1 2 2 1476 1 1 2 2 1634 1 1 2 2 1568 1 1 2 2 152 1 1 2 2 1516 1 1 2 2 1535 1 1 2 2 1514 1 1 2 2 1568 1 1 2 2 1611 1 1 2 2 1528 1 1 2 2 1511 1 1 2 2 1549 1 1 2 2 1533 1 1 2 2 1427 1 1 2 2 1504 1 1 2 2 1491 1 1 2 2 1455 1 1 2 2

1501 1 1 2 2 1637 1 1 2 2 157 1 1 2 2 1633 1 1 2 2 1515 1 1 2 2 1804 1 1 2 2 1538 1 1 2 2 1617 1 1 2 2 1628 1 1 2 2 1672 1 1 2 2 1596 1 1 2 2 1586 1 1 2 2 1578 1 1 2 2 1532 1 1 2 2 1616 2 1 2 2 1615 2 1 2 2 1582 2 1 2 2 1627 2 1 2 2 1607 2 1 2 2 1511 2 1 2 2 1479 2 1 2 2 1514 2 1 2 2 1474 2 1 2 2 1594 2 1 2 2 1571 2 1 2 2 1626 2 1 2 2 149 2 1 2 2 1617 2 1 2 2 1634 2 1 2 2 1613 2 1 2 2 1529 2 1 2 2 1571 2 1 2 2 1652 2 1 2 2 1647 2 1 2 2 1537 2 1 2 2 164 2 1 2 2 1589 2 1 2 2 1152 2 1 2 2 1605 2 1 2 2 1633 2 1 2 2 1483 2 1 2 2 1557 2 1 2 2 1615 2 1 2 2 1438 2 1 2 2 1713 2 1 2 2 1611 2 1 2 2 1586 2 1 2 2 1516 2 1 2 2 1614 2 1 2 2 1627 2 1 2 2 1567 2 1 2 2 1495 2 1 2 2 1535 2 1 2 2 1545 2 1 2 2 1607 2 1 2 2

1513 2 1 2 2 1582 2 1 2 2 1601 2 1 2 2 1555 2 1 2 2 1607 2 1 2 2 1586 2 1 2 2 1524 2 1 2 2 1656 2 1 2 2 154 2 1 2 2 1454 2 1 2 2 1669 2 1 2 2 1631 2 1 2 2 1608 2 1 2 2 1497 2 1 2 2 1639 2 1 2 2 1575 2 1 2 2 1567 2 1 2 2 1528 2 1 2 2 1511 2 1 2 2 1627 2 1 2 2 1612 2 1 2 2 1536 2 1 2 2 1568 1 1 3 2 152 1 1 3 2 1486 1 1 3 2 1441 1 1 3 2 1706 1 1 3 2 144 1 1 3 2 15 1 1 3 2 1537 1 1 3 2 1387 1 1 3 2 1463 1 1 3 2 157 1 1 3 2 1429 1 1 3 2 1667 1 1 3 2 1632 1 1 3 2 1589 1 1 3 2 1474 2 1 3 2 1543 2 1 3 2 1555 2 1 3 2 1599 2 1 3 2 1494 2 1 3 2 1624 2 1 3 2 1577 2 1 3 2 1593 2 1 3 2 1479 2 1 3 2 1517 2 1 3 2 1538 2 1 3 2 159 2 1 3 2 1479 2 1 3 2 1479 2 1 3 2 1522 2 1 3 2 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

60

Table C-2 Example OpenBUGS Certification Summary for Component A

mean sd MC_error val25pc median val975pc start sample b 2584 02948 0005376 2001 2588 3158 100001 100000 mu[1] 1417 02706 0005366 1364 1416 147 100001 100000 mu[2] 1559 008368 4776E-4 1543 1559 1576 100001 100000 mu[3] 1533 01504 0001419 1504 1533 1563 100001 100000

Certification Data for Component B list(Ntot=278) lamb[] asy[] cmp[] typ[] pip[] 1869 1 2 1 1 185 2 2 1 1 1709 3 2 1 1 1856 4 2 1 1 1848 5 2 1 1 1828 6 2 1 1 1861 7 2 1 1 1825 8 2 1 1 1871 9 2 1 1 1776 10 2 1 1 196 1 2 1 1 1781 2 2 1 1 1707 3 2 1 1 1787 4 2 1 1 1829 5 2 1 1 1826 6 2 1 1 1858 7 2 1 1 1869 8 2 1 1 1773 9 2 1 1 1902 10 2 1 1 2174 1 2 1 1 2202 2 2 1 1 2172 3 2 1 1 223 5 2 1 1 2094 6 2 1 1 1949 7 2 1 1 1971 8 2 1 1 2123 9 2 1 1 2107 10 2 1 1 2231 1 2 1 1 2292 2 2 1 1 2216 3 2 1 1 2078 4 2 1 1 2192 5 2 1 1 2034 6 2 1 1 211 7 2 1 1 2123 8 2 1 1 2066 9 2 1 1 2046 10 2 1 1 181 1 2 1 1 1865 2 2 1 1 1806 3 2 1 1 1878 4 2 1 1 1824 5 2 1 1 1768 6 2 1 1 1811 7 2 1 1 1763 8 2 1 1 1865 9 2 1 1 1815 10 2 1 1

1806 1 2 1 1 1801 2 2 1 1 1786 3 2 1 1 1768 4 2 1 1 179 5 2 1 1 167 6 2 1 1 1749 7 2 1 1 1741 8 2 1 1 1781 9 2 1 1 1866 10 2 1 1 1784 1 2 1 1 1735 2 2 1 1 1778 3 2 1 1 1835 4 2 1 1 1794 5 2 1 1 17 6 2 1 1 1799 7 2 1 1 1765 8 2 1 1 1754 9 2 1 1 1792 10 2 1 1 1821 1 2 1 1 1813 2 2 1 1 1813 3 2 1 1 1749 4 2 1 1 1848 5 2 1 1 1711 6 2 1 1 1779 7 2 1 1 1769 8 2 1 1 1725 9 2 1 1 1814 10 2 1 1 1718 1 2 1 1 1753 2 2 1 1 1773 3 2 1 1 1777 4 2 1 1 1803 5 2 1 1 1718 6 2 1 1 1802 7 2 1 1 1834 8 2 1 1 1731 9 2 1 1 1842 10 2 1 1 1768 1 2 1 1 1698 2 2 1 1 1792 3 2 1 1 174 4 2 1 1 1795 5 2 1 1 1693 6 2 1 1 1743 7 2 1 1 1745 8 2 1 1 1714 9 2 1 1 1791 1 2 1 2

1746 2 2 1 2 1803 3 2 1 2 1767 4 2 1 2 1745 5 2 1 2 1824 6 2 1 2 1752 7 2 1 2 1813 8 2 1 2 1648 9 2 1 2 1818 10 2 1 2 177 1 2 1 2 182 2 2 1 2 1803 3 2 1 2 1739 4 2 1 2 1773 5 2 1 2 1729 6 2 1 2 1744 7 2 1 2 1754 8 2 1 2 1671 9 2 1 2 1741 10 2 1 2 179 1 2 2 2 1712 1 2 2 2 1693 1 2 2 2 1754 1 2 2 2 1757 1 2 2 2 1697 1 2 2 2 1722 1 2 2 2 1884 1 2 2 2 1713 1 2 2 2 1759 1 2 2 2 1749 1 2 2 2 1748 1 2 2 2 1632 1 2 2 2 1752 1 2 2 2 1871 1 2 2 2 1688 1 2 2 2 1789 1 2 2 2 1826 1 2 2 2 1751 1 2 2 2 1738 1 2 2 2 1742 1 2 2 2 1761 1 2 2 2 1909 1 2 2 2 1719 1 2 2 2 1755 1 2 2 2 1745 1 2 2 2 1687 1 2 2 2 1709 1 2 2 2 1625 1 2 2 2 1802 1 2 2 2 1734 1 2 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

61

1809 1 2 2 2 1782 1 2 2 2 1712 1 2 2 2 1668 1 2 2 2 1535 1 2 2 2 1835 1 2 2 2 1725 1 2 2 2 1769 1 2 2 2 178 1 2 2 2 1759 1 2 2 2 1678 1 2 2 2 1836 1 2 2 2 1743 1 2 2 2 1885 1 2 2 2 1704 1 2 2 2 1749 1 2 2 2 1739 1 2 2 2 1735 1 2 2 2 1714 1 2 2 2 1703 1 2 2 2 1772 1 2 2 2 1727 1 2 2 2 1738 1 2 2 2 174 1 2 2 2 1718 1 2 2 2 1701 1 2 2 2 1673 1 2 2 2 1887 1 2 2 2 1743 1 2 2 2 1825 1 2 2 2 1849 1 2 2 2 188 1 2 2 2 1768 1 2 2 2 1713 1 2 2 2 166 1 2 2 2 1775 2 2 2 2 1824 2 2 2 2 167 2 2 2 2 1683 2 2 2 2 1693 2 2 2 2 169 2 2 2 2 1864 2 2 2 2 1709 2 2 2 2 1764 2 2 2 2

1741 2 2 2 2 1797 2 2 2 2 1626 2 2 2 2 1724 2 2 2 2 177 2 2 2 2 1744 2 2 2 2 1716 2 2 2 2 1805 2 2 2 2 1827 2 2 2 2 1752 2 2 2 2 1746 2 2 2 2 1666 2 2 2 2 187 2 2 2 2 1718 2 2 2 2 1483 2 2 2 2 1632 2 2 2 2 1759 2 2 2 2 1723 2 2 2 2 1684 2 2 2 2 1887 2 2 2 2 1756 2 2 2 2 1816 2 2 2 2 1792 2 2 2 2 1614 2 2 2 2 1566 2 2 2 2 1741 2 2 2 2 1752 2 2 2 2 1697 2 2 2 2 1762 2 2 2 2 1801 2 2 2 2 1766 2 2 2 2 1747 2 2 2 2 1744 2 2 2 2 1692 2 2 2 2 186 2 2 2 2 1756 2 2 2 2 1896 2 2 2 2 1814 2 2 2 2 1703 2 2 2 2 1722 2 2 2 2 1688 2 2 2 2 1821 2 2 2 2 1758 2 2 2 2 1787 2 2 2 2

1784 2 2 2 2 179 2 2 2 2 1681 2 2 2 2 1685 2 2 2 2 1798 2 2 2 2 171 2 2 2 2 1638 2 2 2 2 1931 2 2 2 2 1786 2 2 2 2 1768 2 2 2 2 1764 2 2 2 2 1819 2 2 2 2 1666 1 2 3 2 1626 1 2 3 2 188 1 2 3 2 1653 1 2 3 2 1733 1 2 3 2 1551 1 2 3 2 1884 1 2 3 2 1718 1 2 3 2 1549 1 2 3 2 1723 1 2 3 2 194 1 2 3 2 164 1 2 3 2 1565 1 2 3 2 1787 1 2 3 2 1655 2 2 3 2 167 2 2 3 2 1673 2 2 3 2 1705 2 2 3 2 1755 2 2 3 2 1625 2 2 3 2 1708 2 2 3 2 1705 2 2 3 2 1862 2 2 3 2 1686 2 2 3 2 1723 2 2 3 2 1909 2 2 3 2 1859 2 2 3 2 1897 2 2 3 2 1972 2 2 3 2 END

Table C-3 Example OpenBUGS Certification Summary for Component B

mean sd MC_error val25pc median val975pc start sample b 09337 03387 0007504 02695 09371 1592 100001 100000 mu[1] 1764 03101 0007483 1704 1763 1825 100001 100000 mu[2] 1748 008014 4719E-4 1733 1748 1764 100001 100000 mu[3] 1735 02338 0003093 1688 1736 1781 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

62

Certification data for Component C list(Ntot=280) lamb[] asy[] cmp[] typ[] pip[] 163 1 3 1 1 1596 2 3 1 1 1574 3 3 1 1 159 4 3 1 1 161 5 3 1 1 1589 6 3 1 1 1569 7 3 1 1 1565 8 3 1 1 1564 9 3 1 1 1548 10 3 1 1 1706 1 3 1 1 1576 2 3 1 1 1538 3 3 1 1 1555 4 3 1 1 1607 5 3 1 1 1548 6 3 1 1 1572 7 3 1 1 1681 8 3 1 1 1492 9 3 1 1 1655 10 3 1 1 1499 1 3 1 1 1576 2 3 1 1 1545 3 3 1 1 1584 4 3 1 1 1577 5 3 1 1 1599 6 3 1 1 1587 7 3 1 1 152 8 3 1 1 1479 9 3 1 1 1512 10 3 1 1 1497 1 3 1 1 1611 2 3 1 1 156 3 3 1 1 1546 4 3 1 1 1529 5 3 1 1 1478 6 3 1 1 1504 7 3 1 1 1551 8 3 1 1 1438 9 3 1 1 1494 10 3 1 1 1452 1 3 1 1 1347 2 3 1 1 1395 3 3 1 1 1442 4 3 1 1 1448 5 3 1 1 1463 6 3 1 1 1402 7 3 1 1 1474 8 3 1 1 1466 9 3 1 1 1439 10 3 1 1 1424 1 3 1 1 1397 2 3 1 1 135 3 3 1 1 1459 4 3 1 1 1602 5 3 1 1 1435 6 3 1 1 1433 7 3 1 1 1481 8 3 1 1 1482 9 3 1 1

1568 10 3 1 1 147 1 3 1 1 1411 2 3 1 1 1512 3 3 1 1 1452 4 3 1 1 1567 5 3 1 1 148 6 3 1 1 1442 7 3 1 1 143 8 3 1 1 1413 9 3 1 1 1486 10 3 1 1 1444 1 3 1 1 1427 2 3 1 1 1439 3 3 1 1 1468 4 3 1 1 1524 5 3 1 1 1424 6 3 1 1 1459 7 3 1 1 1467 8 3 1 1 1411 9 3 1 1 148 10 3 1 1 1485 1 3 1 1 1422 2 3 1 1 1438 3 3 1 1 1439 4 3 1 1 1515 5 3 1 1 1409 6 3 1 1 1469 7 3 1 1 1453 8 3 1 1 1465 9 3 1 1 1363 10 3 1 1 1458 1 3 1 1 1459 2 3 1 1 1465 3 3 1 1 1468 4 3 1 1 1457 5 3 1 1 1412 6 3 1 1 1474 7 3 1 1 1428 8 3 1 1 1425 9 3 1 1 1393 10 3 1 1 1357 1 3 1 2 1447 2 3 1 2 1448 3 3 1 2 1409 4 3 1 2 1389 5 3 1 2 1415 6 3 1 2 1431 7 3 1 2 1381 8 3 1 2 1403 9 3 1 2 1435 10 3 1 2 1364 1 3 1 2 1361 2 3 1 2 1404 3 3 1 2 1411 4 3 1 2 1429 5 3 1 2 1417 6 3 1 2 138 7 3 1 2 1336 8 3 1 2 1387 9 3 1 2

1372 10 3 1 2 1493 1 3 2 2 1516 1 3 2 2 1451 1 3 2 2 144 1 3 2 2 1523 1 3 2 2 1437 1 3 2 2 1485 1 3 2 2 1433 1 3 2 2 1463 1 3 2 2 1534 1 3 2 2 1533 1 3 2 2 145 1 3 2 2 1537 1 3 2 2 1515 1 3 2 2 1511 1 3 2 2 1449 1 3 2 2 1388 1 3 2 2 1475 1 3 2 2 1468 1 3 2 2 1549 1 3 2 2 153 1 3 2 2 1442 1 3 2 2 1385 1 3 2 2 1513 1 3 2 2 1528 1 3 2 2 1486 1 3 2 2 1472 1 3 2 2 1493 1 3 2 2 1477 1 3 2 2 1405 1 3 2 2 1499 1 3 2 2 1458 1 3 2 2 1514 1 3 2 2 1525 1 3 2 2 1527 1 3 2 2 149 1 3 2 2 1568 1 3 2 2 1452 1 3 2 2 1421 1 3 2 2 1455 1 3 2 2 1525 1 3 2 2 1531 1 3 2 2 144 1 3 2 2 1487 1 3 2 2 151 1 3 2 2 1577 1 3 2 2 1506 1 3 2 2 152 1 3 2 2 1483 1 3 2 2 1522 1 3 2 2 1463 1 3 2 2 1478 1 3 2 2 1505 1 3 2 2 1454 1 3 2 2 1555 1 3 2 2 1497 1 3 2 2 1515 1 3 2 2 1494 1 3 2 2 1594 1 3 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

63

1471 1 3 2 2 1394 1 3 2 2 1445 1 3 2 2 1446 1 3 2 2 1499 1 3 2 2 1484 1 3 2 2 1436 1 3 2 2 1565 2 3 2 2 149 2 3 2 2 1485 2 3 2 2 1529 2 3 2 2 1513 2 3 2 2 1492 2 3 2 2 1475 2 3 2 2 1497 2 3 2 2 1503 2 3 2 2 1448 2 3 2 2 1488 2 3 2 2 1496 2 3 2 2 1499 2 3 2 2 1469 2 3 2 2 1515 2 3 2 2 1543 2 3 2 2 1452 2 3 2 2 1458 2 3 2 2 1435 2 3 2 2 1499 2 3 2 2 1499 2 3 2 2 1589 2 3 2 2 1562 2 3 2 2 1547 2 3 2 2 1505 2 3 2 2 1508 2 3 2 2 152 2 3 2 2

1481 2 3 2 2 1547 2 3 2 2 1501 2 3 2 2 1491 2 3 2 2 1465 2 3 2 2 147 2 3 2 2 1483 2 3 2 2 1479 2 3 2 2 148 2 3 2 2 1543 2 3 2 2 1448 2 3 2 2 1531 2 3 2 2 1478 2 3 2 2 1465 2 3 2 2 1454 2 3 2 2 1591 2 3 2 2 1466 2 3 2 2 157 2 3 2 2 1471 2 3 2 2 1506 2 3 2 2 1491 2 3 2 2 1461 2 3 2 2 1485 2 3 2 2 1495 2 3 2 2 1516 2 3 2 2 1474 2 3 2 2 1483 2 3 2 2 1519 2 3 2 2 1475 2 3 2 2 1466 2 3 2 2 1565 2 3 2 2 1491 2 3 2 2 1435 2 3 2 2 1459 2 3 2 2

1421 2 3 2 2 1468 2 3 2 2 1556 2 3 2 2 1424 1 3 3 2 1427 1 3 3 2 1381 1 3 3 2 1394 1 3 3 2 1473 1 3 3 2 133 1 3 3 2 1446 1 3 3 2 1424 1 3 3 2 133 1 3 3 2 1452 1 3 3 2 1512 1 3 3 2 1314 1 3 3 2 1525 1 3 3 2 1475 1 3 3 2 1542 1 3 3 2 1461 2 3 3 2 1443 2 3 3 2 1547 2 3 3 2 1394 2 3 3 2 1364 2 3 3 2 1395 2 3 3 2 1522 2 3 3 2 1483 2 3 3 2 1452 2 3 3 2 1482 2 3 3 2 1439 2 3 3 2 1458 2 3 3 2 155 2 3 3 2 1502 2 3 3 2 152 2 3 3 2 END

Table C-4 Example OpenBUGS Certification Summary for Component C

mean sd MC_error val25pc median val975pc start sample b 09626 01701 0002428 06285 09633 1293 100001 100000 mu[1] 1399 01566 0002413 1369 1399 143 100001 100000 mu[2] 1491 005701 2386E-4 148 1491 1503 100001 100000 mu[3] 1449 0132 0001181 1423 1449 1475 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

64

34 Results Table C-5 lists the results in CopiesnL for the three components

Table C-5 Parameter Values for Components A B and C

Posterior Distribution Summary Statistics Component Parameter Mean SD 25 50 975

A

β 260 029 203 260 318 μ1 1415 027 1362 1415 1467 μ2 1559 010 1539 1559 158 μ3 1533 016 1501 1533 1565

B

β 095 034 029 095 162 μ1 1762 031 1701 1762 1822 μ2 1748 010 1728 1748 1769 μ3 1735 024 1687 1735 1783

C

β 097 017 064 097 130 μ1 1399 016 1368 1399 1429 μ2 1491 008 1476 1491 1507 μ3 1449 014 1421 1449 1477

Since in each case the three data set types represent independent measurements the three estimates can be combined to obtain a consensus value In this way a between data set type variability is included in the final standard uncertainty Table C-6 lists the results in Copies per Nanoliter obtained using the NIST Consensus Builder (NICOB)C3

Table C-6 Certified Values for Components A B and C

Component Mean u(Mean) Uk=2(Mean) A 151 07 14 B 175 03 06 C 145 05 10

4 Mito-to-nuclear DNA ratios (mtDNAnDNA)

41 Data Measurements were also obtained on the same samples using three mitochondrial detection assays mtND1 mtBatz and mtKav and a single nDNA assay (genomic) to compute the ratios mtDNAnDNA For the complete mtDNAnDNA data see

Table 16 mtDNAnDNA Ratios

42 Model For each component the data was analyzed using the following random effects model to account for between assay variability

119903119903119894119894~119873119873120572120572119894119894 1205901205901198861198861198861198861198861198861198861198861198861198862

where 119894119894 denotes assay (for A and B 1 = mtND1 2 = mtBatz 3 = mtKav for C 1 = mtBatz 2 = mtKav) and 120572120572119894119894~119873119873micro120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119886119886119886119886119886119886119886119886119886119886

2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

65

43 OpenBUGS code inits and data Two versions of the same model are required one for the results from the three mtDNA assays available for components A and B mtND1 mtBatz and mtKav and the other for the results from the two mtDNA assays available for component C mtBatz and mtKav mtDNAnDNA code for components A and B ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 13)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for components A and B list(sigt=c(111)) Data for Component A list(Ndata=102) set[] rep[] asy[] rat[] 1 1 1 1931 1 2 1 187 1 1 2 1845 1 2 2 1764 2 1 1 181 2 2 1 1906 2 1 2 1785 2 2 2 1837 3 1 1 1743 3 2 1 1761 3 1 2 1829 3 2 2 1847 4 1 1 1716 4 2 1 180 4 1 2 174 4 2 2 1813 5 1 1 1923 5 2 1 1835 5 1 2 1935 5 2 2 1851 6 1 1 1757 6 2 1 1724 6 3 1 1768 6 1 2 1743 6 2 2 1671 6 3 2 1766 6 1 3 1692 6 2 3 1662 6 3 3 1767 7 1 1 1924 7 2 1 1944 7 3 1 1893 7 1 2 1786

7 2 2 1752 7 3 2 1743 7 1 3 153 7 2 3 1542 7 3 3 1523 7 1 2 158 7 2 2 1557 7 3 2 1592 7 1 3 2103 7 2 3 210 7 3 3 2049 7 1 1 1749 7 2 1 1769 7 3 1 1777 7 1 2 1965 7 2 2 1889 7 3 2 1947 7 1 3 1813 7 2 3 1864 7 3 3 1864 8 1 1 178 8 2 1 1776 8 3 1 1706 8 1 2 1804 8 2 2 1751 8 3 2 1738 8 1 3 1769 8 2 3 1786 8 3 3 1761 9 1 2 1584 9 2 2 1544 10 1 1 1947 10 2 1 1928 10 3 1 1922 10 1 1 1829

10 2 1 1895 10 3 1 1851 10 1 1 1674 10 2 1 1617 11 1 2 1667 11 2 2 1813 11 3 2 1708 11 4 2 1598 11 1 2 1566 11 2 2 1687 11 3 2 1663 11 4 2 1589 11 1 2 1643 11 2 2 1668 11 3 2 1604 11 4 2 1573 12 1 2 1719 12 2 2 1716 12 3 2 167 12 4 2 1666 12 1 2 1581 12 2 2 1573 12 3 2 1601 12 4 2 1612 12 1 2 1546 12 2 2 1542 12 3 2 1538 12 4 2 1587 13 1 1 1617 13 2 1 1599 13 1 1 1622 13 2 1 1579 13 1 1 1575 13 2 1 1528 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

66

Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A

mean sd MC_error val25pc median val975pc start sample mu 1730 1443 004419 1703 1730 1759 100001 100000

data for Component B list(Ndata=87) set[] rep[] asy[] rat[] 1 1 1 2461 1 2 1 2401 1 1 2 2555 1 2 2 2608 2 1 1 2229 2 2 1 2200 2 1 2 2108 2 2 2 2192 3 1 1 2127 3 2 1 2198 3 1 2 2455 3 2 2 2506 4 1 1 2211 4 2 1 2199 4 1 2 2380 4 2 2 2249 5 1 1 2105 5 2 1 1943 5 1 2 2136 5 2 2 2119 6 1 1 1990 6 2 1 1921 6 3 1 1982 6 1 2 1990 6 2 2 1924 6 3 2 1958 6 1 3 1903 6 2 3 1831

6 3 3 1919 7 1 1 2304 7 2 1 2341 7 3 1 2332 7 1 2 2048 7 2 2 2078 7 3 2 2067 7 1 3 2399 7 2 3 2479 7 3 3 2493 7 1 1 2065 7 2 1 2166 7 3 1 2164 7 1 2 2314 7 2 2 2280 7 3 2 2258 7 1 3 1867 7 2 3 1987 7 3 3 1971 8 1 1 2046 8 2 1 2057 8 3 1 2024 8 1 2 1998 8 2 2 2043 8 3 2 1952 8 1 3 2077 8 2 3 2117 8 3 3 2041 9 1 2 1890 9 2 2 1886

10 1 2 1928 10 2 2 1916 10 3 2 1915 10 1 2 1947 10 2 2 1946 10 3 2 1961 10 4 2 1909 10 1 2 1953 10 2 2 1952 10 3 2 1941 10 4 2 1881 11 1 2 1865 11 2 2 1863 11 3 2 1828 11 4 2 1828 11 1 2 1944 11 2 2 1964 11 3 2 1947 11 4 2 1895 11 1 2 1908 11 2 2 1909 11 3 2 1940 11 4 2 1839 12 1 1 1688 12 2 1 1783 12 1 1 1823 12 2 1 1631 12 1 1 1920 12 2 1 1887 END

Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B

mean sd MC_error val25pc median val975pc start sample mu 2045 2246 0111 1999 2045 2088 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

67

mtDNAnDNA code for component C ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 12)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for component C list(sigt=c(11)) mtDNAnDNA data for component C list(Ndata=66) set[] rep[] asy[] rat[] 1 1 1 3201 1 2 1 3299 2 1 1 3347 2 2 1 3287 3 1 1 3217 3 2 1 3384 4 1 1 3200 4 2 1 3363 5 1 1 3294 5 2 1 3013 6 1 1 2805 6 2 1 2825 6 3 1 2829 6 1 2 2709 6 2 2 2738 6 3 2 2727 7 1 1 2797 7 2 1 2642 7 3 1 2792 7 1 2 2804 7 2 2 2789

7 3 2 2842 7 1 1 2715 7 2 1 2606 7 3 1 2687 7 1 2 2799 7 2 2 2628 7 3 2 2559 7 1 1 2639 7 2 1 2535 7 3 1 2657 7 1 2 2953 7 2 2 2757 7 3 2 2802 8 1 1 2927 8 2 1 2776 8 3 1 2762 8 1 2 3044 8 2 2 2830 8 3 2 2825 9 1 1 2501 9 2 1 2522 10 1 1 2832 10 2 1 2840

10 3 1 2788 10 4 1 2725 10 1 1 2835 10 2 1 2790 10 3 1 2736 10 4 1 2679 10 1 1 2803 10 2 1 2773 10 3 1 2757 10 4 1 2752 11 1 1 2599 11 2 1 2508 11 3 1 2551 11 4 1 2609 11 1 1 2617 11 2 1 2483 11 3 1 2626 11 4 1 2602 11 1 1 2665 11 2 1 2535 11 3 1 2570 11 4 1 2650 END

Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C

mean sd MC_error val25pc median val975pc start sample mu 2774 2367 01176 2729 2774 2821 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

68

Table C-10 Results for mtDNAnDNA Posterior Distribution Summary Statistics

Component Mean Std Dev 25 50 975 A 1730 14 1703 1730 1758 B 2045 23 1994 2047 2087 C 2774 23 2729 2774 2821

5 ReferencesC1 Lunn DJ Spiegelhalter D Thomas A Best N Statistics in Medicine 2009283049ndash3082C2 Polson NG Scott JG Bayesian Analysis 20127887ndash902C3 NIST Consensus Builder Software and userrsquos guide freely available at

httpsconsensusnistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

69

Appendix D Performance in Commercial qPCR Chemistries The performance of the SRM 2372a materials in commercial qPCR chemistries was assessed using manufacturerrsquos recommended protocols Each panel in Figures D-1 to D-4 displays the average cycle threshold (Ct) as a function of DNA concentration for the 5-point dilutions of SRM 2372a components In all cases the external calibrant was SRM 2372 component A Error bars represent approximate 95 confidence intervals Table D-1 lists the estimated [DNA] for the three SRM 2372a components for all commercial assays evaluated This Table also lists the Component BComponent A (BA) and Component CComponent A (CA) ratios calculated from the [DNA] estimates

Figure D-1 Results for the Innogenomics Commercial qPCR assays

20

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantLong(IQ_h)

ABCStd

[A] = (384 plusmn 26) ngμL[B] = (441 plusmn 24) ngμL[C] = (516 plusmn 24) ngμL

Rsup2 = 0999 se(Ct) = 01016

17

18

19

20

21

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantShort(IQ_d)

ABCStd

[A] = (435 plusmn 28) ngμL[B] = (543 plusmn 12) ngμL[C] = (510 plusmn 28) ngμL

Rsup2 = 0999 se(Ct) = 007

24

25

26

27

28

29

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYLong(IQHY_h)

ABCStd

[A] = (387 plusmn 24) ngμL[B] = (448 plusmn 21) ngμL[C] = (559 plusmn 31) ngμL

Rsup2 = 0998 se(Ct) = 01314

16

18

20

22

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYShort(IQHY_d)

ABCStd

[A] = (353 plusmn 23) ngμL[B] = (267 plusmn 15) ngμL[C] = (390 plusmn 18) ngμL

Rsup2 = 1000 se(Ct) = 00420

22

24

26

28

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYY(IQHY_y)

ABCStd

[A] = (572 plusmn 13) ngμL[B] = 0 ngμL[C] = (137 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

70

Figure D-2 Results for the Thermo Fisher Commercial qPCR assays

21

22

23

24

25

26

27

28

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HumanHuman(QFH_h)

ABCStd

[A] = (464 plusmn 13) ngμL[B] = (529 plusmn 19) ngμL[C] = (472 plusmn 22) ngμL

Rsup2 = 1000 se(Ct) = 004

22

23

24

25

26

27

28

29

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoHuman(QFD_h)

ABCStd

[A] = (701 plusmn 53) ngμL[B] = (561 plusmn 11) ngμL[C] = (472 plusmn 11) ngμL

Rsup2 = 0994 se(Ct) = 01922

24

26

28

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoMale(QFD_y)

ABCStd

[A] = (706 plusmn 49) ngμL[B] = 0 ngμL[C] = (122 plusmn 02) ngμL

Rsup2 = 0999 se(Ct) = 007

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPLarge(QFHP_h)

ABCStd

[A] = (340 plusmn 18) ngμL[B] = (432 plusmn 05) ngμL[C] = (342 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 00720

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPSmall(QFHP_d)

ABCStd

[A] = (434 plusmn 15) ngμL[B] = (514 plusmn 10) ngμL[C] = (466 plusmn 07) ngμL

Rsup2 = 1000 se(Ct) = 005

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioLarge(QFT_h)

ABCStd

[A] = (362 plusmn 23) ngμL[B] = (430 plusmn 06) ngμL[C] = (378 plusmn 07) ngμL

Rsup2 = 0999 se(Ct) = 00920

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioSmall(QFT_d)

ABCStd

[A] = (468 plusmn 08) ngμL[B] = (505 plusmn 08) ngμL[C] = (523 plusmn 14) ngμL

Rsup2 = 1000 se(Ct) = 00318

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioY(QFT_y)

ABCStd

[A] = (372 plusmn 11) ngμL[B] = 0 ngμL[C] = (110 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 003

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

71

Figure D-3 Results for the Qiagen Commercial qPCR assays

Figure D-4 Results for the Promega Commercial qPCR assays

19

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresAutosomal(QPH_h)

ABCStd

[A] = (554 plusmn 21) ngμL[B] = (479 plusmn 05) ngμL[C] = (381 plusmn 09) ngμL

Rsup2 = 1000 se(Ct) = 00418

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresY(QPH_y)

ABCStd

[A] = (358 plusmn 10) ngμL[B] = 0 ngμL[C] = (118 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 007

16

17

18

19

20

21

22

23

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProHuman(QPP_h)

ABCStd

[A] = (525 plusmn 49) ngμL[B] = (480 plusmn 42) ngμL[C] = (467 plusmn 45) ngμL

Rsup2 = 0999 se(Ct) = 01019

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProDegradation(QPP_d)

ABCStd

[A] = (567 plusmn 18) ngμL[B] = (558 plusmn 08) ngμL[C] = (484 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 00410

15

20

25

30

35

-50 -40 -30 -20 -10 00 10

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProY(QPP_y)

ABCStd

[A] = (538 plusmn 27) ngμL[B] = (00 plusmn 00) ngμL[C] = (162 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 003

16

18

20

22

24

-05 00 05 10 15 20

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorAutosomal(P_h)

ABCStd

[A] = (759 plusmn 64) ngμL[B] = (775 plusmn 24) ngμL[C] = (1056 plusmn 17) ngμL

Rsup2 = 0995 se(Ct) = 01818

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorY(P_y)

ABCStd

[A] = (596 plusmn 51) ngμL[B] = 0 ngμL[C] = (117 plusmn 01) ngμL

Rsup2 = 0994 se(Ct) = 020

18

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantAutosomal(PQ_h)

ABCStd

[A] = (436 plusmn 13) ngμL[B] = (606 plusmn 20) ngμL[C] = (504 plusmn 10) ngμL

Rsup2 = 1000 se(Ct) = 00218

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantDegradation(PQ_d)

ABCStd

[A] = (420 plusmn 18) ngμL[B] = (529 plusmn 16) ngμL[C] = (530 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 00516

18

20

22

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantY(PQ_y)

ABCStd

[A] = (410 plusmn 20) ngμL[B] = 0 ngμL[C] = (139 plusmn 05) ngμL

Rsup2 = 0999 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

72

Table D-1 Summary [DNA] Results for the Commercial qPCR Assays Estimated [DNA] ngμL Ratios

A B C BA CA Assaya xb u(x)c xb u(x)c xb u(x)c yd u(y)e yd u(y)e

IQd 435 28 543 12 510 28 125 008 117 010 IQh 384 26 441 24 516 24 115 010 134 011

IQHYd 353 23 267 15 390 18 076 006 111 009 IQHYh 387 24 448 21 559 31 116 009 144 012

Ph 759 64 775 24 1056 16 102 009 139 012 PQd 420 18 529 16 530 06 126 007 126 006 PQh 436 13 606 20 504 10 139 006 116 004

QFDh 701 53 561 11 472 11 080 006 067 005 QFHh 464 13 529 19 472 22 114 005 102 006

QFHPd 434 15 514 10 466 07 118 005 107 004 QFHPh 340 18 432 05 342 03 127 007 101 005

QFTd 468 08 505 08 523 14 108 003 112 004 QFTh 362 23 430 06 378 07 118 008 104 007 QPHh 554 21 479 05 380 09 086 003 069 003 QPPd 567 18 558 08 484 11 098 003 085 003 QPPh 525 49 480 42 467 45 092 012 089 012

IQHYy 572 13 NDf 137 06

NRg

024 001 Py 596 51 NDf 117 01 020 002

PQy 410 20 NDf 139 05 034 002 QFDy 706 49 NDf 122 02 017 001 QFTy 372 11 NDf 110 06 030 002 QPHy 358 10 NDf 118 03 033 001 QPPy 538 27 0002 0001 162 11 030 003

a Assay codes are listed in Table 8 b Mean of dilution-adjusted results from five dilutions of the SRM 2372a component c Standard uncertainty of the mean estimated from the stdanrd deviation of the five dilutions d Ratio of the estimated [DNA] of component B or C relative to that of component A e Standard uncertainty of the ratio estimated from the standard uncertainties of the individual components f Not detected Component B was prepared using tissue from only female donors it is not expected to

respond to male-specific assays g Not relevant

  • Abstract
  • Keywords
  • Technical Information Contact for this SRM
  • Table of Contents
  • Table of Tables
  • Table of Figures
  • Purpose and Description
  • Warning SRM 2372a is a Human Source Material
  • Storage and Use
  • History and Background
    • Figure 1 Sales History of SRM 2372
      • Experimental Methods
        • Polymerase Chain Reaction Assays
          • Table 1 NIST-Developed Human Nuclear DNA Assays
          • Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays
          • Table 2 NIST-Optimized Human Mitochondrial DNA Assays
          • Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays
            • Droplet Digital Polymerase Chain Reaction Measurements
              • Nuclear DNA (nDNA)
                • Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays
                  • Mitochondrial DNA (mtDNA)
                    • Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays
                    • Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays
                        • Sample Preparation
                        • Absorbance Measurements
                        • SRM Unit Production
                        • Component Homogeneity
                          • Figure 4 Plate Layout Design for Homogeneity Measurements
                            • Component Stability
                              • Figure 5 Plate Layout Design for Stability Measurements
                                • Component Certification Measurements
                                  • Table 6 Tubes used for Certification Measurements
                                  • Figure 6 Basic Plate Layout Design for Certification Measurements
                                    • Mitochondrial DNA Measurements
                                      • Table 7 Run Parameters for Mitochondrial DNA
                                        • Performance in Commercial qPCR Chemistries
                                          • Table 8 Commercial qPCR Chemistries Tested with SRM 2372a
                                          • Figure 7 Plate Layout Design for Performance Assessments
                                              • Measurement Results
                                                • Absorbance Spectrophotometry
                                                  • Table 9 Absorbance at Selected Wavelengths
                                                  • Figure 8 Absorbance Spectra of SRM 2372a Components A B and C
                                                    • Droplet Volume
                                                    • Certification Measurements
                                                      • Homogeneity Results
                                                        • Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component
                                                        • Table 11 Within- and Between-Tube Relative Standard Deviations
                                                          • Stability Results
                                                            • Figure 9 ddPCR Stability Measurements as Function of Time
                                                            • Table 12 Stability Data as λacute Copy Number per Nanoliter of Component
                                                              • Ten-Assay Certification Results
                                                                • Figure 10 Comparison of Assay Performance
                                                                • Table 13 Certification Data as λacute Copy Number per Nanoliter of Component
                                                                  • Estimation of Copy Number Concentration
                                                                    • Table 14 Measured Copy Number Concentrations
                                                                      • Conversion of Copy Number to Mass Concentration
                                                                        • Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution
                                                                            • Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)
                                                                              • Figure 11 mtDNAnDNA for Components A B and C
                                                                              • Table 16 mtDNAnDNA Ratios
                                                                              • Table 17 mtDNAnDNA Summary Results
                                                                                • Performance in Commercial qPCR Chemistries
                                                                                  • Figure 12 qPCR Autosomal Quantitation Performance Summary
                                                                                  • Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary
                                                                                      • Qualification of ddPCR as Fit-For-Purpose
                                                                                        • Figure 14 Absorbance Spectra of SRM 2372 Components A and B
                                                                                        • Digital Polymerase Chain Reaction Assays
                                                                                        • Heat Treatments (Melting of the DNA into Single Strands)
                                                                                        • Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a
                                                                                          • Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a
                                                                                            • Quantitative Evaluation of ssDNA Content
                                                                                              • ddPCR Measurement of ssDNA Proportion
                                                                                              • Proof of Principle
                                                                                                • Table 19 Comparison of Snap-Cooling to Slow-Cooling
                                                                                                  • Time at Denaturation Temperature
                                                                                                    • Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures
                                                                                                      • Temperature
                                                                                                        • Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a
                                                                                                          • DNA Concentration in the ddPCR Solution
                                                                                                            • Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution
                                                                                                              • Equivalence of Assays
                                                                                                                • Table 23 Comparison of Assays
                                                                                                                    • ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials
                                                                                                                      • Table 24 Evaluation of Limiting Ratio
                                                                                                                          • Conclusions and Recommendations
                                                                                                                            • Recommended Certification Values
                                                                                                                              • Table 25 Recommended Values and 95 Uncertainties for Certification
                                                                                                                                • Use of SRM 2372a for Value-Assigning In-House Reference Materials
                                                                                                                                  • Certificate of Analysis
                                                                                                                                  • References
                                                                                                                                  • Appendix A DNA Extraction Method
                                                                                                                                    • Procedure
                                                                                                                                    • Reagents
                                                                                                                                    • Reference
                                                                                                                                      • Appendix B Spectrophotometric Measurement Protocol
                                                                                                                                        • Spectrophotometer Calibration
                                                                                                                                          • Establish Baseline
                                                                                                                                          • Enable Verification of Wavelength Calibration
                                                                                                                                          • Enable Verification of Absorbance Calibration
                                                                                                                                            • For ldquonativerdquo dsDNA Samples
                                                                                                                                              • Verify Optical Balance of Cuvettes
                                                                                                                                                • For NaOH denatured ssDNA Samples
                                                                                                                                                  • Prepare a 04 molL NaOH Stock Solution
                                                                                                                                                  • Prepare the Sample Solutions
                                                                                                                                                  • Measure Sample Solutions
                                                                                                                                                    • After All Sample Measurements are Complete
                                                                                                                                                      • Enable Verification of Spectrophotometer Stability
                                                                                                                                                      • Clean Up
                                                                                                                                                        • Evaluation
                                                                                                                                                        • Reference
                                                                                                                                                          • Appendix C SRM 2372a Statisticianrsquos report
                                                                                                                                                            • 1 Introduction
                                                                                                                                                            • 11 OpenBUGS Analysis
                                                                                                                                                            • 12 OpenBUGS Exemplar Output
                                                                                                                                                            • 2 Homogeneity assessment
                                                                                                                                                            • 21 Data
                                                                                                                                                            • 22 Model
                                                                                                                                                            • 23 OpenBUGS code and data
                                                                                                                                                              • Table C-1 Example OpenBUGS Homogeneity Summary
                                                                                                                                                                • 24 Results
                                                                                                                                                                  • Figure C-1 Homogeneity Assessment
                                                                                                                                                                    • 3 Certification measurements
                                                                                                                                                                    • 31 Data
                                                                                                                                                                    • 32 Model
                                                                                                                                                                    • 33 OpenBUGS code inits and data
                                                                                                                                                                      • Table C-2 Example OpenBUGS Certification Summary for Component A
                                                                                                                                                                      • Table C-3 Example OpenBUGS Certification Summary for Component B
                                                                                                                                                                      • Table C-4 Example OpenBUGS Certification Summary for Component C
                                                                                                                                                                        • 34 Results
                                                                                                                                                                          • Table C-5 Parameter Values for Components A B and C
                                                                                                                                                                          • Table C-6 Certified Values for Components A B and C
                                                                                                                                                                            • 4 Mito-to-nuclear DNA ratios (mtDNAnDNA)
                                                                                                                                                                            • 41 Data
                                                                                                                                                                            • 42 Model
                                                                                                                                                                            • 43 OpenBUGS code inits and data
                                                                                                                                                                              • Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A
                                                                                                                                                                              • Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B
                                                                                                                                                                              • Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C
                                                                                                                                                                              • Table C-10 Results for mtDNAnDNA
                                                                                                                                                                                • 5 References
                                                                                                                                                                                  • Appendix D Performance in Commercial qPCR Chemistries
                                                                                                                                                                                    • Table D-1 Summary [DNA] Results for the Commercial qPCR Assays
Page 2: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .

NIST Special Publication 260-189

Certification of Standard Reference Materialreg 2372a

Human DNA Quantitation Standard Erica L Romsos

Margaret C Kline Biomolecular Measurement Division

Material Measurement Laboratory

David L Duewer Chemical Sciences Division

Material Measurement Laboratory

Blaza Toman Statistical Engineering Division

Information Technology Laboratory

Natalia Farkas Engineering Physics Division

Physical Measurement Laboratory

This publication is available free of charge from httpsdoiorg106028NISTSP260-189

March 2018

US Department of Commerce Wilbur L Ross Jr Secretary

National Institute of Standards and Technology Walter Copan Under Secretary of Commerce for Standards and Technology and Director

Certain commercial entities equipment or materials may be identified in this document to describe an experimental procedure or concept adequately

Such identification is not intended to imply recommendation or endorsement by the National Institute of Standards and Technology nor is it intended to imply that the entities materials or equipment are necessarily the best available for the purpose

National Institute of Standards and Technology Special Publication 260-189 Natl Inst Stand Technol Spec Publ 260-189 83 pages (March 2018)

CODEN NSPUE2

This publication is available free of charge from httpsdoiorg106028NISTSP260-189

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

i

Abstract Standard Reference Material (SRM) 2372a is intended for use in the value assignment of human genomic deoxyribonucleic acid (DNA) forensic quantitation materials A unit of SRM 2372a consists of three well-characterized human genomic DNA materials solubilized in 10 mmolL 2-amino-2-(hydroxymethyl)-13 propanediol hydrochloride (Tris HCl) and 01 mmolL ethylenediaminetetraacetic acid disodium salt (disodium EDTA) pH 80 buffer (TE-4) The three component genomic DNA materials labeled A B and C are respectively derived from a single male donor a single female donor and 13 mixture of a male and a female donor A unit of the SRM consists of one 05 mL tube of each component each tube containing approximately 55 microL of DNA solution Each of these tubes is labeled and is sealed with a color-coded screw cap This publication documents the production analytical methods and statistical evaluations involved in production of this SRM

Keywords Human mitochondrial DNA (mtDNA)

Human nuclear DNA (nDNA) Standard Reference Material (SRM)

Technical Information Contact for this SRM Please address technical questions about this SRM to srmsnistgov where they will be assigned to the appropriate Technical Project Leader responsible for support of this material For sales and customer service inquiries please contact srminfonistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

ii

Table of Contents Purpose and Description 1 Warning SRM 2372a is a Human Source Material 1 Storage and Use 1 History and Background 2 Experimental Methods 4

Polymerase Chain Reaction Assays 4 Droplet Digital Polymerase Chain Reaction Measurements 7

Nuclear DNA (nDNA) 7 Mitochondrial DNA (mtDNA) 8

Sample Preparation 9 Absorbance Measurements 10 SRM Unit Production 10 Component Homogeneity 10 Component Stability 11 Component Certification Measurements 12 Mitochondrial DNA Measurements 13 Performance in Commercial qPCR Chemistries 13

Measurement Results 15 Absorbance Spectrophotometry 15 Droplet Volume 16 Certification Measurements 16

Homogeneity Results 16 Stability Results 18 Ten-Assay Certification Results 20 Estimation of Copy Number Concentration 22 Conversion of Copy Number to Mass Concentration 22

Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA) 23 Performance in Commercial qPCR Chemistries 25

Qualification of ddPCR as Fit-For-Purpose 27 Digital Polymerase Chain Reaction Assays 28 Heat Treatments (Melting of the DNA into Single Strands) 28 Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a 29 Quantitative Evaluation of ssDNA Content 30

ddPCR Measurement of ssDNA Proportion 30 Proof of Principle 31 Time at Denaturation Temperature 33 Temperature 35 DNA Concentration in the ddPCR Solution 36 Equivalence of Assays 38

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials 40 Conclusions and Recommendations 43

Recommended Certification Values 43 Use of SRM 2372a for Value-Assigning In-House Reference Materials 44

Certificate of Analysis 44 References 44

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

iii

Appendix A DNA Extraction Method 46 Procedure 46 Reagents 46 Reference 47

Appendix B Spectrophotometric Measurement Protocol 48 Spectrophotometer Calibration 48

Establish Baseline 48 Enable Verification of Wavelength Calibration 48 Enable Verification of Absorbance Calibration 48

For ldquonativerdquo dsDNA Samples 48 Verify Optical Balance of Cuvettes 48

For NaOH denatured ssDNA Samples 49 Prepare a 04 molL NaOH Stock Solution 49 Prepare the Sample Solutions 49 Measure Sample Solutions 50

After All Sample Measurements are Complete 50 Enable Verification of Spectrophotometer Stability 50 Clean Up 50

Evaluation 50 Appendix C SRM 2372a Statisticianrsquos report 51

1 Introduction 51 11 OpenBUGS Analysis 51 12 OpenBUGS Exemplar Output 51 2 Homogeneity assessment 52 21 Data 52 22 Model 52 23 OpenBUGS code and data 52 24 Results 55 3 Certification measurements 57 31 Data 57 32 Model 57 33 OpenBUGS code inits and data 58 34 Results 64 4 Mito-to-nuclear DNA ratios (mtDNAnDNA) 64 41 Data 64 42 Model 64 43 OpenBUGS code inits and data 65 5 References 68

Appendix D Performance in Commercial qPCR Chemistries 69

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

iv

Table of Tables Table 1 NIST-Developed Human Nuclear DNA Assays 5 Table 2 NIST-Optimized Human Mitochondrial DNA Assays 6 Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays 8 Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays 8 Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays 8 Table 6 Tubes used for Certification Measurements 12 Table 7 Run Parameters for Mitochondrial DNA 13 Table 8 Commercial qPCR Chemistries Tested with SRM 2372a 14 Table 9 Absorbance at Selected Wavelengths 15 Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component 17 Table 11 Within- and Between-Tube Relative Standard Deviations 18 Table 12 Stability Data as λacute Copy Number per Nanoliter of Component 19 Table 13 Certification Data as λacute Copy Number per Nanoliter of Component 21 Table 14 Measured Copy Number Concentrations 22 Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution 23 Table 16 mtDNAnDNA Ratios 24 Table 17 mtDNAnDNA Summary Results 25 Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a 29 Table 19 Comparison of Snap-Cooling to Slow-Cooling 32 Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures 34 Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 36 Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution 37 Table 23 Comparison of Assays 39 Table 24 Evaluation of Limiting Ratio 41 Table 25 Recommended Values and 95 Uncertainties for Certification 43 Table C-1 Example OpenBUGS Homogeneity Summary 55 Table C-2 Example OpenBUGS Certification Summary for Component A 60 Table C-3 Example OpenBUGS Certification Summary for Component B 61 Table C-4 Example OpenBUGS Certification Summary for Component C 63 Table C-5 Parameter Values for Components A B and C 64 Table C-6 Certified Values for Components A B and C 64 Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A 66 Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B 66 Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C 67 Table C-10 Results for mtDNAnDNA 68 Table D-1 Summary [DNA] Results for the Commercial qPCR Assays 72

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

i

Table of Figures Figure 1 Sales History of SRM 2372 3 Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays 6 Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays 7 Figure 4 Plate Layout Design for Homogeneity Measurements 11 Figure 5 Plate Layout Design for Stability Measurements 11 Figure 6 Basic Plate Layout Design for Certification Measurements 12 Figure 7 Plate Layout Design for Performance Assessments 14 Figure 8 Absorbance Spectra of SRM 2372a Components A B and C 16 Figure 9 ddPCR Stability Measurements as Function of Time 18 Figure 10 Comparison of Assay Performance 20 Figure 11 mtDNAnDNA for Components A B and C 24 Figure 12 qPCR Autosomal Quantitation Performance Summary 26 Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary 26 Figure 14 Absorbance Spectra of SRM 2372 Components A and B 27 Figure 15 Comparison of Snap-Cooling to Slow-Cooling 32 Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures 33 Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 35 Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution 37 Figure 19 Comparison of Assays 38 Figure 20 Evaluation of Limiting Ratio 40 Figure 21 Degrees of Equivalence 42 Figure C-1 Homogeneity Assessment 56 Figure D-1 Results for the Innogenomics Commercial qPCR assays 69 Figure D-2 Results for the Thermo Fisher Commercial qPCR assays 70 Figure D-3 Results for the Qiagen Commercial qPCR assays 70 Figure D-4 Results for the Promega Commercial qPCR assays 71

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

1

Purpose and Description Standard Reference Material (SRM) 2372a is intended for use in the value assignment of human genomic deoxyribonucleic acid (DNA) forensic quantitation materials A unit of SRM 2372a consists of three well-characterized human genomic DNA materials solubilized in 10 mmolL 2-amino-2-(hydroxymethyl)-13 propanediol hydrochloride (Tris HCl) and 01 mmolL ethylenediaminetetraacetic acid disodium salt (disodium EDTA) pH 80 buffer (TE-4) The three component genomic DNA materials labeled A B and C are respectively derived from a single male donor a single female donor and 13 mixture of a male and a female donor A unit of the SRM consists of one 05 mL tube of each component each tube containing approximately 55 microL of DNA solution Each of these tubes is labeled and is sealed with a color-coded screw cap NIST is guided by and adheres to the ethical principles set forth in the Belmont Report1 SRM 2372a was developed after an appropriate human subject research determination by the NIST Human Subjects Protections Office

Warning SRM 2372a is a Human Source Material The three SRM 2372a components are human genomic DNA extracts prepared at NIST from buffy coat white blood cells from single-source anonymous donors All buffy coats were purchased from Interstate Blood Bank (Memphis TN) The supplier reported that each donor unit used in the preparation of this product was tested by FDA-licensed tests and found to be negative for human immunodeficiency virus (HIV-1 RNA and Anti-HIV 12) hepatitis B (HBV DNA and HBsAg) hepatitis C (HCV RNA and Anti-HCV) West Nile virus (WNV RNA) and syphilis Additionally each donor was tested according to FDA guidelines for Trypanosoma cruzi (Chagas) and found to be negativenon-reactive However no known test method can offer complete assurance that infectious agents are absent from this material Accordingly this human blood-based product should be handled at the Biosafety Level 12 SRM 2372a components and derived solutions should be disposed of in accordance with local state and federal regulations

Storage and Use Until required for use SRM 2372a should be stored in the dark between 2 degC and 8 degC The SRM 2372a component tubes should be mixed briefly and centrifuged (without opening the tube cap) prior to removing sample aliquots for analysis For the certified values to be applicable materials should be withdrawn immediately after opening the tubes and processed without delay Certified values do not apply to any material remaining in recapped tubes The certification only applies to the initial use and the same results are not guaranteed if the remaining material is used later

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

2

History and Background SRM 2372a is the second member of the SRM 2372 Human DNA Quantitation Standard series The supply of the initial material SRM 2372 was exhausted in May 2017 SRM 2372a is derived from commercially obtained buffy coat cells from different donors than those used in SRM 2372 A series of interlaboratory studies coordinated by what was then the Biotechnology Division of the National Institute of Standards and Technology (NIST) during the 1990s demonstrated that the DNA quantitation technologies then in use by the forensic human identification community adequately met the needs of the DNA typing tools then available34 However our 2004 study demonstrated that the increasingly sensitive and informative polymerase chain reaction (PCR)-based multiplex typing tools required more reliable determinations of DNA input quantity for best performance5 Quantitative PCR (qPCR) methods for DNA quantitation require calibration to an external standard of defined DNA concentration ([DNA]) expressed in units of nanograms of DNA per microliter of extract solution The true [DNA] of a commercial secondary standard may vary from its nominal concentration depending on supplier and lot number To enable providers of secondary standards to more reliably evaluate their materials and forensic testing laboratories to validate their use of commercial quantitation methods and supplies in 2007 we delivered the human DNA calibration standard SRM 2372 Human DNA Quantitation Standard6 The quantity of DNA in each of the three components of SRM 2372 was originally assigned based on the absorbance of double-stranded DNA (dsDNA) at 260 nm7 However by 2012 the absorbance properties of the materials had increased beyond the certified uncertainty limits due to slow but progressive changes in DNA tertiary structure Sales of the SRM were restricted until the DNA quantities were reassigned based on measurements of single-stranded DNA (ssDNA) in early 20138-10 Figure 1 displays the sales history of this relatively high-selling SRM While SRM 2372 was fit for the practical purpose of harmonizing DNA quantitation measurements to an accessible and reproducible reference material the conversion of both dsDNA and ssDNA absorbance to mass [DNA] is through conventional proportionality constants that are not metrologically traceable to the modern reference system for measurements the International System of Units (SI) Responding to requests from the forensic and clinical communities for calibration standards that provide traceability to the SI we and others have established droplet digital polymerase chain reaction (ddPCR) as a potentially primary measurement process that can provide traceable results11-18 The copy numbers of human nuclear DNA (nDNA) in the reaction mixture per partition symbolized as λ were determined by ddPCR using ten PCR assays optimized for use with our ddPCR platform These assays target loci on eight chromosomes with three assays on widely separated locations on one chromosome (see Table 1)

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

3

Figure 1 Sales History of SRM 2372

Each small black dot represents the sale of one unit of SRM 2372 The red lines summarize the average rate of sales as originally certified (2007 to 2012) and after re-certification (2013 to 2017) The magenta labels state the average yearly sales rate during these two periods

The nDNA copy number concentration human haploid genome equivalents (HHGE) per nanoliter component solution and symbolized here as λacute is calculated as

copy number [DNA] = λ = λ(VF) HHGEnL

where V is the average droplet volume and F is the total dilution factor the volume fraction of the component solution in the ddPCR reaction mixture The manufacturer of the ddPCR platform used a Bio-Rad QX200 Droplet Digital PCR System (Hercules CA) system does not provide metrologically traceable droplet volume information This lack of information prevented metrologically traceable conversion of the number per partition measurements to the desired number per reaction volume units In consequence NIST developed a measurement method that provides metrologically traceable droplet volumes19 We show elsewhere that the mass [DNA] nanograms of human DNA per microliter component can be traceably estimated as

mass [DNA] = 3031timesλacute ngmicroL

where the constant is the product of the number of nucleotide basepairs (bp) in the human reference genome the average bp molecular weight Avogadrorsquos number of bp per mole and

2372

2372(Recertif ied)

050

010

00

1713 unitsyear

1416 unitsyear

2007 2009 2011 2013 2015 2017

Uni

ts S

old

Calendar Year

Human DNA Quantitation StandardSRM 2372

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

4

unit scaling factors18 The three components of SRM 2372a are certified for both copy number and mass [DNA] Mitochondrial DNA (mtDNA) analysis within the forensic community is useful when the nDNA may be degraded or is in low-to-undetectable quantity The challenges of mtDNA testing include contamination of the standard due to routine laboratory use degradation of plasmids or synthetic oligomers or improper commutability of the standard relative to casework samples Commonly cell line materials are used as standards for qPCR with human genomic DNA which may lead to an incorrect mtDNA-to-nDNA ratio (mtDNAnDNA) for unknown samples Non-certified values of mtDNAnDNA for the three components are provided to bridge the current gap between well characterized mitochondrial DNA quantification assays and the availability of commercial standards With the understanding that the forensic DNA community primarily uses qPCR as a method for DNA quantification the performance of the three SRM 2372a materials was evaluated in eleven commercial qPCR chemistries currently used employed within the forensic DNA community

Experimental Methods Production and value-assignment of any certified reference material (CRM) is a complex process In addition to acquiring processing and packaging the materials included in the CRM the materials must be sufficiently homogenous across the units stable over time within its packaging and well-characterized to deliver results that are fit-for-purpose within the intended measurement community20 The process has been particularly complicated for SRM 2372a because we needed to determine the metrological properties of the required ddPCR measurement systems before we could trust the results they provide The following sections describe the methods and processes used to produce and characterize the SRM 2372a materials including sample preparation SRM unit production PCR assays ddPCR measurement processes spectrophotometric absorbance measurements homogeneity testing stability testing ddPCR nDNA and mtDNAnDNA measurements and performance in commercial qPCR assessments Many of these processes were mutually dependent therefore the sections are presented in a ldquodefinitionalrdquo order that sequentially introduces terminology Polymerase Chain Reaction Assays

Ten PCR assays were developed for characterization of SRM 2372a Each assay was confirmed to target one locus per HHGE using the National Center for Biotechnology Informationrsquos BLASTblastn system21 Primers were purchased from Eurofins Operon (Huntsville AL) TaqMan probes labeled with 6-carboxyfluorescein (FAM) were purchased from Thermo Fisher (Waltham MA) FAM-labeled Blackhole Quencher and Blackhole Quencher+ probes were purchased from LGC Biosearch Technologies (Novato CA) Table 1 details the ten nDNA assays used to certify copy number and mass [DNA] of the SRM 2372a components FAM-labeled TaqMan probes were purchased from Thermo Fisher

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

5

and used in monoplex reactions VIC (a proprietary dye)-labeled probes were purchased from the same source and were used when nDNA and mtDNA were analyzed in the same reaction mixture Figure 2 displays exemplar droplet patterns for the assays All assays were developed at NIST and have been optimized for ddPCR18 Table 2 details three mitochondrial DNA (mtDNA) assays used to estimate mtDNAnDNA in the SRM 2372a components Figure 3 displays exemplar droplet patterns for the assays These assays were optimized for ddPCR from real-time quantitative PCR (qPCR) assays from the literature All primers probes were diluted with TE-4 buffer purchased from Affymetrix-USB Products (Affymetrix Inc Cleveland OH) to a nominal 5 micromol working solution for all single-plex assays (individual nDNA and mtDNA assays) For duplex assays (combined nDNA and mtDNA) primer and probe working solutions were diluted to a nominal 10 micromol

Table 1 NIST-Developed Human Nuclear DNA Assays

Assay Target

Chromosome Band Accession Primers and Probe a

Amplicon Length bp

NEIF Gene EIF5B

Chr 2 p111-q111 NC_00000212

F gccaaacttcagccttctcttc R ctctggcaacatttcacactaca PB+ tcatgcagttgtcagaagctg

67

2PR4 Gene RPS27A

Chr 2 p16 NC_00000212

F cgggtttgggttcaggtctt R tgctacaatgaaaacattcagaagtct PB tttgtctaccacttgcaaagctggccttt

97

POTP STR TPOX

Chr 2 p253 NC_00000212

F ccaccttcctctgcttcacttt R acatgggtttttgcctttgg PT caccaactgaaatatg

60

NR4Q Gene DCK

Chr 4 q133-q211 NC_00000412

F tggtgggaatgttcttcagatga R tcgactgagacaggcatatgtt PB+ tgtatgagaaacctgaacgatggt

83

D5 STR D5S2500

Chr 5 q112 NC_00000510

F ttcatacaggcaagcaatgcat R cttaaagggtaaatgtttgcagtaatagat PT ataatatcagggtaaacaggg

75

ND6 STR D6S474

Chr 6 q21-22 NC_00000612

F gcatggctgagtctaaagttcaaag R gcagcctcagggttctcaa PB+ cccagaaccaaggaagatggt

82

D9 STR D9S2157

Chr 9 q342 NC_00000912

F ggctttgctgggtactgctt R ggaccacagcacatcagtcact PT cagggcacatgaat

60

HBB1 Gene HBB

Chr 11 p155 NC_00001110

F gctgagggtttgaagtccaactc R ggtctaagtgatgacagccgtacct PT agccagtgccagaagagccaagga

76

ND14 STR D14S1434

Chr 14 q3213C NC_0000149

F tccaccactgggttctatagttc R ggctgggaagtcccacaatc PB+ tcagactgaatcacaccatcag

109

22C3 Gene PMM1

Chr 22 q132 NC_00002210

F cccctaagaggtctgttgtgttg R aggtctggtggcttctccaat PB caaatcacctgaggtcaaggccagaaca

78

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

6

a F Forward primer R Reverse primer PB Blackhole quencher probe (FAM labeled) PB+ Blackhole Plus quencher probe (FAM labeled) PT TaqMan MGB probe (FAM labeled)

Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays

Table 2 NIST-Optimized Human Mitochondrial DNA Assays

Assay Location Primers and Probe a Amplicon Length bp Ref

mtND1 34853504 F ccctaaaacccgccacatct

69 22 35333553 R gagcgatggtgagagctaaggt 35063522 PT ccatcaccctctacatc

mtBatz 84468473 F aatattaaacacaaactaccacctacct

79 23 85038524 R tggttctcagggtttgttataa 84758491 PT cctcaccaaagcccata

mtKav 1318813308 F ggcatcaaccaaccacaccta

105 24 1336913392 R attgttaaggttgtggatgatgga 1331013325 PT cattcctgcacatctg

a F Forward primer R Reverse primer PT TaqMan MGB probe (FAM labeled)

2PR4

POTP

NEI

F

R4Q

5

D5

ND6

D9

HBB1

ND1

4

22C3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

7

Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays

Droplet Digital Polymerase Chain Reaction Measurements

Nuclear DNA (nDNA) All nDNA ddPCR measurements used in this study followed the same measurement protocol

bull Droplets were generated on the Auto Droplet Generator using the Bio-Rad ldquoddPCR ldquoSupermix for Probes (No dUTP)rdquo (Bio-Rad cat 186-3024 Control 64079080) and ldquoDroplet Generation Oil for Probesrdquo (Bio-Rad cat 186-4110 Control 64046051) Note desoxyuridine 5-triphosphatase (dUTP) is an enzyme sometimes included in PCR reaction mixtures to minimize PCR carryover contamination

bull Table 3 details the composition of the ddPCR mastermix setup for a given assay and sample where ldquomastermix refers to reaction mixture of manufacturerrsquos proprietary PCR reagent ldquoSupermixrdquo forward and reverse PCR primers for a given assay probe for that assay the sample DNA and sometimes additional water to produce the desired value dilution factor F This setup was used for all assays and samples

bull Non-Template Controls (NTCs) for each assay were included in every analysis bull Droplets were thermal cycled on a ProFlex thermal cycler (Applied Biosystems) for

95 degC for 10 min followed by 60 cycles of 94 degC for 05 min and 61 degC for 1 min then 98 degC for 10 min before a 4 degC hold until the plate was removed

bull Droplets were read on the QX200 Droplet Reader and analysis was performed using the QuantaSoft Analysis Software version 1740917

mtND1 mtBatz mtKav

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

8

bull The numbers of positive and negative droplets at the end of 60 cycles were determined and were exported into a spreadsheet for further analysis Assay-specific intensity thresholds were determined by visual inspection for each assay for each measurement session

Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays

Concentration nDNA Mastermix Microliter

per Reaction 2X ldquoSupermix for Probes (no dUTPs)rdquo 125

5 μmolL Forward Primer 1875 5 μmolL Reverse Primer 1875 5 μmolL Probe (FAM) 125

PCR Grade Water 50 14 Diluted Sample 25 Total Volume 250

Mitochondrial DNA (mtDNA) All the mtDNA measurements used in this study followed the nDNA protocol described above but with different mastermix setups The duplex mtDNA measurements used the setup described in Table 4 The monoplex mtDNA measurements used the setup described in Table 5

Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 125

10 μmolL Forward Primer 09375 10 μmolL Reverse Primer 09375 10 μmolL Probe (VIC) 0625

PCR Grade Water 95 nDNA Mastermix from Table 3 05 Total Volume 250

Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 120

10 μmolL Forward Primer 09 10 μmolL Reverse Primer 09 10 μmolL Probe (FAM) 06

PCR Grade Water 72 Diluted Sample (See Table 7) 24

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

9

Total Volume 240 Sample Preparation

The buffy coat white blood cells used to prepare SRM 2372a materials were purchased from Interstate Blood Bank These materials were from four single-source anonymous individuals supplied to us labeled as Female1 Female2 Male1 and Male2 In our laboratories each donor buffy coat bag was aliquoted (5 mL per aliquot) into sterile 50 mL conical tubes and stored at 4 degC prior to extraction The modified salt-out manual extraction protocol described in Appendix A was performed for each aliquot with rehydration of the DNA in TE-4 buffer From 150 mL to 250 mL of solubilized DNA for each component was accumulated through the extraction of the 5 mL aliquots The DNA was stored in perfluoroalkoxy polymer (PFA) containers that had been bleach sterilized thoroughly rinsed with deionized water rinsed with ethanol and air-dried The minimum volume for production of 2000 units was 120 mL Preliminary estimates of the mass [DNA] of the re-solubilized components were obtained using a Nanodrop microvolume spectrophotometer Dilutions of the component materials were made until the absorbance at 260 nm was approximately 1 corresponding to a nominal mass [DNA] of 50 ngμL of dsDNA7 After addition of the TE-4 buffer the containers were placed on an orbital shaker and gently rotated for several hours prior to refrigerated storage at 4 ordmC The four single-source components were initially screened with the nDNA ddPCR measurement process using four nDNA assays 2PR4 NR4Q ND6 and 22C3 Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate The gravimetric 1rarr4 dilution was prepared by weighing an empty PFA vial and recording the weight then pipetting 30 microL TE-4 into the vial and recording the weight then pipetting 10 microL DNA and recording the final weight The dilution factor was calculated in a spreadsheet These screening measurements verified that the individual component materials were fit-for-purpose The four single-source components were evaluated with the nDNA ddPCR measurement process using the ten nDNA assays Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate Quantitative information was derived from the average of the ten nDNA assays to prepare the component C mixture Thirty milliliters of sample Male2 (estimated at 452 ngμL mass [DNA]) were combined with 80 mL of sample Female2 (estimated at 507 ngμL mass [DNA]) and 10 mL of TE-4 buffer to create a 1 male nDNA to 3 female nDNA producing a 14 malehuman mixture After addition of the TE-4 buffer the component C container was gently rotated on the orbital shaker for several hours prior to refrigerated storage at 4 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

10

Absorbance Measurements

In February 2017 absorbance spectra of the recently prepared fully diluted nominally dsDNA and ssDNA solutions of the three components were acquired with a BioCary 100 spectrophotometer using the measurement protocol described in Appendix B In December 2017 additional spectra were acquired for the remaining bulk solution of components A and B and for SRM 2372 components A and B These solutions had been stored at 4 degC in well-sealed PFA containers There was no remaining bulk component C solution SRM Unit Production

In the morning of the day a component was transferred into the sterile component tubes (Sarstedt Biosphere SC micro tube-05 mL Newton NC) used to package the SRM the container holding the diluted component was removed from the refrigerator and placed on a slow orbital shaker for 2 h The tubes were opened and placed under the laminar flow hood in rows of 5 within 80-hole tube racks for a total of 30 tubes per rack Tubes were placed in every second row of the racks to facilitate filling with an Eppendorf Repeater Xstream pipette (Eppendorf North America Inc Hauppauge NY) fitted with a 1-mL positive displacement tip set to dispense 55 microL per tube The component container was removed from the shaker a magnetic stir-bar was added the container was placed on a magnetic stir plate in the laminar flow hood and the material was stirred gently Filling proceeded until 2200 vials were filled with pipet tip replacement after filling every 200 tubes The filling process consisted of two individuals manually filling each vial within a biosafety cabinet (one physically pipetting each sample the other verifying that each tube was filled) and four individuals tightly closing the lids to each tube on a sterilized and covered bench When all tubes in a rack were filled the rack was moved out of the hood the tubes checked for proper filling their lids closed and the tubes transferred to 100-unit pre-labeled storage boxes As the 100-unit boxes were filled they were transferred to the labeling room where SRM labels were applied by hand and colored inserts were added to the lids red for component A white for component B and blue for component C After all tubes for a given component were labeled the 22 boxes were placed in a locked refrigerator Filling of each of the components took 60 min to 90 min All units were transferred to the refrigerator within 4 hours of beginning the filling process All units of the three materials were allowed to equilibrate for a minimum of two weeks prior to beginning homogeneity analysis Component Homogeneity

The ddPCR measurement protocol described above was used to assess the nDNA homogeneity of the three components One tube from the approximate center of each of the 22 storage boxes was assayed using the NEIF and ND6 assays (see Table 1) A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Three technical replicates were obtained for every tube A total

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

11

of six plates were used two plates per component one with the NEIF assay and one with the ND6 assay Figure 4 details the layout of the samples

Figure 4 Plate Layout Design for Homogeneity Measurements

Component Stability

Stability of the materials was evaluated at 4 degC room temperature (RT asymp 22 degC) and 37 degC Tubes from the homogeneity study were divided among the three temperatures for the stability study Additionally two never-opened tubes (one from box 8 and one from box 12 of each component) were held at each temperature Two tubes at each temperature for each component were opened sampled and resealed 4 times over the course of 8 weeks Measurements were made at week 1 week 3 week 5 and week 8 The measurement on week 8 was from an unopened tube (from box 8) for each of the components at all temperatures The data from the homogeneity study was used as timepoint 0 in the stability study anchored to 4 degC by the storage recommendations A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used for both the NEIF and ND6 assays All temperatures components and both assays fit onto one plate for timepoints 1 to 5 Twelve NTCs placed in rows G and H were included for each assay Figure 5 details the sample layout

Figure 5 Plate Layout Design for Stability Measurements

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

ND6 ND6 ND6

NEIFNEIFNEIFComponent A

Component A

Component B

Component B

Component C

Component C

1 2 3 4 5 6 7 8 9 10 11 12A A-7 A-7 B-7 B-7 C-7 C-7 A-7 A-7 B-7 B-7 C-7 C-7B A-10 A-10 B-10 B-10 C-10 C-10 A-10 A-10 B-10 B-10 C-10 C-10C A-9 A-9 B-9 B-9 C-9 C-9 A-9 A-9 B-9 B-9 C-9 C-9D A-12 A-12 B-12 B-12 C-12 C-12 A-12 A-12 B-12 B-12 C-12 C-12E A-8 A-8 B-8 B-8 C-8 C-8 A-8 A-8 B-8 B-8 C-8 C-8F A-11 A-11 B-11 B-11 C-11 C-11 A-11 A-11 B-11 B-11 C-11 C-11G NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTCH NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

NEIF

4C

RT

37C

ND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

12

Component Certification Measurements

After establishing homogeneity and stability measurements intended to be used to certify nDNA values and inform mtDNA values were started using randomly selected SRM tubes from the boxes listed in Table 7 The ten nDNA assays described in Table 1 and two of the three mtDNA assays described in Table 2 were run simultaneously on the same sample plate A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used in each of the six measurement campaigns

Table 6 Tubes used for Certification Measurements

Component Plate A B C

dd20170711 Box 1 Box 1 Box 1 dd20170712a Box 10 tube a Box 10 tube a Box 10 tube a dd20170712b Box 10 tube b Box 10 tube b Box 10 tube b dd20170713 Box 20 Box 20 Box 20 dd20170714 Box 5 Box 5 Box 5 dd20170801 Box 5 Box 5 Box 5

Every sample was analyzed in duplicate with every assay in all six plates Component A was always located in rows B and C component B in rows D and E component C in rows F and G and NTCs in rows A and H Figure 6 displays the basic layout of the ten nDNA and two mtDNA assays In addition to the plate column assignment scheme shown the three assay groupings were also ordered as Group 2 Group 3 Group 1 and Group 3 Group 1 Group 2 to address potential positional bias

Group 1 Group 2 Group 3 2PR4 POTP NEIF NR4Q D5 ND6 D9 HBB1 ND14 22C3 mtND1 mtBatz 1 2 3 4 5 6 7 8 9 10 11 12 A NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Comp A B Comp A C Comp B D Comp B E Comp C F Comp C G

H NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Figure 6 Basic Plate Layout Design for Certification Measurements

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

13

Mitochondrial DNA Measurements

A ldquoduplexed reactionrdquo was used to determine mtDNAnDNA where 1 μL of one of the nDNA mastermix was added to 49 μL of the mtDNA mastermix resulting in a 1rarr50 volumetric dilution of the nDNA The nDNA was targeted with a FAM-labeled probe while the mtDNA was targeted with a VIC labeled probe The primer and probe concentrations were increased from 5 micromol to 10 micromol in the duplexed reactions Additional testing of the mtDNA assays were performed in monoplex with FAM-labeled probes to determine the additional sources of bias between sampling assay setup and dilution factors Table 7 outlines the different methods examined to determine mtDNAnDNA for the three components

Table 7 Run Parameters for Mitochondrial DNA Droplet

Generation nDNA mtDNA

Plate Component Assay Dilution Reps Assay Dilution Reps dd20170824 A manualDG HBB1 1rarr4 3 mtND1 1rarr200 9 dd20170825 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170829 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170831 AB autoDG HBB1 1rarr3 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr6 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr12 2 mtND1 1rarr10rarr100 3 dd20170907 ABC manualDG HBB1 1rarr3 2 mtBatz 1rarr10rarr100 3 dd20170913 AB manualDG NEIF 1rarr4 3 mtND1 mtBatz mtKav 1rarr10rarr100 3 dd20170913 C manualDG NEIF 1rarr4 3 mtBatz mtKav 1rarr10rarr100 3 Performance in Commercial qPCR Chemistries

Table 8 lists the 11 qPCR chemistries that were commercially available in late 2017 All three of the SRM 2372a components were evaluated with these chemistries using component A of SRM 2372 as the reference standard For each component of SRM 2372a four tubes were pulled from box 22 and combined into one tube A 1rarr5 dilution of all samples was prepared followed by three serial 1rarr2 dilutions A five-point serial dilution of each SRM 2372a component was amplified in triplicate for each chemistry This consisted of the neat material the 1rarr5 dilution then the three serial 1rarr2 dilutions A six-point serial dilution of SRM 2372 Component A was also created for the standard curve for the commercial qPCR chemistries This consisted of the neat material followed by a 1rarr5 dilution then followed by four 1rarr2 dilutions serially The six-point serial dilution series of SRM 2372 Component A was also amplified in triplicate These commercial chemistries were amplified in individual plates on an Applied Biosystems 7500 HID Real-Time PCR Instrument (Foster City CA) following the manufacturersrsquo recommended protocols for reaction mix setup and cycling conditions Figure 7 details the plate layout for the qPCR plates Data was analyzed with the HID Real-Time PCR Analysis Software v12 and further analyzed in a spreadsheet

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

14

Table 8 Commercial qPCR Chemistries Tested with SRM 2372a

Manufacturer Commercial Kit Target Codea

Innogenomics

InnoQuant Long IQh Short IQd

InnoQuantHY Long IQHYh Short IQHYd Y IQHYy

Thermo Fisher Quantifiler Duo Human QFDh Male QFDy

Quantifiler HP Large QFHPh Small QFHPd

Quantifiler Human Human QFHh

Quantifiler Trio Large QFTh Small QFTd Y QFTy

Qiagen

Quantiplex Autosomal QPh

Quantiplex Hyres Autosomal QPHh Y QPHy

Quantiplex Pro Human QPPh Degradation QPPd Y QPPy

Promega

Plexor Autosomal Ph Y Py

PowerQuant Autosomal PQh Degradation PQd Y PQy

a Assays are coded with a short acronym derived from the kit name followed by the following codes for the type of target ldquohrdquo autosomal non-degradation described as ldquoAutosomalrdquo ldquoHumanrdquo or ldquoLongrdquo ldquodrdquo autosomal degradation described as ldquoDegradationrdquo or ldquoShortrdquo ldquoyrdquo Y-chromosome described as ldquoMalerdquo or ldquoYrdquo

1 2 3 4 5 6 7 8 9 10 11 12

A

B 2372-A_57 2372-A_57 2372-A_57 2372a-neat_A

2372a-neat_A

2372a-neat_A

2372a-neat_B

2372a-neat_B

2372a-neat_B

2372a-neat_C

2372a-neat_C

2372a-neat_C

C 2372-A_11

4 2372-A_11

4 2372-A_11

4 2372a-15_A

2372a-15_A

2372a-15_A

2372a-15_B

2372a-15_B

2372a-15_B

2372a-15_C

2372a-15_C

2372a-15_C

D 2372-A_5

7 2372-A_5

7 2372-A_5

7 2372a-110_A

2372a-110_A

2372a-110_A

2372a-110_B

2372a-110_B

2372a-110_B

2372a-110_C

2372a-110_C

2372a-110_C

E 2372-A_2

85 2372-A_2

85 2372-A_2

85 2372a-120_A

2372a-120_A

2372a-120_A

2372a-120_B

2372a-120_B

2372a-120_B

2372a-120_C

2372a-120_C

2372a-120_C

F 2372-A_1

425 2372-A_1

425 2372-A_1

425 2372a-140_A

2372a-140_A

2372a-140_A

2372a-140_B

2372a-140_B

2372a-140_B

2372a-140_C

2372a-140_C

2372a-140_C

G 2372-A_0

7125 2372-A_0

7125 2372-A_0

7125

H NTC NTC NTC

Figure 7 Plate Layout Design for Performance Assessments

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

15

Measurement Results After preparing the SRM 2372a bulk solutions their approximate 50 ngmicroL mass [DNA] were confirmed spectrophotometrically The bulk solutions were then packaged into individual tubes The materials contained in the tubes were characterized for homogeneity stability nDNA copy number [DNA] and mtDNAnDNA ratio The following sections detail the results of these measurements Absorbance Spectrophotometry

By widely accepted convention an aqueous DNA solution having an absorbance of 10 at 260 nm corresponds to a mass [DNA] of 50 ngmicroL for dsDNA and (37 to 40) ngmicroL for ssDNA7 Absorbance at (230 270 280 and 330) nm are traditionally used in the assessment of DNA quality725 Due to minor baseline shifts during acquisition of some spectra it was necessary to bias-adjusted all spectra to have absorbance of 00 at 330 nm Table 9 lists the resulting bias-adjusted values for selected wavelengths and the conventional mass [DNA] conversion of the absorbance at 260 nm The close agreement between the mass concentration values for the nominally dsDNA and the converted-to-ssDNA spectra confirm that the bulk solutions were correctly prepared to be asymp 50 ngμL dsDNA The slight increase in absorbance over the 10 months between acquiring the spectra suggests that like SRM 2372 the tertiary structure of the dsDNA in the SRM 2372a materials may be changing slowly with time Figure 8 displays the spectra

Table 9 Absorbance at Selected Wavelengths

Wavelength nm Form Date Component 230 260 270 280 ngμL dsDNA 2142017 A 0394 0931 0758 0488 466

B 0480 1112 0905 0583 556 C 0451 1006 0815 0524 503

ssDNA 2142017 A 0390 0614 0540 0316 454 B 0457 0723 0634 0372 535 C 0411 0641 0563 0330 474

dsDNA 12142017 A 0402 0956 0775 0503 478 B 0484 1128 0914 0589 564

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

16

Figure 8 Absorbance Spectra of SRM 2372a Components A B and C

dsDNA is denatured to ssDNA by diluting with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to DNA mass concentration

Droplet Volume

Droplet volumes for the two lots of ldquoSupermixrdquo used in the studies described in this report were determined using NISTrsquos Special Test Method 11050S-D as described in Dagata et al19 The lot used for most of the ddPCR analyses produced (07349 plusmn 00085) nL droplets where the ldquoplusmnrdquo value is the combined standard uncertainty The lot used for analyses performed after October 11 2017 produced (07376 plusmn 00085) nL droplets NIST measurements have demonstrated that droplet volumes are not influenced by the presence or absence of DNA in the sample solution Measurements repeated over time on an earlier lot of the Supermix product used in these studies suggest that droplet volumes remain constant over the shelf-life of the lot However droplet volumes for different lots of this product as well as for different Supermix products have been observed to differ by several percent We intend to document these measurements in a future report Other researchers have demonstrated that non-constant droplet volume distributions may contribute to ddPCR measurement imprecision and bias2627 However these effects become significant only with distributions wider than observed using Special Test Method 11050S-D19 and at copydroplet values larger than the λ le 05 copydroplet typically used in our measurements Certification Measurements

Homogeneity Results Table 10 lists the λacute nDNA copy number values adjusted for both droplet volume and sample dilution from the homogeneity measurements The data was analyzed using the three-level Gaussian hierarchical ANOVA model described in Appendix C Table 11 lists the values of the within- and between-tube variance components expressed as percent relative standard deviations At the 95 confidence level all components are homogeneous

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372a-A2141721417-NaOH121417

220 260 300 340

Wavelength nm

2372a-B2141721417-NaOH121517

220 260 300 340

Wavelength nm

2372a-C21417

21417-NaOH

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

17

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 NEIF A-1-01 1632 1529 1511 B-1-01 1790 1909 1885 C-1-01 1493 1385 1510 ND6 1616 1640 1582 NAe 1870 1860 1565 1562 1466

NEIF A-1-02 1531 1525 1549 B-1-02 1712 1719 1704 C-1-02 1516 1513 1577 ND6 1615 1589 1601 1775 1718 1756 1490 1547 1570

NEIF A-1-03 1524 1540 1533 B-1-03 1693 1755 1749 C-1-03 1451 1528 1506 ND6 1582 1152 1555 1824 1483 1896 1485 1505 1471

NEIF A-1-04 1454 1458 1427 B-1-04 1754 1745 1739 C-1-04 1440 1486 1520 ND6 1627 1605 1607 1670 1632 1814 1529 1508 1506

NEIF A-1-05 1465 1540 1504 B-1-05 1757 1687 1735 C-1-05 1523 1472 1483 ND6 1607 1633 1586 1683 1759 1703 1513 1520 1491

NEIF A-1-06 1512 1491 1491 B-1-06 1697 1709 1714 C-1-06 1437 1493 1522 ND6 1511 1483 1524 1693 1723 1722 1492 1481 1461

NEIF A-1-07 1418 1480 1455 B-1-07 1722 1625 1703 C-1-07 1485 1477 1463 ND6 1479 1557 1656 1690 1684 1688 1475 1547 1485

NEIF A-1-08 1496 1560 1501 B-1-08 1884 1802 1772 C-1-08 1433 1405 1478 ND6 1514 1615 1540 1864 1887 1821 1497 1501 1495

NEIF A-1-09 1588 1581 1637 B-1-09 1713 1734 1727 C-1-09 1463 1499 1505 ND6 1474 1438 1454 1709 1756 1758 1503 1491 1516

NEIF A-1-10 1574 1513 1570 B-1-10 1759 1809 1738 C-1-10 1534 1458 1454 ND6 1594 1713 1669 1764 1816 1787 1448 1465 1474

NEIF A-1-11 1557 1508 1633 B-1-11 1749 1782 1740 C-1-11 1533 1514 1555 ND6 1571 1611 1631 1741 1792 1784 1488 1470 1483

NEIF A-1-12 1500 1476 1515 B-1-12 1748 1712 1718 C-1-12 1450 1525 1497 ND6 1626 1586 1608 1797 1614 1790 1496 1483 1519

NEIF A-1-13 1672 1634 1804 B-1-13 1632 1668 1701 C-1-13 1537 1527 1515 ND6 1490 1516 1497 1626 1566 1681 1499 1479 1475

NEIF A-1-14 1551 1568 1538 B-1-14 1752 1535 1673 C-1-14 1515 1490 1494 ND6 1617 1614 1639 1724 1741 1685 1469 1480 1466

NEIF A-1-15b 1793 1722 1733 B-1-15 1871 1835 1887 C-1-15 1511 1568 1594 ND6 1818 1753 1762 1770 1752 1798 1515 1543 1565

NEIF A-1-16 1573 1520 1617 B-1-16 1688 1725 1743 C-1-16 1449 1452 1471 ND6 1634 1627 1575 1744 1697 1710 1543 1448 1491

NEIF A-1-17 1516 1516 1628 B-1-17 1789 1769 1825 C-1-17 1388 1421 1394 ND6 1613 1567 1567 1716 1762 1638 1452 NAe 1435

NEIF A-1-18 1612 1535 1672 B-1-18 1826 1780 1849 C-1-18 1475 1455 1445 ND6 1529 1495 1528 1805 1801 1931 1458 1531 1459

NEIF A-1-19 1598 1514 1596 B-1-19 1751 1759 1880 C-1-19 1468 1525 1446 ND6 1571 1535 1511 1827 1766 1786 1435 1478 1421

NEIF A-1-20 1575 1568 1586 B-1-20 1738 1678 1768 C-1-20 1549 1531 1499 ND6 1652 1545 1627 1752 1747 1768 1499 1465 1468

NEIF A-1-21 1658 1611 1578 B-1-21 1742 1836 1713 C-1-21 1530 1440 1484 ND6 1647 1607 1612 1746 1744 1764 1499 1454 NAe

NEIF A-1-22 1547 1528 1532 B-1-22 1761 1743 1660 C-1-22 1442 1487 1436 ND6 1537 1513 1536 1666 1692 1819 1589 1591 1556

NEIF A-1-15c 1456 1448 1523 ND6 1525 1402 1436

NEIF A-2-15d 1679 1599 1674 ND6 1581 1550 1584

a Tube coded as Component-Tube-Box b These results were identified as a probable error in the preparation or volumetric delivery of the sample

solution into the microfluidic device and were not used in the homogeneity evaluations c Repeat assessment of the solution in sample A-1-15 d Results for a second tube from box 15 analyzed at the same time as the repeat assessment of A-1-15 e Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

18

Table 11 Within- and Between-Tube Relative Standard Deviations

Within-Tube Component NEIF ND6 Between-Tube

A 28 40 33 B 27 38 25 C 23 18 19

Stability Results Table 12 lists the λacute volume- and dilution-adjusted nDNA copy number values from the stability measurements Figure 9 summarizes these results The results for the tubes held at 4 degC room temperature (RT asymp 22 degC) and 37 degC are very similar for all three components The absence of between-temperature differences identifies the between-week differences as attributable to variation in the measurement processes rather than sample instability These results indicate that the SRM 2372a genomic copy number is thermally stable from 4 degC to 37 degC over an extended period

Figure 9 ddPCR Stability Measurements as Function of Time

Each dot-and-bar symbol represents the mean λacute and the approximate 95 expanded uncertainty for all results for a given component at a given storage temperature The thick black horizontal lines represent an approximate 95 confidence interval on the λacute values estimated from the homogeneity data

13

14

15

16

17

18

19

0 1 2 3 4 5 6 7 8

λacute H

aplo

id n

DNA

Copi

es μL

Weeks

A

0 1 2 3 4 5 6 7 8

Weeks

B

4 degC 22 degC 37 degC

0 1 2 3 4 5 6 7 8

Weeks

C

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

19

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Week Temp Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3

1 4 degC NEIF A-10 1568 1441 B-10 1666 1653 C-10 1424 1394 ND6 A-10 1474 1494 B-10 1655 1705 C-10 1461 1394

NEIF A-07 1500 1463 B-07 1718 1640 C-07 1446 1452 ND6 A-07 1479 1479 B-07 1708 1686 C-07 1522 1482

RT (22 degC) NEIF A-12 1674 1608 B-12 1805 1777 C-12 1506 1495 ND6 A-12 1522 1507 B-12 1759 1788 C-12 1478 1461 NEIF A-09 1635 1539 B-09 1791 1793 C-09 1411 NAb ND6 A-09 1543 1519 B-09 1661 1682 C-09 1482 1496

37 degC NEIF A-11 1614 1637 B-11 1748 1744 C-11 1515 1464 ND6 A-11 1494 1502 B-11 1740 1697 C-11 1495 1535 NEIF A-08 1592 1614 B-08 1691 1725 C-08 1528 1500 ND6 A-08 1581 1595 B-08 1727 1699 C-08 1458 1483 3 4 degC NEIF A-10 1520 1706 B-10 NAb 1733 C-10 1427 1473 ND6 A-10 1543 1624 B-10 1670 1755 C-10 1443 1364 NEIF A-07 1537 1570 B-07 1549 1565 C-07 1424 1512 ND6 A-07 1517 1479 B-07 1705 1723 C-07 1483 1439

RT (22 degC) NEIF A-12 1521 1607 B-12 1817 1842 C-12 1566 1471 ND6 A-12 1644 1597 B-12 1845 1786 C-12 1504 1511 NEIF A-09 1432 1626 B-09 1825 1856 C-09 1392 1387 ND6 A-09 1539 1528 B-09 1734 1737 C-09 1535 1465

37 degC NEIF A-11 1651 1685 B-11 1759 1779 C-11 1437 1440 ND6 A-11 1680 1594 B-11 1806 1806 C-11 1471 1509 NEIF A-08 1580 1613 B-08 1740 1791 C-08 1582 1590 ND6 A-08 1592 1524 B-08 1769 1694 C-08 1468 1520 5 4 degC NEIF A-10 1486 1440 B-10 1626 1551 C-10 1381 1330 ND6 A-10 1555 1577 B-10 1673 1625 C-10 1547 1395 NEIF A-07 1387 1429 B-07 1723 1787 C-07 1330 1314 ND6 A-07 1538 1522 B-07 1862 1909 C-07 1452 1458

RT (22 degC) NEIF A-12 1601 1631 B-12 1718 1752 C-12 1546 1541 ND6 A-12 1699 1677 B-12 1772 1721 C-12 1490 1471 NEIF A-09 1381 1484 B-09 1675 1694 C-09 1371 1376 ND6 A-09 1625 1615 B-09 1904 1815 C-09 1468 1472

37 degC NEIF A-11 1456 1517 B-11 1720 1704 C-11 1505 1493 ND6 A-11 1650 1658 B-11 1148 1152 C-11 1527 1600 NEIF A-08 1590 1666 B-08 1745 1723 C-08 1469 1540 ND6 A-08 1235 1244 B-08 1809 1749 C-08 1508 1529 8 4 degC NEIF A-08a 1667 1632 1589 B-08a 1880 1884 1940 C-08a 1525 1475 1542 ND6 A-08a 1599 1593 1590 B-08a 1859 1897 1972 C-08a 1550 1502 1520

RT (22 degC) NEIF A-08b 1693 1666 1634 B-08b 1769 1884 1771 C-08b 1595 1557 1545 ND6 A-08b 1644 1620 1647 B-08b 1862 1738 1811 C-08b 1611 1585 1535

37 degC NEIF A-08c 1620 1576 1608 B-08c 1934 1952 1956 C-08c 1595 1600 1569 ND6 A-08c 1576 1553 1532 B-08c 1889 1844 1889 C-08c 1569 1514 1581

a Tube coded as Component-Tube-Box b Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

20

Ten-Assay Certification Results The original intention was to certify the copy number [DNA] of the three components with results from five independently prepared sets of samples analyzed over four days Each sample of the three components was from a new tube from different boxes and was evaluated with 10 nDNA assays The assay positions were alternated across these five ldquo10-assay certificationrdquo plates The λ results for these five plates are systematically somewhat larger than those for the homogeneity and stability data On review we realized that a different workspace and its suite of pipettes had been used to prepare the 1rarr10 dilution of the 1rarr4 diluted components into the reaction mastermix The ldquohighrdquo workstation was used for stability and homogeneity measurements while the ldquolowrdquo workstation was used for certification measurements Gravimetric testing of the pipettes established that the volumes delivered by these alternate pipettes was sufficiently biased to account for the observed λ differences A sixth ldquocertificationrdquo plate was prepared using the original (and now re-verified) pipettes The λ results or this plate agreed well with the homogeneity and stability data Table 13 lists the λacute volume- and dilution-adjusted nDNA copy number values for all six of the ten-assay datasets Figure 10 displays the extent of agreement among the ten assays While there are small-but-consistent differences between the assays there is only a 4 difference between the assays providing smallest and largest λ

Figure 10 Comparison of Assay Performance

The blue dot-and-bars represent the mean and its combined standard uncertainty for the A B and C components for each assay relative to the over-all grand mean for each component The black dot-and-bars represent combination of three component results for each assay The solid horizontal line represents the median difference of the assay means from the grand mean the dotted horizontal lines bound an approximate 95 confidence region about the median

2PR4POTPNEIFR4Q5 D5ND6 D9 HBB1ND14 22C3-10

-8

-6

-4

-2

0

2

4

6

8

10

Rela

tive

Diffe

renc

e fr

om M

ean

Assays

Component MeansAssay MeansMedianU95

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

21

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

7112017 2PR4 1667 1698 1869 1960 1630 1706 POTP 1655 1685 1850 1781 1596 1576 NEIF 1647 1665 1709 1707 1574 1538 R4Q5 1594 1658 1856 1787 1590 1555 D5 1655 1683 1848 1829 1610 1607 ND6 1629 1647 1828 1826 1589 1548 D9 1655 1652 1861 1858 1569 1572 HBB1 1644 1611 1825 1869 1565 1681 ND14 1645 1707 1871 1773 1564 1492 22C3 1545 1603 1776 1902 1548 1655

7122017 2PR4 1668 1739 2174 2231 1499 1497 POTP 1587 1702 2202 2292 1576 1611 NEIF 1678 1614 2172 2216 1545 1560 R4Q5 1628 1762 NA 2078 1584 1546 D5 1738 1649 2230 2192 1577 1529 ND6 1667 1678 2094 2034 1599 1478 D9 1695 1653 1949 2110 1587 1504 HBB1 1757 1750 1971 2123 1520 1551 ND14 1771 1764 2123 2066 1479 1438 22C3 1762 1759 2107 2046 1512 1494

7122017 2PR4 1638 1667 1810 1806 1452 1424 POTP 1676 1638 1865 1801 1347 1397 NEIF 1616 1624 1806 1786 1395 1350 R4Q5 1559 1587 1878 1768 1442 1459 D5 1608 1652 1824 1790 1448 1602 ND6 1611 1589 1768 1670 1463 1435 D9 1517 1607 1811 1749 1402 1433 HBB1 1557 1582 1763 1741 1474 1481 ND14 1604 1625 1865 1781 1466 1482 22C3 1675 1616 1815 1866 1439 1568

7132017 2PR4 1888 1939 1784 1821 1470 1444 POTP 1868 1815 1735 1813 1411 1427 NEIF 1953 1902 1778 1813 1512 1439 R4Q5 1779 1871 1835 1749 1452 1468 D5 1924 2124 1794 1848 1567 1524 ND6 1861 1787 1700 1711 1480 1424 D9 1824 1844 1799 1779 1442 1459 HBB1 1809 1799 1765 1769 1430 1467 ND14 1864 1857 1754 1725 1413 1411 22C3 1930 1942 1792 1814 1486 1480

7142017 2PR4 1597 1590 1718 1768 1485 1458 POTP 1533 1404 1753 1698 1422 1459 NEIF 1462 1533 1773 1792 1438 1465 R4Q5 1493 1519 1777 1740 1439 1468 D5 1538 1586 1803 1795 1515 1457 ND6 1492 1516 1718 1693 1409 1412 D9 1528 1538 1802 1743 1469 1474 HBB1 1548 1581 1834 1745 1453 1428 ND14 1513 1514 1731 1714 1465 1425 22C3 1562 NA 1842 NA 1363 1393

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

22

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

101217 2PR4 1431 1411 1791 1770 1357 1364 POTP 1407 1395 1746 1820 1447 1361 NEIF 1487 1398 1803 1803 1448 1404 R4Q5 1450 1439 1767 1739 1409 1411 D5 1394 1422 1745 1773 1389 1429 ND6 1396 1393 1824 1729 1415 1417 D9 1414 1390 1752 1744 1431 1380 HBB1 1401 1382 1813 1754 1381 1336 ND14 1412 1460 1648 1671 1403 1387 22C3 1435 1410 1818 1741 1435 1372

Estimation of Copy Number Concentration It was decided that all three sets of results (homogeneity stability and certification) would be used to derive the certified values with the λ of the first five ldquoten-assayrdquo plates adjusted by a data-driven estimate of the volumetric bias This bias was incorporated into the statistical model providing better estimates of the λ means while capturing all of the variance components (homogeneity stability between-assay and between data sets) as random effects using Bayesian MCMC programmed in OpenBUGS28 Table 14 lists the estimated values and their uncertainties See Appendix C for details

Table 14 Measured Copy Number Concentrations

Genomic Copies per Nanoliter of Solution Component Mean a u(Mean) b U95(Mean) c CV95

d A 151 07 14 93 B 175 03 06 34 C 145 05 10 69

a Mean the experimentally determined most probable estimate of the true value b u(Mean) the standard uncertainty of the Mean c Mean plusmnU95(Mean) the interval which is believed with about 95 confidence to contain the true value d CV95 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the counting unit one29 2) the validity of the Poisson endpoint transformation for digital PCR endpoint assays when applied to samples providing le 05 copies per droplet 3) the determination that the copies are dispersed as dsDNA and 4) calibrated mean droplet volume measurements made at NIST during sample dilution and mastermix preparation Conversion of Copy Number to Mass Concentration As described in Reference 18 metrologically valid conversion from copies per nanoliter to nanograms per microliter requires quantitative estimates for the number of nucleotide bases per reference human genome and the average nucleotide mass Using the values estimated in that report and asserting that each target copy corresponds to one HHGE the required transformation is

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

23

[DNA] ngmicroL

= [DNA] HHGE

nLtimes

3301 ngHHGE

The combined standard uncertainty is

119906119906 [DNA] ng

microL = 1199061199062 [DNA] HHGE

nL + 0531100

times[DNA] HHGE

nL 2

Both the measured and the conversion factor uncertainties are associated with ldquolargerdquo degrees of freedom Therefor the approximate 95 confidence uncertainty is estimated

11988011988095 [DNA] ng

microL = 2 times 119906119906

[DNA] ngmicroL

Table 15 lists the transformation of the Table 14 copy number values These values agree well with the conventional 260 nm absorbance values averaging (4 plusmn 5) larger

Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution

ngmicroL Component Value a u(Measurement)b u(Conversion)c U95(Value)d CV95

e A 498 23 03 47 93 B 578 10 03 21 36 C 479 17 03 33 70

a Estimated true value b Standard uncertainty component from the ddPCR measurements c Standard uncertainty estimated from the relative uncertainty of the conversion factor d Value plusmnU95(Value) is believed with about 95 confidence to contain the true value e 100U95(Value)Value the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the measured estimates of copies per nanoliter and 2) determinations made by us from internationally recognized literature resources of the average number of nucleotide bases per HHGE and the mean molecular weights of the bases18 Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)

While not intended to provide certified values mtDNAnDNA was measured as part of 13 experiments over 5 months Measurements were made using dilution-adjusted λ for the mtND1 mtBatz andor mtKav mitochondrial assays referenced against the dilution-adjusted λ of either the NEIF or HBB1 nDNA assays By chance both DNAs in the component C mixture have a binding site mutation that silences the mtND1 probe Figure 11 graphically summarizes the results by assay and component Table 16 lists the data Table 17 summarizes results from a random effects model that accounts for between-assay variability See Appendix C for details

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

24

Due to the limited number of mtDNA assays used in these measurements these results are metrologically traceable only to the assays used

Figure 11 mtDNAnDNA for Components A B and C

Each dot-and-bar symbol represents a mean mtDNAnDNA and its approximate 95 expanded uncertainty The dotted horizontal lines represent approximate 95 confidence intervals on the ratios

Table 16 mtDNAnDNA Ratios

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

20170711a 1 mtND1 1931 1870 2461 2401 1 mtBatz 1845 1764 2555 2608 3201 3299

20170711b 1 mtND1 1810 1906 2229 2200 1 mtBatz 1785 1837 2108 2192 3347 3287

20170712 1 mtND1 1743 1761 2127 2198 1 mtBatz 1829 1847 2455 2506 3217 3384

20170713 1 mtND1 1716 1800 2211 2199 1 mtBatz 1740 1813 2380 2249 3200 3363

20170714 1 mtND1 1923 1835 2105 1943 1 mtBatz 1935 1851 2136 2119 3294 3013

20170814 1 mtND1 1780 1776 1706 2046 2057 2024 1 mtBatz 1804 1751 1738 1998 2043 1952 2927 2776 2762 1 mtKav 1769 1786 1761 2077 2117 2041 3044 2830 2825

20170824 1 mtND1 1947 1928 1922 2 mtND1 1829 1895 1851 3 mtND1 1674 1617

20170825 1 mtBatz 1667 1813 1708 1598 1928 1916 1915 2832 2840 2788 2725 2 mtBatz 1566 1687 1663 1589 1947 1946 1961 1909 2835 2790 2736 2679 3 mtBatz 1643 1668 1604 1573 1953 1952 1941 1881 2803 2773 2757 2752

20170829 1 mtBatz 1719 1716 1670 1666 1865 1863 1828 1828 2599 2508 2551 2609 2 mtBatz 1581 1573 1601 1612 1944 1964 1947 1895 2617 2483 2626 2602 3 mtBatz 1546 1542 1538 1587 1908 1909 1940 1839 2665 2535 2570 2650

20170831 1 mtND1 1617 1599 1688 1783

ND1

ND1

ND1Ka

v

Kav

Kav

Batz

Batz

Batz

A

B

C

160

210

260

mtD

NAn

DNA

Ratio

Assay

Bind

ing s

ite m

utat

ion

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

25

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

2 mtND1 1622 1579 1823 1631 3 mtND1 1575 1528 1920 1887

20170907 1 mtBatz 1584 1544 1890 1886 2501 2522 20170913 1 mtND1 1757 1724 1768 1990 1921 1982

1 mtBatz 1743 1671 1766 1990 1924 1958 2805 2825 2829 1 mtKav 1692 1662 1767 1903 1831 1919 2709 2738 2727

20171016 1 mtND1 1924 1944 1893 2304 2341 2332 1 mtBatz 1786 1752 1743 2048 2078 2067 2797 2642 2792 1 mtKav 1530 1542 1523 2399 2479 2493 2804 2789 2842 2 mtBatz 1580 1557 1592 2715 2606 2687 2 mtKav 2103 2100 2049 2799 2628 2559 3 mtND1 1749 1769 1777 2065 2166 2164 3 mtBatz 1965 1889 1947 2314 2280 2258 2639 2535 2657 3 mtKav 1813 1864 1864 1867 1987 1971 2953 2757 2802

Table 17 mtDNAnDNA Summary Results

mtDNAnDNA Component Value a u(Value)b U95(Value)c CV95

d A 1737 16 38 17 B 2058 22 44 21 C 2791 23 47 17

a Estimated true value b Standard uncertainty c Value plusmnU95(Value) is believed with about 95 confidence to contain the true value d 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) Performance in Commercial qPCR Chemistries

Traditional qPCR assays only quantify relative to a reference standard Thus qPCR assays do not provide absolute [DNA] for the individual materials but they can provide [DNA] of one component relative to another Figure 12 summarizes the BA and CA ratios for the commercial autosomal (Human) targets from the eleven assays evaluated (see Table 8) The Y-chromosome (Male) assays cannot be summarized in this manner as the (all female) component B does not include a Y chromosome Figure 13 summaries the mass [DNA] results for the (all male) component A and (male-female mixture) component C Since this material was prepared as a nominal 13 malefemale mixture the expected malehuman (total) ratio is 14 sjc As displayed in these figures there is intrinsic variability between the SRM 2372a components and between the commercial qPCR chemistries examined This reflects the difference in the targets and amplification chemistries for each of the commercial kits Many of the commercial chemistries on the market to date contain multi-copy targets for both autosomal and Y-chromosomal quantification to increase the sensitivity of the assays

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

26

Figure 12 qPCR Autosomal Quantitation Performance Summary

Open circles denote the BA CA ratios the 17 commercial qPCR assays evaluated each labeled with the code listed in Table 8 The solid black circle represents ratios as estimated by ddPCR The red lines bound the approximate 95 confidence intervals on the ddPCR-estimated ratios

Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary

Open squares denote the ngμL values for SRM 2372a Components A (male) and C (mixture of male and female) for the seven commercial Y-chromosome qPCR assays evaluated each labeled with the code listed in Table 8 The dotted red lines represent malehuman ratios of 13 14 and 15

Appendix D provides a graphical analysis of all the data generated in the performance assessments for each of the targets amplified in the eleven commercial qPCR chemistries and the [DNA] values for all of the assays

dd

IQ_h

IQ_d

IQHY_h

IQHY_d

QFD_h

QFH_h QFHP_h

QFHP_dQFT_h

QFT_d

QPH_h

QPP_hQPP_d

P_h

PQ_h

PQ_d

06

08

10

12

14

07 09 11 13

CA

BA

IQHY_y

QFD_y

QFT_y

QPH_y

QPP_y

P_y

PQ_y

1413

15

10

12

14

16

18

30 40 50 60 70

[DNA

] C n

gμL

[DNA]A nguL

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

27

Qualification of ddPCR as Fit-For-Purpose We have documented the requirements for establishing metrological traceability of ddPCR human genomic assay results expressed as copies per nanoliter of sample in a series of three peer-reviewed papers151617 Through the experimental evidence they are based upon these papers demonstrate that ddPCR can accurately enumerate the number of independently migrating genomic copies in an aqueous extract Our most recent paper establishes the metrological traceability of these ddPCR results when transformed into the nanogram per microliter units used by the forensic and clinical communities18 However our previous studies have not addressed a fundamental requirement for correctly interpreting digital PCR measurements In broad simplification end-point assay systems count the number of independently partitioning entities regardless of the number of copies the entities contain With assays that target just one site per HHGE entities that migrate as ssDNA will yield counts that are twice the value had they migrated as dsDNA30 Should some originally dsDNA be converted over time to ssDNA ddPCR measurements of the number of independently migrating entities would increase although the number of HHGE would be unchanged Measurements of mass [DNA] provided by absorbance spectrophotometry and ddPCR reflect different properties absorbance at 260 nm is sensitive to the spatial configuration of the nucleotides as well as their number while ddPCR counts independently migrating entities The SRM 2372a materials have absorbance spectra shown in Figure 8 consistent with being predominantly dsDNA The SRM 2372 materials now have the absorbance spectra shown in Figure 14 which are consistent with them being entirely ssDNA

Figure 14 Absorbance Spectra of SRM 2372 Components A and B dsDNA is denatured to ssDNA by diluting the source material with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to mass [DNA]

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372-A200720122012-NaOH121417

0

4

6

0

4

6

220 260 300 340

Wavelength nm

2372-B200720122012-NaOH121517

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

28

To ensure the validity of ddPCR-based value assignments it is thus necessary to evaluate the nature of the partitioning entities Should a solution of what is thought to be dsDNA contain an appreciable fraction of ssDNA then the ddPCR results would require adjustment31 To avoid further delay in making SRM 2372a available to the forensic community the following sections detail a series of experiments that have enabled us to conclude with high confidence that the SRM 2372a components have a dsDNA conformation and that for this material ldquoentitiesrdquo can be considered dsDNA ldquocopiesrdquo The remaining stock of the SRM 2372 components also have a dsDNA conformation however the strands in the old materials are much less robustly connected than they are in the new materials Digital Polymerase Chain Reaction Assays

Three nDNA assays were used HBB1 NEIF and ND6 The HBB1 and NEIF assays reproducibly yield very similar results for DNA extracts from many different donors The HBB1 target is at the tip of the short (p) arm of chromosome 11 The NEIF target is in the centromere region of chromosome 2 The ND6 assay was selected for use because it typically yields the lowest results of the ten nDNA assays used to certify the SRM 2372a components (see Figure 10 below) The ND6 target is about halfway down the long (q) arm of chromosome 6 Heat Treatments (Melting of the DNA into Single Strands)

In all experiments the non-heated control and the heat-treated diluted solutions were stored in the same type of PCR reaction tube The starting point for the heat treatment was to determine if there was a difference between snap-cooling on ice and slower cooling at 25 degC Two tubes of the diluted material were placed in a GeneAmp 9700 thermal cycler (Applied Biosystems) and cycled up to 98 degC and held for 15 min After 15 min one tube was removed and placed on ice while the remaining tube was cooled at 1 degCs in the thermal cycler to 25 degC then placed on ice until prepared for ddPCR Additional heat treatments included cycling samples on a GeneAmp 9700 thermal cycler

a) from 98 degC for 5 min then cooling to 25 degC for 2 min During the 25 degC hold a randomly selected tube was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 45 min at 98 degC

b) from 94 degC for 1 min then cooling to 25 degC for 1 min During the 1 min hold at 25 degC a random sample was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 9 min at 94 degC

c) from 90 degC for 1 min then cooling to 25 degC for 1 min for a total of 20 cycles Samples were pulled at 25 degC after 1 min 3 min 5 min 10 min and 20 min and placed on ice

After the above experiments heat treatments were performed using a ProFlex thermal cycler This thermal cycler has six temperature zones that can be le 5 degC apart This enables heating

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

29

tubes at the same time but at different temperatures After treatment tubes were cooled at 5 degCs to 4 degC in the thermal cycler While designed for use with PCR plates a MicroAmp 96-well tray for the VeriFlex block enabled its use with single PCR reaction tubes Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a

As a first step in evaluating whether the DNA in the SRM materials disperse as dsDNA or ssDNA a direct comparison of the five materials was made using the HBB1 and NEIF assays Table 18 reports summary statistics for the ddPCR measurements reported as λ copiesdroplet and the estimated mass [DNA] values from the ddPCR results18 from the conventional dsDNA relationship7 and from the conventional ssDNA relationship after denaturing with 04 molL NaOH89 While the mass [DNA] values differ somewhat between the three methods there is no evidence for the SRM materials segregating exclusively or even mostly as ssDNA The extent of the hypochromic interactions among the nucleotides does not accurately reflect the extent to which the two strands of dsDNA have separated

Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a

λ copies per droplet λ copies per droplet ngmicroL Sample Assay Nrep

a Mean CVb ncmbc Mean CVb ddPCRd dsDNAe ssDNAf

2372-A HBB1 4 03645 19 8 03644 20 566g 524 57 2372-A NEIF 4 03644 23 2372-B HBB1 4 03513 23 8 03487 35 580h 536 61 2372-B NEIF 4 03461 47 2372a-A HBB1 4 02603 27 8 02649 30 474 465 454 2372a-A NEIF 4 02694 24 2372a-B HBB1 4 02908 34 8 02891 37 517 556 534 2372a-B NEIF 4 02873 45 2372a-C HBB1 4 02757 28 8 02711 28 485 502 474 2372a-C NEIF 4 02665 16

a Number of technical replicates for the given assay b CV = 100(standard deviation)mean c Combined number of technical replicates ignoring potential differences between the assays d Estimated from the ddPCR λ measurements listed in this Table18 e Estimated from the conventional relationship between the mass [DNA] of dsDNA and absorbance at

260 nm7 using absorbance spectra acquired shortly after solutions were produced f Estimated from the relationship between the mass [DNA] of ssDNA and absorbance at 260 nm after 11

dilution with 04 molL NaOH89 g Values adjusted to measured (132 plusmn19) volume loss of archived units due to evaporation h Values adjusted to measured (71 plusmn25) volume loss of archived due to evaporation

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

30

Quantitative Evaluation of ssDNA Content

Absorbance spectrophotometry is routinely used to quantitatively assess DNA tertiary structure3233 On storage ldquoat physiological pH the intrastrand repulsion between the negatively charged phosphate groups forces the double helix into more rigid rodlike conformationrdquo34 suggesting a mechanism for the observed slow changes in the SRM 2372 materials As demonstrated in Table 18 absorbance at 260 nm does not necessarily correlate with ddPCR behavior Yet ldquofurthermore the repulsion between phosphate groups on opposite strands tends to separate the complementary strandsrdquo34 Therefore a method that is independent of hypochromism is needed to adequately evaluate the proportion (if any) of ldquotruerdquo (completely disjoint) ssDNA in the SRM materials Circular dichroism spectroscopy is sensitive to the conformational structures of DNA in solution35 and different DNA configurations migrate differentially in agarose gel electrophoresis7 Both techniques have been used to qualitatively discriminate ssDNA from dsDNA in ldquopurerdquo single-sequence materials but neither technique appears to be able to quantify small proportions of ssDNA in a complex multi-chromosome genomic material Wilson and Ellison have proposed a real-time chamber digital PCR (cdPCR) technique that can estimate excess proportions of ssDNA copies from the crossing-threshold distribution31 However their method is insufficiently sensitive for low ssDNA proportions It is well known that heating DNA beyond the melting temperature separates mammalian dsDNA into ssDNA36 albeit we have shown using cdPCR that boiling dsDNA for some minutes also renders some target sites non-amplifiable15 In the following section we document a ddPCR-based approach for estimating the proportion if any of ssDNA in a sample that involves less destructive heating While the method is not yet optimized the results from these preliminary studies demonstrate that the SRM 2372 and 2372a components disperse into the ddPCR droplets as dsDNA ddPCR Measurement of ssDNA Proportion Given two otherwise identical aliquots of a given DNA material one subjected to some treatment and the other untreated the ratio between the average number of copies per droplet of treated material and that of its untreated sibling λtλu quantitatively estimates the change in the number of independently dispersing entities Should the treatment reduce the number of accessible amplifiable targets (AAT) λtλu will be less than 1 Reduction in AAT may result from damage to the target itself andor conformational changes to the fragment containing the target that render the target inaccessible or non-amplifiable16 Assuming no denaturation of dsDNA to ssDNA either because of an ineffective treatment or the material having started as ldquopurerdquo ssDNA if the treatment does not itself damage targets then λtλu will equal 1 Should treatment convert at least some dsDNA to ssDNA without damaging targets the number of AAT will increase and λtλu will be greater than 1 Assuming complete conversion of ldquopurerdquo dsDNA to ldquopurerdquo ssDNA without other damage then λtλu will equal 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

31

Values greater than 1 but less than 2 indicate some conversion of dsDNA to ssDNA but also some combination of (a) incomplete denaturation of dsDNA to ssDNA (b) partial renaturation of ssDNA to dsDNA after treatment but before ddPCR analysis (c) damage to targets during the process or ndash of greatest interest ndash (d) the original sample material was an impure mixture of dsDNA and ssDNA The following discussions graphically summarize sets of λtλu followed by tabular listings of the relevant experimental conditions ddPCR measurements and the numerical λtλu values The information in these tables includes

Plate Index Experiment identifier indicates date performed in YYYYMMDD format Material Sample material SRM 2372-A 2372-B 2372a-A 2372a-B or 2372a-C Assay ddPCR assay HBB1 NEIF or ND6 degC Denaturation temperature in degC Min Length of time the treated material was held at denaturation temperature in

minutes Cool Method used to cool treated material from denaturation temperature Snap

Slow R1 or R5 where ldquoSnaprdquo indicates cooling on ice ldquoSlowrdquo cooling at room temperature to 25 degC ldquoR1rdquo cooling in a thermal cycler to 25 degC at 1 degCs and ldquoR5rdquo cooling in a thermal cycler to 4 degC at 5 degCs

F Dilution factor expressed as (volume Material) per (final volume of diluted solution) 12 14 or 18

n Number of technical replicates Mean Mean of technical replicates CV Coefficient of Variation expressed as a percentage 100 Standard Deviation

Mean

λtλu the copies per droplet result of the treated solution λt divided by the copies per droplet result of its untreated sibling λu

u(λtλu) Standard uncertainty of λtλu 119906119906(120582120582t 120582120582ufrasl ) = 120582120582t 120582120582ufrasl CV(120582120582t)100

2

+ CV(120582120582u)100

2

Proof of Principle Figure 15 summarizes two replicate sets of measurements performed to assess whether the SRM 2372 and 2372a materials respond similarly whether the denatured materials renature if allowed to cool slowly and whether results are (reasonably) repeatable Table 19 details the experimental conditions The λtλu values for the four materials are similar all much greater than 1 but not approaching 2 Most of the ratios are on or close to the ldquoslow cooling is the same as snap coolingrdquo line suggesting that renaturation is not a significant concern While the replicate results for the SRM 2372-A and 2372a-A materials do not overlap as closely as those for SRM 2372-B and 2372a-B similar treatment conditions produce similar results

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

32

Figure 15 Comparison of Snap-Cooling to Slow-Cooling

Symbols mark the means of replicate measurements made using the same dilutions on successive days with bars representing plusmn one standard combined uncertainty Results are labeled by material ldquoArdquo and ldquoBrdquo for SRM 2372 components and ldquoaArdquo ldquoaBrdquo and ldquoaCrdquo for SRM 2372a components The diagonal black line represents equality between the snap- and slow-cooling treatments

Table 19 Comparison of Snap-Cooling to Slow-Cooling

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171218 2372-A HBB1 14 4 03830 29

dd20171218 2372-A HBB1 98 15 Slow 14 2 05644 18 1474 0051 dd20171218 2372-A HBB1 98 15 Snap 14 4 05852 15 1528 0051

dd20171218 2372a-A HBB1 14 4 03029 18

dd20171218 2372a-A HBB1 98 15 Slow 14 4 04430 21 1463 0040 dd20171218 2372a-A HBB1 98 15 Snap 14 4 04454 33 1471 0055

dd20171218 2372-B HBB1 14 4 03594 31

dd20171218 2372-B HBB1 98 15 Slow 14 4 05806 18 1615 0058 dd20171218 2372-B HBB1 98 15 Snap 14 4 05825 13 1620 0055

dd20171218 2372a-B HBB1 14 4 03383 15

dd20171218 2372a-B HBB1 98 15 Slow 14 4 05043 38 1491 0061 dd20171218 2372a-B HBB1 98 15 Snap 14 4 05254 20 1553 0039

dd20171219 2372-A HBB1 14 4 03357 23

dd20171219 2372-A HBB1 98 15 Slow 14 4 05340 16 1591 0046 dd20171219 2372-A HBB1 98 15 Snap 14 4 05340 12 1591 0042

A

aA

B

aB

A

aA

B

aB

14

15

16

17

14 15 16 17

λ tλ

u Sn

ap C

oole

d

λtλu Slow Cooled

Rep 1Rep 2Snap = Slow

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

33

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171219 2372a-A HBB1 14 4 02735 24

dd20171219 2372a-A HBB1 98 15 Slow 14 4 04151 03 1518 0037 dd20171219 2372a-A HBB1 98 15 Snap 14 4 03918 21 1433 0045 dd20171219 2372-B HBB1 14 4 03346 22

dd20171219 2372-B HBB1 98 15 Slow 14 4 05476 17 1637 0045 dd20171219 2372-B HBB1 98 15 Snap 14 4 05480 29 1638 0059 dd20171219 2372a-B HBB1 14 4 03158 19

dd20171219 2372a-B HBB1 98 15 Slow 14 4 04798 16 1519 0038 dd20171219 2372a-B HBB1 98 15 Snap 14 4 04856 17 1538 0039

Time at Denaturation Temperature Figure 16 summarizes λtλu for the SRM 2372-B material as a function of denaturation temperature and the cumulative time the solution is held at that temperature Table 20 details the experimental conditions

Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures

The symbols represent the ratio of the mean result of three technical replicates of the untreated and each of the treated solutions the bars represent combined standard uncertainties The lines denote regression fits to the function λtλu = atimese-bt where t is the cumulative time Values and the standard uncertainties of the coefficients are listed in the legend

12

14

16

18

20

0 5 10 15 20

λ t λ

u

Cummulative Treatment Time min

degC a u (a ) b u (b )90 1804 0039 00198 0000494 1788 0029 00099 0002998 1855 0019 00067 00019

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

34

For all three of the denaturation temperatures studied the relationship between λtλu and cumulative time t can be modeled as

λtλu = ae-bt

where a is the extrapolated value of λtλu at t = 0 and b is the rate of change in AAT per minute As expected the rate of loss due to the treatment increases with increasing treatment temperature and cumulative time at that temperature37 Given the much longer treatment time for the 98 degC series the agreement among the three t = 0 parameter (a) values is remarkable This suggests that the first 1 min of treatment efficiently separates dsDNA into ssDNA The pooled relative standard uncertainty of these regression parameter is 17 somewhat smaller than the 25 pooled relative standard uncertainties of the individual λtλu However the plusmnstandard uncertainty interval of the 1-min λtλu results for the 94 degC and 90 degC both overlap the extrapolated t = 0 value suggesting that reliable results can be obtained using a single treatment of short duration

Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171228 2372-B HBB1 14 3 03650 17 dd20171228 2372-B HBB1 98 5 R1 14 3 06157 12 1687 0035 dd20171228 2372-B HBB1 98 10 R1 14 3 05493 08 1505 0028 dd20171228 2372-B HBB1 98 15 R1 14 3 05019 13 1375 0029 dd20171228 2372-B HBB1 98 20 R1 14 3 04583 21 1256 0034 dd20171228 2372-B HBB1 98 25 R1 14 3 04154 32 1138 0041 dd20171228 2372-B HBB1 98 30 R1 14 3 03784 14 1037 0023 dd20171228 2372-B HBB1 98 35 R1 14 3 03408 10 0934 0019 dd20171228 2372-B HBB1 98 40 R1 14 3 02980 01 0816 0014 dd20171228 2372-B HBB1 98 40 R1 14 3 02815 08 0771 0015

dd20171229 2372-B HBB1 14 3 03708 27

dd20171229 2372-B HBB1 94 1 R1 14 3 06598 26 1779 0056 dd20171229 2372-B HBB1 94 2 R1 14 3 06553 32 1767 0063 dd20171229 2372-B HBB1 94 3 R1 14 3 06301 20 1699 0045 dd20171229 2372-B HBB1 94 4 R1 14 3 06252 14 1686 0037 dd20171229 2372-B HBB1 94 5 R1 14 3 06577 14 1773 0039 dd20171229 2372-B HBB1 94 6 R1 14 3 06156 07 1660 0030 dd20171229 2372-B HBB1 94 7 R1 14 3 06118 10 1650 0033 dd20171229 2372-B HBB1 94 8 R1 14 3 06256 12 1687 0035 dd20171229 2372-B HBB1 94 9 R1 14 3 05999 10 1618 0032

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

35

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180104 2372-B HBB1 14 4 03713 25 dd20180104 2372-B HBB1 90 1 R1 14 3 06916 22 1863 0052 dd20180104 2372-B HBB1 90 3 R1 14 3 06465 06 1741 0032 dd20180104 2372-B HBB1 90 5 R1 14 3 06431 20 1732 0045 dd20180104 2372-B HBB1 90 10 R1 14 3 06014 25 1620 0049 dd20180104 2372-B HBB1 90 15 R1 14 3 06024 10 1623 0032 dd20180104 2372-B HBB1 90 20 R1 14 3 06003 14 1617 0035

Temperature Figure 17 summarizes the effect of denaturation temperature using a single treatment of 05 min on the 2372-B and 2372a-B materials Table 21 details the experimental conditions

Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

The symbols represent the ratio of the mean result of four technical replicates of the untreated and three replicates of the treated solutions the bars represent combined standard uncertainties In all cases the duration of the treatment was 05 min The lines denote regression fits to the linear function λtλu = a +bT where Tis the denaturation temperature

The SRM 2372 and 2372a materials respond very differently to changes in denaturation temperature The older spectrophotometrically ssDNA material suffers somewhat greater loss in λtλu as temperature increases The new spectrophotometrically dsDNA material suffers much greater loss in λtλu as temperature decreases

14

15

16

17

18

19

20

82 86 90 94 98

λ tλ

u

Temperature degC

2372-B

2372a-B

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

36

Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180109 2372-B HBB1 14 4 03577 25 dd20180109 2372-B HBB1 100 05 R5 14 3 06314 16 1765 0051

dd20180104a 2372-B HBB1 14 4 03577 05

dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032 dd20180104a 2372-B HBB1 94 05 R5 14 3 06459 25 1806 0046 dd20180104a 2372-B HBB1 90 05 R5 14 2 06591 08 1842 0017 dd20180104a 2372-B HBB1 88 05 R5 14 3 06589 05 1842 0013 dd20180104a 2372-B HBB1 84 05 R5 14 3 06729 10 1881 0022 dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

We hypothesize that in the older SRM 2372 material the between-strand bond is relatively fragile and can be completely severed at relatively low temperature while avoiding much temperature-dependent AAT loss In contrast the between-strand bond in the SRM 2372a material requires higher temperature to completely disassociate the strands This suggests that there is not a sample-independent ldquooptimumrdquo combination of denaturation temperature and treatment duration However 98 degC for 05 min appears to provide comparable results for these two materials DNA Concentration in the ddPCR Solution Figure 18 summarizes the effect of [DNA] in the ddPCR solution with SRM 2372-B at a denaturation temperature of 98 degC and treatment duration of 05 min Table 22 details the experimental conditions

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

37

Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution

The solid black symbols represent the ratio of the mean result of four technical replicates of the untreated and treated solutions with the λtλu values shown on the left-hand vertical axis the bars represent combined standard uncertainties The solid blue circles represent λu and the open red circles λt with values shown on the right-hand vertical axis at this graphical scale the standard deviations of the technical replicates are covered by the symbol The diagonal blue and red lines represent the log-linear relationships between just the 18 and 14 λ values

The primary effect of [DNA] appears to be the loss of linearity in λt with copiesdroplet values greater than about 1 copiesdroplet There may be some advantage to use somewhat greater dilutions than the routine 14 however the small increase in λtλu may be offset by the increased variability with small λu This suggests the use of dilutions that give λu values of (01 to 04) copiesdroplet

Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180117 2372-B HBB1 12 4 07127 11 dd20180117 2372-B HBB1 98 05 R5 12 4 12331 14 1730 0031

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 3 06151 15 1843 0040

dd20180117 2372-B HBB1 18 4 01549 18 dd20180117 2372-B HBB1 98 05 R5 18 4 02897 20 1870 0051

12 14 18

170

175

180

185

190

02

04

06

08

101214

λ T

arge

t D

ropl

et

λ tλ

u

Dilution Factor

λ₊λᵤλᵤλ₊

λtλu

λu

λt

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

38

Equivalence of Assays Figure 19 summarizes the agreement between the HBB1 NEIF and ND6 assays at a denaturation temperature of 98 degC and treatment duration of 05 min Table 23 details the experimental conditions

Figure 19 Comparison of Assays

The symbols compared the λtλu results from the HBB1 and either the NEIF or ND6 assays Each λtλu value is the ratio of the mean result of four technical replicates of an untreated and three or four replicates of the treated solutions The bars represent combined standard uncertainties The diagonal line represents equality between the assays

There is little or no difference between the λtλu results from the HBB1 NEIF and ND6 assays suggesting that the chromosomal location of the assay target has little influence Since the NEIFHBB1 comparison includes results for both the SRM 2372 and 2372a materials this equivalence is not material-specific

14

16

18

20

14 16 18 20

λ tλ

u NE

IF o

r ND6

λtλu HBB1

NEIFHBB1

ND6HBB1

2nd Assay = HBB1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

39

Table 23 Comparison of Assays

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu)

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045 dd20180105 2372a-B ND6 94 05 R5 14 3 05526 36 1711 0074 dd20180105 2372a-B ND6 90 05 R5 14 3 05326 18 1649 0051 dd20180105 2372a-B ND6 88 05 R5 14 3 05077 22 1572 0052 dd20180105 2372a-B ND6 86 05 R5 14 3 05038 23 1560 0053 dd20180105 2372a-B ND6 84 05 R5 14 3 04906 10 1519 0041

dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

40

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials

Figure 20 summarizes the currently available λtλu results obtained by denaturing at 98 degC for 05 min rapidly cooling in the thermal cycler with dilutions that provide asymp 03 copiesdroplet The figure also displays the consensus result and its 95 confidence region estimated using the Hierarchical Bayes method provided in the NIST Consensus Builder (NICOB) system3839 Table 24 details the experimental conditions

Figure 20 Evaluation of Limiting Ratio

The symbols represent the ratio of the mean result of three or four technical replicates of the untreated and treated solutions the bars represent combined standard uncertainties The thick horizontal line represents a consensus value covering all displayed results the dotted horizontal lines bracket the 95 confidence interval about the consensus value

The Hierarchical Bayes consensus result for these values is 183 with a 95 confidence interval of plusmn003 All but three of the 16 λtλu values are within one standard uncertainty of this consensus value with most of the λtλu means lying within the 95 confidence interval This suggests that there are no statistically significant differences between the SRM 2372 and 2372a materials with regard to ssDNA proportion

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

17

18

19

20

λ tλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

41

Table 24 Evaluation of Limiting Ratio

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180104a 2372-B HBB1 14 4 03577 05 dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032

dd20180108 2372-B HBB1 14 3 03463 09 dd20180108 2372-B HBB1 98 05 R5 14 3 06562 14 1895 0031

dd20180109 2372-B HBB1 14 3 03577 25 dd20180109 2372-B HBB1 98 05 R5 14 3 06495 21 1816 0059

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 4 06151 15 1843 0040

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

42

The agreement among the measured λtλu is more directly visualized using estimated unilateral ldquoDegrees of Equivalencerdquo d Each d is the absolute difference between the measured and consensus values

119889119889 = |120582120582119905119905 120582120582119906119906frasl minus consensus|

and is associated with a 95 confidence interval U95(d) that combines the uncertainty estimates of the measured and consensus values The U95(d) calculation details are specific to the consensus estimation method39 Values of d plusmn U95(d) that include 0 indicate that the λtλu result does not differ from the majority at about a 95 level of confidence Figure 21 displays the d for the λtλu results for the Hierarchical Bayes consensus value39 By this figure of merit all the λtλu can be considered as members of a single population

Figure 21 Degrees of Equivalence

The symbols represent the Degree of Equivalence d for the λtλu results relative to the Hierarchical Bayes consensus value The bars represent 95 level of confidence uncertainty intervals d plusmn U95(d) The horizontal line marks the d = 0 ldquono differencerdquo value Results with d plusmn U95(d) that include this line are not significantly different from the majority

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

-024

-018

-012

-006

000

006

012

018

024

Degr

ees o

f Equ

ival

ence

λtλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

43

Conclusions and Recommendations With high confidence all tubes of the SRM 2372a solution have the same nDNA copy number content within measurement repeatability Additionally within measurement repeatability all tubes of the SRM 2372a have the same mtDNAnDNA With high confidence the nDNA copy number content of the SRM 2372a solution is thermally stable from 4 degC to 37 degC over an extended period The solution should not be shipped or stored below 4 degC With high confidence the ddPCR measurements used to establish the entity and mass concentrations and the mtDNAnDNA ratios are adequately precise and unbiased With high confidence the DNA in the SRM 2372a solutions is now dsDNA and can be expected to remain dsDNA for at least ten years when stored at 4 degC Recommended Certification Values

Since the component materials are believed to be homogeneous are of very similar composition and were prepared and processed similarly the observed differences in the measured entity concentration uncertainties are likely artifacts of the measurement process To limit over-interpretation of these differences we recommend that the copy per nanoliter values of the three components be assigned same 95 relative uncertainty interval (CV95) based on the largest of the measured CV95 values (93 ) rounded up to 10 Since the ngμL transformation does not appreciably inflate this relative uncertainty we recommend that the mass concentration values also be assigned CV95 of 10 Similarly we recommend that the mtDNAnDNA ratios be rounded to integer values all three components be assigned the same CV95 and that this uncertainty be based on the largest (21 ) of the measured CV95 rounded up to 25 Table 25 lists the values that we recommend be used in the Certificate of Analysis

Table 25 Recommended Values and 95 Uncertainties for Certification

Component Units Value U95(Value) a A Copies per nanoliter 151 15 B Copies per nanoliter 175 18 C Copies per nanoliter 145 15 A ngμL 498 50

B ngμL 578 58 C ngμL 479 48 A mtDNAnDNA 174 4

B mtDNAnDNA 206 5 C mtDNAnDNA 279 7

a The interval Value plusmn U95(Value) is believed with about 95 confidence to contain the true value of the measurand

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

44

Use of SRM 2372a for Value-Assigning In-House Reference Materials

The within-component and between-assay variability observed for the commercial qPCR assays is not unexpected since different qPCR methods exploit qualitatively different molecular targets Given the intrinsic variability of the human genome the prospect of developing a human DNA source material that responds equally to all current and future qPCR assays is improbable All three components should be used to elucidate potential bias when using these materials to value assign in-house DNA reference and control materials for qPCR assays

Certificate of Analysis In accordance with ISO Guide 31 2000 a NIST SRM certificate is a document containing the name description and intended purpose of the material the logo of the US Department of Commerce the name of NIST as a certifying body instructions for proper use and storage of the material certified property value(s) with associated uncertainty(ies) method(s) used to obtain property values the period of validity if appropriate and any other technical information deemed necessary for its proper use A Certificate is issued for an SRM certified for one or more specific physical or engineering performance properties and may contain NIST reference information or both values in addition to certified values A Certificate of Analysis is issued for an SRM certified for one or more specific chemical properties Note ISO Guide 31 is updated periodically check with ISO for the latest version [httpswwwnistgovsrmsrm-definitions] For the most current version of the Certificate of Analysis for NIST SRM 2372a Human DNA Quantification Standard please visit httpswww-snistgovsrmorsview_detailcfmsrm=2372a

References

1 National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research Department of Health Education and Welfare (1978) Belmont Report Ethical Principles and Guidelines for the Protection of Human Subjects of Research DHEW Publication No (OS) 78-0012 httpvideocastnihgovpdfohrp_belmont_reportpdf

2 CDCNIH Biosafety in Microbiological and Biomedical Laboratories 5th ed HHS publication No (CDC) 21-1112 Chosewood LC Wilson DE Eds US Government Printing Office Washington DC (2009) httpwwwcdcgovbiosafetypublicationsbmbl5indexhtm

3 Duewer DL Kline MC Redman JW Newall PJ Reeder DJ NIST Mixed Stain Studies 1 and 2 Interlaboratory Comparison of DNA Quantification Practice and Short Tandem Repeat Multiplex Performance with Multiple-Source Samples J Forensic Sci 200146(5)1199-1210 PMID 11569565

4 Kline MC Duewer DL Redman JW Butler JM NIST Mixed Stain Study 3 DNA quantitation accuracy and its influence on short tandem repeat multiplex signal intensity Anal Chem 200375(10)2463-2469 httpsdoiorg101021ac026410i

5 Kline MC Duewer DL Redman JW Butler JM Results from the NIST 2004 DNA Quantitation Study J Forensic Sci 200550(3)571-578 PMID 15932088

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

45

6 Kline MC Duewer DL Travis JC Smith MV Redman JW Vallone PM Decker AE Butler JM Production and Certification of Standard Reference Material 2372 Human DNA Quantitation Standard Anal Bioanal Chem 2009394(4)1183-1192 httpsdoiorg101007s00216-009-2782-0

7 Sambrook J and Russell DW (2001) Molecular Cloning a Laboratory Manual Cold Spring Harbor Laboratory Press Cold Spring Harbor New York

8 EU-JRC Protocol NK603 ndash Community Reference Laboratory Event-specific method for the quantification of maize line NK603 using real-time PCR Section 33 Spectrophotometric Measurement of DNA Concentration httpgmo-crljrceceuropaeusummariesNK603-WEB-ProtocolValidationpdf

9 ISO 215712005(E) Foodstuffs ndash Methods of analysis for the detection of genetically modified organisms and derived products ndash Nucleic acid extraction Annex B Methods for quantitation of extracted DNA pp 34-36 International Organization for Standardization Geneva (2005)

10 Certificate of Analysis Standard Reference Material 2372 Human DNA Quantitation Standard (2013) Gaithersburg MD USA Standard Reference Materials Program National Institute of Standards and Technology httpswww-snistgovsrmorsview_certcfmsrm=2372

11 Vogelstein B Kinzler KW Digital PCR Proc Natl Acad Sci USA 1999969236-9241 12 Hindson BJ Ness KD Masquelier DA Belgrader P Heredia NJ Makarewicz AJ Bright IJ Lucero

MY Hiddessen AL Legler TC Kitano TK Hodel MR Petersen JF Wyatt PW Steenblock ER Shah PH Bousse LJ Troup CB Mellen JC Wittmann DK Erndt NG Cauley TH Koehler RT So AP Dube S Rose KA Montesclaros L Wang S Stumbo DP Hodges SP Romine S Milanovich FP White HE Regan JF Karlin-Neumann GA Hindson CM Saxonov S Colston BW High-Throughput Droplet Digital PCR System for Absolute Quantitation of DNA Copy Number Anal Chem 2011838604-8610 httpsdoiorg101021ac202028g

13 Pinheiro LB Coleman VA Hindson CM Herrmann J Hindson BJ Bhat S Emslie KR Evaluation of a Droplet Digital Polymerase Chain Reaction Format for DNA Copy Number Quantification AnalChem 2012841003-1011 httpsdoiorg101021ac202578x

14 Corbisier P Pinheiro L Mazoua S Kortekaas AM Chung PY Gerganova T Roebben G Emons H Emslie K DNA copy number concentration measured by digital and droplet digital quantitative PCR using certified reference materials Anal Bioanal Chem 2015407(7)1831-1840 httpsdoiorg101007s00216-015-8458-z

15 Duewer DL Kline MC Romsos EL Real-time cdPCR opens a window into events occurring in the first few PCR amplification cycles Anal Bioanal Chem 2015407(30)9061-9069 httpsdoiorg101007s00216-015-9073-8

16 Kline MC Romsos EL Duewer DL Evaluating Digital PCR for the Quantification of Human Genomic DNA Accessible Amplifiable Targets Anal Chem 201688(4)2132-2139 httpsdoiorg101021acsanalchem5b03692

17 Kline MC Duewer DL Evaluating Droplet Digital Polymerase Chain Reaction for the Quantification of Human Genomic DNA Lifting the Traceability Fog Anal Chem 201789(8)4648-4654 httpsdoiorg101021acsanalchem7b00240

18 Duewer DL Kline MC Romsos EL Toman B Evaluating Droplet Digital PCR for the quantification of human genomic DNA Conversion from Copies per Nanoliter to Nanogram Genomic DNA per Microliter Anal Bioanal Chem 2018in press httpsdoiorg101007s00216-018-0982-1

19 Dagata JA Farkas N Kramer JA Method for measuring the volume of nominally 100 μm diameter spherical water in oil emulsion droplets NIST Special Publication 260-184 2016 httpsdoiorg106028NISTSP260-184

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

20 ISO 170342016 General requirements for the competence of reference material producers International Organization for Standardization Geneva (2016)

21 NCBI Standard Nucleotide BLAST httpsblastncbinlmnihgovBlastcgiPAGE_TYPE=BlastSearch

22 Timken MD Swango KL Orrego C Buoncristiani MR A Duplex Real-Time qPCR Assay for the Quantification of Human Nuclear and Mitochondrial DNA in Forensic Samples Implications for Quantifying DNA in Degraded Samples J Forensic Sci 200550(5)1044-60 PMID 16225209

23 Walker JA Hedges DJ Perodeau BP Landry KE Stoilova N Laborde ME Shewale J Sinha SK Batzer MA Multiplex polymerase chain reaction for simultaneous quantitation of human nuclear mitochondrial and male Y-chromosome DNA application in human identification Anal Biochem 200533789ndash97 httpsdoiorg101016jab200409036

24 Kavlick MF Lawrence HS Merritt RT Fisher C Isenberg A Robertson JM Budowle B Quantification of Human Mitochondrial DNA Using Synthesized DNA Standards J Forensic Sci 201156(6) 1457ndash1463 httpsdoiorg101111j1556-4029201101871x

25 Stulnig TM Amberger A Exposing contaminating phenol in nucleic acid preparations BioTechniques 199416(3) 403-404

26 Majumdar N Wessel T Marks J Digital PCR Modeling for Maximal Sensitivity Dynamic Range and Measurement Precision PLoS ONE 201510(3)e0118833 httpsdoi101371journalpone0118833

27 Tellinghuisen J Partition Volume Variability in Digital Polymerase Chain Reaction Methods Polydispersity Causes Bias but Can Improve Precision Anal Chem 2016881283-12187 httpsdoiorg101021acsanalchem6b03139

28 Lunn DJ Spiegelhalter D Thomas A Best N (2009) Statistics in Medicine 283049ndash3082 29 De Biegravevre P Dybkaer R Fajgelj A Hibbert DB Metrological traceability of measurement results in

chemistry Concepts and implementation Pure Appl Chem 201183(10)1873-1935 httpsdoiorg101007s00769-011-0805-y

30 Huggett JF Foy CA Benes V Emslie K Garson JA Haynes R Hellemans J Kubista M Mueller RD Nolan T Pfaffl MW Shipley GL Vandesompele J Wittwer CT Bustin SA The Digital MIQE Guidelines Minimum Information for Publication of Quantitative Digital PCR Experiments Clin Chem 201359(6)892-902 httpsdoiorg101373clinchem2013206375

31 Wilson PJ Ellison SLR Extending digital PCR analysis by modelling quantification cycle data BMC Bioinfo 201617421 httpsdoiorg101186s12859-016-1275-3

32 Wang X Son A Effects of pretreatment on the denaturation and fragmentation of genomic DNA for DNA hybridization Environ Sci Processes Impacts 2013152204 httpsdoiorg101039c3em00457k

33 Nwokeoji AO Kilby PM Portwood DE Dickman MJ Accurate Quantification of Nucleic Acids Using Hypochromicity Measurements in Conjunction with UV Spectrophotometry Anal Chem 201789(24)13567-13574 https101021acsanalchem7b04000

34 Farkas DH Kaul KL Wiedbrauk DL Kiechle FL Specimen collection and storage for diagnostic molecular pathology investigation Arch Pathol Lab Med 1996120(6)591-6

35 Kypr J Kejnovskaacute I Renčiuk D Vorličkovaacute M Circular dichroism and conformational polymorphism of DNA Nucleic Acids Res 200937(6)1713-1725 httpsdoiorg101093nargkp026

36 Strachan T Read AP (1999) Human Molecular Genetics 2nd Ed John Wiley amp Sons NY

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

37 Bhat S McLaughlin JL Emslie KR Effect of sustained elevated temperature prior to amplification on template copy number estimation using digital polymerase chain reaction Analyst 2011136724ndash32 httpsdoiorg101039c0an00484g

38 NIST Consensus Builder (NICOB) httpsconsensusnistgov 39 Koepke A Lafarge T Toman B Possolo A NIST Consensus Builder Userrsquos Manual

httpsconsensusnistgovNISTConsensusBuilder-UserManualpdf

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

Appendix A DNA Extraction Method The following method has been adapted from Miller et al (1988)A1 Procedure

1 Transfer approximately 5 mL of buffy coat into a 50 mL sterile polypropylene centrifuge tube and add 30 mL Red Blood Cell Lysis Buffer

2 Mix by gentle inversion and let stand at room temperature for 15 min

3 Centrifuge at approximately 209 rads (2000 rpm) for 20 min

4 Discard supernatant

5 Add 9 mL of Nucleic Lysis Buffer and re-suspend the pelleted cells by agitating with a plastic transfer pipette

6 Add 300 microL of 20 sodium dodecyl sulfate (SDS) and 18 mL of Protease K Solution (freshly prepared 2 mg to 5 mg Protease K per 1 mL Protease K Buffer)

7 Mix by inversion followed by incubation overnight or longer at 37 ordmC to 40 ordmC in a shaking water bath The length of incubation is determined by the time required to solubilize the cells present in the tube

8 Add 4 mL saturated ammonium acetate solution (in deionized water stored at 2 ordmC to 8 ordmC) Shake vigorously for 15 s

9 Centrifuge at approximately 471 rads (4500 rpm) for 10 min

10 Pipette supernatant into a 50 mL sterile polypropylene centrifuge tube and add two volumes of room temperature absolute ethanol Mix by gentle inversion until the DNA aggregates and floats to the top of the tube

11 Spool the aggregated DNA from the current tube into a fresh tube of 80 ethanol using a sterile inoculation loop At this step multiple tubes of precipitated DNA may be pooled to make a large lot of the purified material Mix by gentle inversion

12 Transfer the aggregated DNA to a fresh dry 50 mL polypropylene centrifuge tube Remove as much ethanol as possible by squeezing the DNA against the wall of the tube with a sterile inoculation loop Air-dry the material in a laminar flow hood

13 Resolubilize DNA in TE-4 buffer and store material at 4 ordmC Reagents

Red Blood Cell Lysis Buffer 144 mmolL ammonium chloride 1 mmolL sodium bicarbonate in deionized (DI) water Store at 2 ordmC to 8 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

Nucleic Lysis Buffer 10 mmolL Tris-HCl 400 mmolL sodium chloride 20 mmolL disodium ethylenediaminetetraacetic acid in DI Water adjust pH to 82 Store at 2 ordmC to 8 ordmC

Protease K Buffer

20 mmolL disodium ethylenediaminetetraacetic acid 10 (mass per volume) Sodium Dodecyl Sulfate (SDS) in DI water Store at 2 ordmC to 8 ordmC

Protease K

Sigma-Aldrich cat P6556 Tritirachium album lyophilized powder Store at ndash20 ordmC

Ammonium Acetate

192 molL ammonium acetate (saturated) in DI water Store at 2 ordmC to 8 ordmC

Absolute Ethanol (200 proof)

Store at room temperature Reference

A1 Miller SA Dykes DD Polesky HF A simple salting out procedure for extracting DNA from human nucleated cells Nucleic Acids Res 198816(3)1215 httpsdoiorg101093nar1631215

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

48

Appendix B Spectrophotometric Measurement Protocol The following is for use with a scanning recording spectrophotometer The procedures include the spectrophotometer calibration measurement of the ldquonativerdquo dsDNA and measurement of the NaOH denatured ssDNA The procedure for evaluating solutions after treatment with NaOH is adapted from references89 Spectrophotometer Calibration

Establish Baseline Set the spectrophotometer to acquire data from 220 nm to 345 nm in steps of 01 nm Zero the instrument with nothing (air) in the sample and reference cells Acquire an air-air baseline of the instrument with nothing (air) in the sample and reference cells If the trace has significant structure or excessive noise take corrective action If the trace is acceptable proceed with data collection Enable Verification of Wavelength Calibration SRM 2034 is a solution of holmium oxide in 10 perchloric acid contained in a sealed quartz cuvette delivering an intrinsic UVVis wavelength standard for absorption spectrophotometryB1 Evaluation of the location of peak maxima in the spectra of this solution enables verification of the spectrophotometers wavelength calibration Place SRM 2034 in the sample cell Create and name a file to hold the verification spectra Perform three replicate SRM 2034-to-air scans saving each scan to the file Remove SRM 2034 from the spectrophotometer and store appropriately Enable Verification of Absorbance Calibration SRM 2031 consists of three metal-on-fused silica filters held in cuvette-sized filter holders plus an empty filter holder The three filters are certified for transmittance and transmittance density (-log10(transmittance) effectively absorbance) at several wavelengths with nominal wavelength-independent transmittances of 10 30 and 90 Evaluation of the absorbance of each of these three filters at the certified 250 nm 280 nm and 340 nm wavelengths enables verification of the spectrophotometers absorbance calibration Place the empty SRM 2031 filter holder in the reference cell Remove the protective covers from the 10 transmittance filter and place in the sample cell Perform three replicate scans saving each scan to file Remove the filter from the spectrophotometer replace the protective covers and store appropriately Repeat for the 30 and 90 filters Close the file containing the verification spectra For ldquonativerdquo dsDNA Samples

Verify Optical Balance of Cuvettes The measurements of the DNA-containing samples are made using a 100 microL quartz cuvette in the sample cell compared to an optically matched cuvette in the reference cell The optical equivalence of these two cuvettes can be verified by filling both cuvettes with appropriate volumes of a sample solution matrix (TE-4 buffer) that does not contain DNA Fill two dry

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

49

clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes followed by a methanol rinse in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquonativerdquo dsDNA remove and clean both the reference cuvette and the sample cuvette For NaOH denatured ssDNA Samples

Prepare a 04 molL NaOH Stock Solution To a 15 mL disposable centrifuge tube add 016 g NaOH pellets add deionized (DI) water to the 10 mL line and mix until all the NaOH pellets have dissolved Centrifuge at 524 rads (5000 rpm) for 5 min For every 9 samples to be analyzed aliquot 1 mL of this 04 molL NaOH stock solution into a 15 mL screw top tube Use these tubes to prepare the NaOH blanks and DNA samples In a 15 mL screw cap vial mix 200 microL TE-4 buffer with 200 microL 04 molL sodium hydroxide solution to achieve a final NaOH concentration of 02 molL Vortex 5 s then centrifuge at 1257 rads (12 000 rpm) for 1 min Fill two dry clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Prepare the Sample Solutions For each unit of material to be measured (Tube or vial)

a) Vortex the unit for 5 s b) Centrifuge at 1257 rads (12 000 rpm) for 1 min c) Remove a 100 microL aliquot and place in a new labeled vial d) Add 100 microL of the freshly prepared 04 molL NaOH solution (with the same pipet) e) Vortex 5 s f) Centrifuge at 1257 rads (12 000 rpm) for 1 min

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

50

Measure Sample Solutions Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes rinse with methanol in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the NaOH denatured DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquodenaturedrdquo ssDNA close the file holding the spectra for this series Remove and clean both the reference cuvette and the sample cuvette After All Sample Measurements are Complete

Enable Verification of Spectrophotometer Stability The stability of the spectrophotometer over the course of the data collection can be evaluated by comparing spectra for SRM 2034 and 2031 taken at the start of data collection with spectra taken at the end of data collection Repeat the wavelength and absorbance verification measurements described above collecting the spectra in a suitably named file Clean Up Transfer all relevant data files from the spectrophotometers storage to a transfer device Clean the spectrophotometer work area as necessary and restore the SRMs and cuvettes to their appropriate storage locations Evaluation

Export the files to a spreadsheet Average the replicate spectra Subtract the average absorbance at A320 from the absorbance at A260 resulting in the corrected absorbance at A260 The conventional spectrophotometric estimate of the mass concentration of the dsDNA in the sample is achieved by multiplying the corrected absorbance by 50 The conventional estimate for the NaOH-treated ssDNA is achieved by multiplying the corrected absorbance by 37times2 where the ldquotimes2rdquo adjusts for the 1rarr2 volumetric dilution of the sample Note reliable spectrophotometric measurements require corrected A260 to be greater than 005 Reference

B1 Travis JC Acosta JC Andor G Bastie J Blattner P Chunnilall CJ Crosson SC Duewer DL Early EA Hengstberger F Kim C-S Liedquist L Manoocheri F Mercader F Metzdorf J Mito A Monardh LAG Nevask S Nilsson M Noeumll M Rodriguez AC Ruiacutez A Schirmacher A Smith MV Valencia G van Tonder N Zwinkels J Intrinsic Wavelength Standard Absorption Bands in Holmium Oxide Solution for UVvisible Molecular Absorption Spectrophotometry J Phys Chem Ref Data 200534(1)41-56

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

51

Appendix C SRM 2372a Statisticianrsquos report Extracted from Blaza Toman

November 30 2017 1 Introduction Candidate material for SRM 2372a is composed of three components labeled A B and C All components are human genomic DNA and prepared at NIST from buffy coat white blood cells from single-source anonymous donors Component A and B are single source (malefemale respectively) and Component C is a 1 male to 3 female mixture of two single-source donors Droplet digital PCR (ddPCR) assays were used to obtain measurements of the concentration of copies per microliter of sample Ten NIST-developed human nuclear DNA and three literature-derived human mitochondrial DNA PCR assays were used in this study

11 OpenBUGS Analysis All parameters in the following sections were estimated using Markov Chain Monte Carlo (MCMC) implemented in the OpenBUGS freewareC1 OpenBUGS is a software application for the Bayesian analysis of complex statistical models

12 OpenBUGS Exemplar Output All following sections list the OpenBUGs model (ldquocoderdquo) initialization values (ldquoinitsrdquo) if they are needed and the data used to evaluate the various data sets discussed in the body of this report homogeneity stability nDNA certification and mtDNAnDNA value assignment The provided information in each of the following sections is sufficient to produce estimates but the estimates may not exactly reproduce the exemplar results provided The MCMC paradigm does not produce ldquoanalyticalrdquo values but approaches them asymptotically as the number o model updates increases All exemplar results presented here are based upon the following bull 10-fold ldquothinningrdquo to minimize autocorrelation there were 10 MCMC estimates for every

model update used in parameter estimation bull a ldquoburn inrdquo of 5 000 updates before beginning parameter estimation and bull 10 000 model updates

The following statistics are provided in the tables listing the exemplar results

mean arithmetic mean of the parameterrsquos model update values sd standard deviation of the parameterrsquos model update values MC_error Monte Carlo standard deviation of the mean meanradic(number of model updates) val25pc 25th percentile of the model update values median 50th percentile of the model update values val975pc 975th percentile of the model update values startnumber of ldquoburn inrdquo updates sample number of model update values

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

52

2 Homogeneity assessment 21 Data

Samples of the material of all three components were drawn from 22 different boxes Two assays (NEIF ND6) were used to obtain 3 replicates of the measurement in copies per nanoliter for each sample and component For the complete homogeneity data see

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

22 Model The data were analyzed using a three-level Gaussian hierarchical ANOVA model

119910119910119894119894119894119894119894119894~119873119873120572120572119894119894119894119894119894119894 120590120590119908119908119894119894119905119905ℎ119894119894119894119894 119887119887119887119887119887119887 1198941198941198941198942

where ~N() denotes ldquodistributed as a normal distribution of some mean μ and variancerdquo 119894119894 denotes component (1 for A 2 for B 3 for C) 119895119895 denotes assay (1 for NEIF 2 for ND6) 119896119896 denotes box (1hellip22) and 120572120572119894119894119894119894119894119894~119873119873micro119894119894119894119894120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119887119887119887119887119887119887119894119894

2

All variance components were given a Half Cauchy prior as described in Reference C2 Note The Half Cauchy prior is specified by ldquo~dnorm(0 00016)I(0001)rdquo The ldquoIrdquo notation is for censoring It cuts off any random draws from the Normal distribution that are lower than 0001 An informative prior is needed when there are less than four data per distribution and the half Cauchy is a good general-purpose prior

23 OpenBUGS code and data Homogeneity code ModelBegin for(i in 13)xiNsb[i] ~ dnorm(0 00016)I(0001) chSqNsb[i] ~ dgamma(0505) sigb[i] lt- xiNsb[i]sqrt(chSqNsb[i]) for(i in 13)for(j in 12)mu[ij]~dnorm(010E-5) for(i in 13) muA[i]lt-mean(mu[i])sigbP[i]lt-100(muA[i]sqrt(sigb[i])) for(i in 13)for(j in 12)xiNs[ij] ~ dnorm(0 00016)I(0001) chSqNs[ij]~dgamma(0505) sigt[ij]lt-xiNs[ij]sqrt(chSqNs[ij]) sigtP[ij]lt-100(mu[ij]sqrt(sigt[ij])) for(i in 13)for(j in 12)for(k in 122)alpha[ijk]~dnorm(mu[ij]sigb[i]) for(idx in 1393)lamb[idx]~dnorm(alpha[cmp[idx]asy[idx]box[idx]]sigt[cmp[idx]asy[idx]]) ModelEnd Inits None required

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

53

Homogeneity Data lamb[] asy[] box[] cmp[] rep[] 1632 1 1 1 1 1529 1 1 1 2 1511 1 1 1 3 1616 2 1 1 1 1640 2 1 1 2 1582 2 1 1 3 1531 1 2 1 1 1525 1 2 1 2 1549 1 2 1 3 1615 2 2 1 1 1589 2 2 1 2 1601 2 2 1 3 1524 1 3 1 1 1540 1 3 1 2 1533 1 3 1 3 1582 2 3 1 1 1152 2 3 1 2 1555 2 3 1 3 1454 1 4 1 1 1458 1 4 1 2 1427 1 4 1 3 1627 2 4 1 1 1605 2 4 1 2 1607 2 4 1 3 1465 1 5 1 1 1540 1 5 1 2 1504 1 5 1 3 1607 2 5 1 1 1633 2 5 1 2 1586 2 5 1 3 1512 1 6 1 1 1491 1 6 1 2 1491 1 6 1 3 1511 2 6 1 1 1483 2 6 1 2 1524 2 6 1 3 1418 1 7 1 1 1480 1 7 1 2 1455 1 7 1 3 1479 2 7 1 1 1557 2 7 1 2 1656 2 7 1 3 1496 1 8 1 1 1560 1 8 1 2 1501 1 8 1 3 1514 2 8 1 1 1615 2 8 1 2 1540 2 8 1 3 1588 1 9 1 1 1581 1 9 1 2 1637 1 9 1 3 1474 2 9 1 1 1438 2 9 1 2 1454 2 9 1 3 1574 1 10 1 1 1513 1 10 1 2 1570 1 10 1 3 1594 2 10 1 1 1713 2 10 1 2 1669 2 10 1 3

1557 1 11 1 1 1508 1 11 1 2 1633 1 11 1 3 1571 2 11 1 1 1611 2 11 1 2 1631 2 11 1 3 1500 1 12 1 1 1476 1 12 1 2 1515 1 12 1 3 1626 2 12 1 1 1586 2 12 1 2 1608 2 12 1 3 1672 1 13 1 1 1634 1 13 1 2 1804 1 13 1 3 1490 2 13 1 1 1516 2 13 1 2 1497 2 13 1 3 1551 1 14 1 1 1568 1 14 1 2 1538 1 14 1 3 1617 2 14 1 1 1614 2 14 1 2 1639 2 14 1 3 1456 1 15 1 1 1448 1 15 1 2 1523 1 15 1 3 1525 2 15 1 1 1402 2 15 1 2 1436 2 15 1 3 1573 1 16 1 1 1520 1 16 1 2 1617 1 16 1 3 1634 2 16 1 1 1627 2 16 1 2 1575 2 16 1 3 1516 1 17 1 1 1516 1 17 1 2 1628 1 17 1 3 1613 2 17 1 1 1567 2 17 1 2 1567 2 17 1 3 1612 1 18 1 1 1535 1 18 1 2 1672 1 18 1 3 1529 2 18 1 1 1495 2 18 1 2 1528 2 18 1 3 1598 1 19 1 1 1514 1 19 1 2 1596 1 19 1 3 1571 2 19 1 1 1535 2 19 1 2 1511 2 19 1 3 1575 1 20 1 1 1568 1 20 1 2 1586 1 20 1 3 1652 2 20 1 1 1545 2 20 1 2 1627 2 20 1 3 1658 1 21 1 1

1611 1 21 1 2 1578 1 21 1 3 1647 2 21 1 1 1607 2 21 1 2 1612 2 21 1 3 1547 1 22 1 1 1528 1 22 1 2 1532 1 22 1 3 1537 2 22 1 1 1513 2 22 1 2 1536 2 22 1 3 1790 1 1 2 1 1909 1 1 2 2 1885 1 1 2 3 1870 2 1 2 2 1860 2 1 2 3 1712 1 2 2 1 1719 1 2 2 2 1704 1 2 2 3 1775 2 2 2 1 1718 2 2 2 2 1756 2 2 2 3 1693 1 3 2 1 1755 1 3 2 2 1749 1 3 2 3 1824 2 3 2 1 1483 2 3 2 2 1896 2 3 2 3 1754 1 4 2 1 1745 1 4 2 2 1739 1 4 2 3 1670 2 4 2 1 1632 2 4 2 2 1814 2 4 2 3 1757 1 5 2 1 1687 1 5 2 2 1735 1 5 2 3 1683 2 5 2 1 1759 2 5 2 2 1703 2 5 2 3 1697 1 6 2 1 1709 1 6 2 2 1714 1 6 2 3 1693 2 6 2 1 1723 2 6 2 2 1722 2 6 2 3 1722 1 7 2 1 1625 1 7 2 2 1703 1 7 2 3 1690 2 7 2 1 1684 2 7 2 2 1688 2 7 2 3 1884 1 8 2 1 1802 1 8 2 2 1772 1 8 2 3 1864 2 8 2 1 1887 2 8 2 2 1821 2 8 2 3 1713 1 9 2 1 1734 1 9 2 2 1727 1 9 2 3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

54

1709 2 9 2 1 1756 2 9 2 2 1758 2 9 2 3 1759 1 10 2 1 1809 1 10 2 2 1738 1 10 2 3 1764 2 10 2 1 1816 2 10 2 2 1787 2 10 2 3 1749 1 11 2 1 1782 1 11 2 2 1740 1 11 2 3 1741 2 11 2 1 1792 2 11 2 2 1784 2 11 2 3 1748 1 12 2 1 1712 1 12 2 2 1718 1 12 2 3 1797 2 12 2 1 1614 2 12 2 2 1790 2 12 2 3 1632 1 13 2 1 1668 1 13 2 2 1701 1 13 2 3 1626 2 13 2 1 1566 2 13 2 2 1681 2 13 2 3 1752 1 14 2 1 1535 1 14 2 2 1673 1 14 2 3 1724 2 14 2 1 1741 2 14 2 2 1685 2 14 2 3 1871 1 15 2 1 1835 1 15 2 2 1887 1 15 2 3 1770 2 15 2 1 1752 2 15 2 2 1798 2 15 2 3 1688 1 16 2 1 1725 1 16 2 2 1743 1 16 2 3 1744 2 16 2 1 1697 2 16 2 2 1710 2 16 2 3 1789 1 17 2 1 1769 1 17 2 2 1825 1 17 2 3 1716 2 17 2 1 1762 2 17 2 2 1638 2 17 2 3 1826 1 18 2 1 1780 1 18 2 2 1849 1 18 2 3 1805 2 18 2 1 1801 2 18 2 2 1931 2 18 2 3 1751 1 19 2 1 1759 1 19 2 2 1880 1 19 2 3 1827 2 19 2 1 1766 2 19 2 2

1786 2 19 2 3 1738 1 20 2 1 1678 1 20 2 2 1768 1 20 2 3 1752 2 20 2 1 1747 2 20 2 2 1768 2 20 2 3 1742 1 21 2 1 1836 1 21 2 2 1713 1 21 2 3 1746 2 21 2 1 1744 2 21 2 2 1764 2 21 2 3 1761 1 22 2 1 1743 1 22 2 2 1660 1 22 2 3 1666 2 22 2 1 1692 2 22 2 2 1819 2 22 2 3 1493 1 1 3 1 1385 1 1 3 2 1510 1 1 3 3 1565 2 1 3 1 1562 2 1 3 2 1466 2 1 3 3 1516 1 2 3 1 1513 1 2 3 2 1577 1 2 3 3 1490 2 2 3 1 1547 2 2 3 2 1570 2 2 3 3 1451 1 3 3 1 1528 1 3 3 2 1506 1 3 3 3 1485 2 3 3 1 1505 2 3 3 2 1471 2 3 3 3 1440 1 4 3 1 1486 1 4 3 2 1520 1 4 3 3 1529 2 4 3 1 1508 2 4 3 2 1506 2 4 3 3 1523 1 5 3 1 1472 1 5 3 2 1483 1 5 3 3 1513 2 5 3 1 1520 2 5 3 2 1491 2 5 3 3 1437 1 6 3 1 1493 1 6 3 2 1522 1 6 3 3 1492 2 6 3 1 1481 2 6 3 2 1461 2 6 3 3 1485 1 7 3 1 1477 1 7 3 2 1463 1 7 3 3 1475 2 7 3 1 1547 2 7 3 2 1485 2 7 3 3 1433 1 8 3 1

1405 1 8 3 2 1478 1 8 3 3 1497 2 8 3 1 1501 2 8 3 2 1495 2 8 3 3 1463 1 9 3 1 1499 1 9 3 2 1505 1 9 3 3 1503 2 9 3 1 1491 2 9 3 2 1516 2 9 3 3 1534 1 10 3 1 1458 1 10 3 2 1454 1 10 3 3 1448 2 10 3 1 1465 2 10 3 2 1474 2 10 3 3 1533 1 11 3 1 1514 1 11 3 2 1555 1 11 3 3 1488 2 11 3 1 1470 2 11 3 2 1483 2 11 3 3 1450 1 12 3 1 1525 1 12 3 2 1497 1 12 3 3 1496 2 12 3 1 1483 2 12 3 2 1519 2 12 3 3 1537 1 13 3 1 1527 1 13 3 2 1515 1 13 3 3 1499 2 13 3 1 1479 2 13 3 2 1475 2 13 3 3 1515 1 14 3 1 1490 1 14 3 2 1494 1 14 3 3 1469 2 14 3 1 1480 2 14 3 2 1466 2 14 3 3 1511 1 15 3 1 1568 1 15 3 2 1594 1 15 3 3 1515 2 15 3 1 1543 2 15 3 2 1565 2 15 3 3 1449 1 16 3 1 1452 1 16 3 2 1471 1 16 3 3 1543 2 16 3 1 1448 2 16 3 2 1491 2 16 3 3 1388 1 17 3 1 1421 1 17 3 2 1394 1 17 3 3 1452 2 17 3 1 1435 2 17 3 3 1475 1 18 3 1 1455 1 18 3 2 1445 1 18 3 3 1458 2 18 3 1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

55

1531 2 18 3 2 1459 2 18 3 3 1468 1 19 3 1 1525 1 19 3 2 1446 1 19 3 3 1435 2 19 3 1 1478 2 19 3 2 1421 2 19 3 3 1549 1 20 3 1

1531 1 20 3 2 1499 1 20 3 3 1499 2 20 3 1 1465 2 20 3 2 1468 2 20 3 3 1530 1 21 3 1 1440 1 21 3 2 1484 1 21 3 3 1499 2 21 3 1

1454 2 21 3 2 1442 1 22 3 1 1487 1 22 3 2 1436 1 22 3 3 1589 2 22 3 1 1591 2 22 3 2 1556 2 22 3 3 END

Table C-1 Example OpenBUGS Homogeneity Summary

mean sd MC_error val25pc median val975pc start sample mu[11] 15460 0122 000123 1522 1546 1570 5001 10000 mu[12] 15640 0137 000141 1537 1564 1591 5001 10000 mu[21] 17490 0113 000118 1727 1749 1772 5001 10000 mu[22] 17480 0129 000125 1722 1748 1774 5001 10000 mu[31] 14870 0073 000073 1473 1487 1501 5001 10000 mu[32] 14960 0069 000063 1482 1496 1509 5001 10000 muA[1] 15550 0092 000092 1536 1555 1573 5001 10000 muA[2] 17490 0085 000086 1732 1749 1765 5001 10000 muA[3] 14910 0050 000046 1481 1491 1501 5001 10000 sigbP[1] 3269 04985 000514 2376 3244 4324 5001 10000 sigbP[2] 2548 04385 000461 1733 2533 3472 5001 10000 sigbP[3] 1850 0292 000260 1329 1832 2477 5001 10000

sigtP[11] 2774 0298 000260 2268 2750 3442 5001 10000 sigtP[12] 4036 0429 000438 3290 4005 4986 5001 10000 sigtP[21] 2664 0301 000321 2155 2640 3313 5001 10000 sigtP[22] 3807 0395 000377 3116 3772 4662 5001 10000 sigtP[31] 2313 0250 000209 1880 2291 2873 5001 10000 sigtP[32] 1812 0195 000170 1473 1797 2243 5001 10000

24 Results The material appears to be homogeneous as can be seen in Figure C-1 Based on these results the between-tube variability is of the same order as within-tube variability

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

56

Component A Assay NEIF Component A Assay ND6 box plot gamma[11]

[111][112][113]

[114]

[115][116]

[117]

[118]

[119]

[1110][1111]

[1112]

[1113]

[1114]

[1115]

[1116][1117]

[1118]

[1119][1120]

[1121]

[1122]

114

box plot gamma[12]

[121][122]

[123]

[124][125]

[126]

[127][128]

[129]

[1210]

[1211][1212]

[1213]

[1214]

[1215]

[1216]

[1217]

[1218][1219]

[1220][1221]

[1222]

115

Component B Assay NEIF Component B Assay ND6

box plot gamma[21]

[211]

[212]

[213][214]

[215][216]

[217]

[218]

[219]

[2110][2111]

[2112]

[2113][2114]

[2115]

[2116]

[2117]

[2118]

[2119]

[2120]

[2121]

[2122]

129

box plot gamma[22]

[221]

[222][223]

[224][225][226][227]

[228]

[229]

[2210][2211]

[2212]

[2213]

[2214]

[2215]

[2216][2217]

[2218]

[2219]

[2220][2221][2222]

128

Component C Assay NEIF Component C Assay ND6

box plot gamma[31]

[311]

[312]

[313][314]

[315][316]

[317]

[318]

[319][3110]

[3111]

[3112]

[3113]

[3114]

[3115]

[3116]

[3117]

[3118]

[3119]

[3120]

[3121]

[3122]109

box plot gamma[32]

[321][322]

[323]

[324][325]

[326]

[327][328][329]

[3210]

[3211]

[3212]

[3213][3214]

[3215]

[3216]

[3217]

[3218]

[3219]

[3220][3221]

[3222]

11

Figure C-1 Homogeneity Assessment

Note These results were obtained using the 1-4 gravimetric dilution-adjusted λ but uncorrected for droplet volume

or the 1-10 volumetric dilution That is the vertical axis in each panel should be multiplied by 07349times10 = 7349 This constant difference in scale does not impact the homogeneity analysis Error bars span approximate 95 confidence intervals

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

57

3 Certification measurements 31 Data

Originally the experiment consisted of 5 plates containing the 3 components (A B C) and 12 assays Ten of the assays were for genomic DNA identification (for the certified value) and two of the assays were for mitochondrial detection which is used for informational values (and are not included in this data set) Each of the plates contained samples from a new vial from different boxes (noted in the ldquovial usedrdquo portion the number references the box) The assay positions were alternated across the five plates as well For the plate design and which tubes were used see

Figure 6 Basic Plate Layout Design for Certification Measurements It was later determined that the homogeneity and stability data sets were obtained using a different set of pipettes than the certification data set and since all data was to be combined for certification another plate was obtained using the original pipette For the complete certification data see

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component Stability data was measured at different temperatures over several weeks using assays NEIF and ND6 Only the 4degC data was also used for certification For the complete stability data see

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

32 Model It was decided that all three sets of data should be used to derive the certified values In this way the uncertainty of the certified values would better reflect any potential heterogeneity in the samples as well as any additional unknown factors that may have effected the results Because two different pipettes were used to obtain different sets of data there was a significant bias in some of the measurements Specifically the ldquocertificationrdquo data from the first five plates was obtained using pipette set ldquo1rdquo and the remaining measurements were obtained using pipette set ldquo2rdquo This bias was introduced into the statistical model as part of the mean Possible effects of different assays and type of data (heterogeneity certification and stability) were accounted for in the statistical model in terms of a random effect Specifically the response for each component (A B C) was represented as

119910119910119894119894119894119894119894119894~1198731198731205741205741198941198941198941198941198941198941205901205901198941198942 where ~N() denotes ldquodistributed as a normal distribution of some mean and variancerdquo i = 123 denotes type of measurement heterogeneity certification and stability j = 1hellip10 denotes assay 2PR4 POTP NEIF R4Q5 D5 ND6 D9 HBB1 ND14 22C3 k = 1 2 denotes pipette set 120574120574119894119894119894119894119894119894 = 120572120572119894119894119894119894 + 120573120573119894119894 120572120572119894119894119894119894~119873119873(micro1198941198941205901205901205721205722) 120573120573119894119894 = 120573120573 for data type 1 and 120573120573119894119894 = 0 for data types 2 and 3 For a given 2372a component the variance 1205901205901205721205722 measures the between assay variability and 1205901205901198941198942 measures the repeatability variability for each data type The data was analyzed using Bayesian MCMC programmed in OpenBUGSC2 with non-informative Gaussian priors for the means and Gamma priors for the variance components

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

58

33 OpenBUGS code inits and data Certification code ModelBegin for(i in 13)mu[i]~dnorm(010E-5) sig[i]~dgamma(10E-510E-5) sigalph~dgamma(10E-510E-5) for(i in 110)alpha[1i]~dnorm(mu[1] sigalph) for(i in 12)alpha[2i]~dnorm(mu[2] sigalph) alpha[3i]~dnorm(mu[3]sigalph) b~dnorm(010E-5) beta[1]lt-b beta[2]lt-00 for(i in 1Ntot)mean[i]lt-alpha[typ[i]asy[i]]+beta[pip[i]] for(i in 1Ntot)lamb[i]~dnorm(mean[i]sig[typ[i]]) ModelEnd Certification Inits list(sig=c(111) sigalph=1) Certification Data for Component A list(Ntot=275) lamb[] asy[] cmp[] typ[] pip[] 1667 1 1 1 1 1655 2 1 1 1 1647 3 1 1 1 1594 4 1 1 1 1655 5 1 1 1 1629 6 1 1 1 1655 7 1 1 1 1644 8 1 1 1 1645 9 1 1 1 1545 10 1 1 1 1698 1 1 1 1 1685 2 1 1 1 1665 3 1 1 1 1658 4 1 1 1 1683 5 1 1 1 1647 6 1 1 1 1652 7 1 1 1 1611 8 1 1 1 1707 9 1 1 1 1603 10 1 1 1 1668 1 1 1 1 1587 2 1 1 1 1678 3 1 1 1 1628 4 1 1 1 1738 5 1 1 1 1667 6 1 1 1 1695 7 1 1 1 1757 8 1 1 1 1771 9 1 1 1 1762 10 1 1 1 1739 1 1 1 1 1702 2 1 1 1 1614 3 1 1 1 1762 4 1 1 1 1649 5 1 1 1 1678 6 1 1 1 1653 7 1 1 1

175 8 1 1 1 1764 9 1 1 1 1759 10 1 1 1 1638 1 1 1 1 1676 2 1 1 1 1616 3 1 1 1 1559 4 1 1 1 1608 5 1 1 1 1611 6 1 1 1 1517 7 1 1 1 1557 8 1 1 1 1604 9 1 1 1 1675 10 1 1 1 1667 1 1 1 1 1638 2 1 1 1 1624 3 1 1 1 1587 4 1 1 1 1652 5 1 1 1 1589 6 1 1 1 1607 7 1 1 1 1582 8 1 1 1 1625 9 1 1 1 1616 10 1 1 1 1888 1 1 1 1 1868 2 1 1 1 1953 3 1 1 1 1779 4 1 1 1 1924 5 1 1 1 1861 6 1 1 1 1824 7 1 1 1 1809 8 1 1 1 1864 9 1 1 1 193 10 1 1 1 1939 1 1 1 1 1815 2 1 1 1 1902 3 1 1 1 1871 4 1 1 1 2124 5 1 1 1

1787 6 1 1 1 1844 7 1 1 1 1799 8 1 1 1 1857 9 1 1 1 1942 10 1 1 1 1597 1 1 1 1 1533 2 1 1 1 1462 3 1 1 1 1493 4 1 1 1 1538 5 1 1 1 1492 6 1 1 1 1528 7 1 1 1 1548 8 1 1 1 1513 9 1 1 1 1562 10 1 1 1 159 1 1 1 1 1404 2 1 1 1 1533 3 1 1 1 1519 4 1 1 1 1586 5 1 1 1 1516 6 1 1 1 1538 7 1 1 1 1581 8 1 1 1 1514 9 1 1 1 1431 1 1 1 2 1407 2 1 1 2 1487 3 1 1 2 145 4 1 1 2 1394 5 1 1 2 1396 6 1 1 2 1414 7 1 1 2 1401 8 1 1 2 1412 9 1 1 2 1435 10 1 1 2 1411 1 1 1 2 1395 2 1 1 2 1398 3 1 1 2 1439 4 1 1 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

59

1422 5 1 1 2 1393 6 1 1 2 139 7 1 1 2 1382 8 1 1 2 146 9 1 1 2 141 10 1 1 2 1632 1 1 2 2 1531 1 1 2 2 1524 1 1 2 2 1454 1 1 2 2 1465 1 1 2 2 1512 1 1 2 2 1418 1 1 2 2 1496 1 1 2 2 1588 1 1 2 2 1574 1 1 2 2 1557 1 1 2 2 15 1 1 2 2 1672 1 1 2 2 1551 1 1 2 2 1573 1 1 2 2 1516 1 1 2 2 1612 1 1 2 2 1598 1 1 2 2 1575 1 1 2 2 1658 1 1 2 2 1547 1 1 2 2 1529 1 1 2 2 1525 1 1 2 2 154 1 1 2 2 1458 1 1 2 2 154 1 1 2 2 1491 1 1 2 2 148 1 1 2 2 156 1 1 2 2 1581 1 1 2 2 1513 1 1 2 2 1508 1 1 2 2 1476 1 1 2 2 1634 1 1 2 2 1568 1 1 2 2 152 1 1 2 2 1516 1 1 2 2 1535 1 1 2 2 1514 1 1 2 2 1568 1 1 2 2 1611 1 1 2 2 1528 1 1 2 2 1511 1 1 2 2 1549 1 1 2 2 1533 1 1 2 2 1427 1 1 2 2 1504 1 1 2 2 1491 1 1 2 2 1455 1 1 2 2

1501 1 1 2 2 1637 1 1 2 2 157 1 1 2 2 1633 1 1 2 2 1515 1 1 2 2 1804 1 1 2 2 1538 1 1 2 2 1617 1 1 2 2 1628 1 1 2 2 1672 1 1 2 2 1596 1 1 2 2 1586 1 1 2 2 1578 1 1 2 2 1532 1 1 2 2 1616 2 1 2 2 1615 2 1 2 2 1582 2 1 2 2 1627 2 1 2 2 1607 2 1 2 2 1511 2 1 2 2 1479 2 1 2 2 1514 2 1 2 2 1474 2 1 2 2 1594 2 1 2 2 1571 2 1 2 2 1626 2 1 2 2 149 2 1 2 2 1617 2 1 2 2 1634 2 1 2 2 1613 2 1 2 2 1529 2 1 2 2 1571 2 1 2 2 1652 2 1 2 2 1647 2 1 2 2 1537 2 1 2 2 164 2 1 2 2 1589 2 1 2 2 1152 2 1 2 2 1605 2 1 2 2 1633 2 1 2 2 1483 2 1 2 2 1557 2 1 2 2 1615 2 1 2 2 1438 2 1 2 2 1713 2 1 2 2 1611 2 1 2 2 1586 2 1 2 2 1516 2 1 2 2 1614 2 1 2 2 1627 2 1 2 2 1567 2 1 2 2 1495 2 1 2 2 1535 2 1 2 2 1545 2 1 2 2 1607 2 1 2 2

1513 2 1 2 2 1582 2 1 2 2 1601 2 1 2 2 1555 2 1 2 2 1607 2 1 2 2 1586 2 1 2 2 1524 2 1 2 2 1656 2 1 2 2 154 2 1 2 2 1454 2 1 2 2 1669 2 1 2 2 1631 2 1 2 2 1608 2 1 2 2 1497 2 1 2 2 1639 2 1 2 2 1575 2 1 2 2 1567 2 1 2 2 1528 2 1 2 2 1511 2 1 2 2 1627 2 1 2 2 1612 2 1 2 2 1536 2 1 2 2 1568 1 1 3 2 152 1 1 3 2 1486 1 1 3 2 1441 1 1 3 2 1706 1 1 3 2 144 1 1 3 2 15 1 1 3 2 1537 1 1 3 2 1387 1 1 3 2 1463 1 1 3 2 157 1 1 3 2 1429 1 1 3 2 1667 1 1 3 2 1632 1 1 3 2 1589 1 1 3 2 1474 2 1 3 2 1543 2 1 3 2 1555 2 1 3 2 1599 2 1 3 2 1494 2 1 3 2 1624 2 1 3 2 1577 2 1 3 2 1593 2 1 3 2 1479 2 1 3 2 1517 2 1 3 2 1538 2 1 3 2 159 2 1 3 2 1479 2 1 3 2 1479 2 1 3 2 1522 2 1 3 2 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

60

Table C-2 Example OpenBUGS Certification Summary for Component A

mean sd MC_error val25pc median val975pc start sample b 2584 02948 0005376 2001 2588 3158 100001 100000 mu[1] 1417 02706 0005366 1364 1416 147 100001 100000 mu[2] 1559 008368 4776E-4 1543 1559 1576 100001 100000 mu[3] 1533 01504 0001419 1504 1533 1563 100001 100000

Certification Data for Component B list(Ntot=278) lamb[] asy[] cmp[] typ[] pip[] 1869 1 2 1 1 185 2 2 1 1 1709 3 2 1 1 1856 4 2 1 1 1848 5 2 1 1 1828 6 2 1 1 1861 7 2 1 1 1825 8 2 1 1 1871 9 2 1 1 1776 10 2 1 1 196 1 2 1 1 1781 2 2 1 1 1707 3 2 1 1 1787 4 2 1 1 1829 5 2 1 1 1826 6 2 1 1 1858 7 2 1 1 1869 8 2 1 1 1773 9 2 1 1 1902 10 2 1 1 2174 1 2 1 1 2202 2 2 1 1 2172 3 2 1 1 223 5 2 1 1 2094 6 2 1 1 1949 7 2 1 1 1971 8 2 1 1 2123 9 2 1 1 2107 10 2 1 1 2231 1 2 1 1 2292 2 2 1 1 2216 3 2 1 1 2078 4 2 1 1 2192 5 2 1 1 2034 6 2 1 1 211 7 2 1 1 2123 8 2 1 1 2066 9 2 1 1 2046 10 2 1 1 181 1 2 1 1 1865 2 2 1 1 1806 3 2 1 1 1878 4 2 1 1 1824 5 2 1 1 1768 6 2 1 1 1811 7 2 1 1 1763 8 2 1 1 1865 9 2 1 1 1815 10 2 1 1

1806 1 2 1 1 1801 2 2 1 1 1786 3 2 1 1 1768 4 2 1 1 179 5 2 1 1 167 6 2 1 1 1749 7 2 1 1 1741 8 2 1 1 1781 9 2 1 1 1866 10 2 1 1 1784 1 2 1 1 1735 2 2 1 1 1778 3 2 1 1 1835 4 2 1 1 1794 5 2 1 1 17 6 2 1 1 1799 7 2 1 1 1765 8 2 1 1 1754 9 2 1 1 1792 10 2 1 1 1821 1 2 1 1 1813 2 2 1 1 1813 3 2 1 1 1749 4 2 1 1 1848 5 2 1 1 1711 6 2 1 1 1779 7 2 1 1 1769 8 2 1 1 1725 9 2 1 1 1814 10 2 1 1 1718 1 2 1 1 1753 2 2 1 1 1773 3 2 1 1 1777 4 2 1 1 1803 5 2 1 1 1718 6 2 1 1 1802 7 2 1 1 1834 8 2 1 1 1731 9 2 1 1 1842 10 2 1 1 1768 1 2 1 1 1698 2 2 1 1 1792 3 2 1 1 174 4 2 1 1 1795 5 2 1 1 1693 6 2 1 1 1743 7 2 1 1 1745 8 2 1 1 1714 9 2 1 1 1791 1 2 1 2

1746 2 2 1 2 1803 3 2 1 2 1767 4 2 1 2 1745 5 2 1 2 1824 6 2 1 2 1752 7 2 1 2 1813 8 2 1 2 1648 9 2 1 2 1818 10 2 1 2 177 1 2 1 2 182 2 2 1 2 1803 3 2 1 2 1739 4 2 1 2 1773 5 2 1 2 1729 6 2 1 2 1744 7 2 1 2 1754 8 2 1 2 1671 9 2 1 2 1741 10 2 1 2 179 1 2 2 2 1712 1 2 2 2 1693 1 2 2 2 1754 1 2 2 2 1757 1 2 2 2 1697 1 2 2 2 1722 1 2 2 2 1884 1 2 2 2 1713 1 2 2 2 1759 1 2 2 2 1749 1 2 2 2 1748 1 2 2 2 1632 1 2 2 2 1752 1 2 2 2 1871 1 2 2 2 1688 1 2 2 2 1789 1 2 2 2 1826 1 2 2 2 1751 1 2 2 2 1738 1 2 2 2 1742 1 2 2 2 1761 1 2 2 2 1909 1 2 2 2 1719 1 2 2 2 1755 1 2 2 2 1745 1 2 2 2 1687 1 2 2 2 1709 1 2 2 2 1625 1 2 2 2 1802 1 2 2 2 1734 1 2 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

61

1809 1 2 2 2 1782 1 2 2 2 1712 1 2 2 2 1668 1 2 2 2 1535 1 2 2 2 1835 1 2 2 2 1725 1 2 2 2 1769 1 2 2 2 178 1 2 2 2 1759 1 2 2 2 1678 1 2 2 2 1836 1 2 2 2 1743 1 2 2 2 1885 1 2 2 2 1704 1 2 2 2 1749 1 2 2 2 1739 1 2 2 2 1735 1 2 2 2 1714 1 2 2 2 1703 1 2 2 2 1772 1 2 2 2 1727 1 2 2 2 1738 1 2 2 2 174 1 2 2 2 1718 1 2 2 2 1701 1 2 2 2 1673 1 2 2 2 1887 1 2 2 2 1743 1 2 2 2 1825 1 2 2 2 1849 1 2 2 2 188 1 2 2 2 1768 1 2 2 2 1713 1 2 2 2 166 1 2 2 2 1775 2 2 2 2 1824 2 2 2 2 167 2 2 2 2 1683 2 2 2 2 1693 2 2 2 2 169 2 2 2 2 1864 2 2 2 2 1709 2 2 2 2 1764 2 2 2 2

1741 2 2 2 2 1797 2 2 2 2 1626 2 2 2 2 1724 2 2 2 2 177 2 2 2 2 1744 2 2 2 2 1716 2 2 2 2 1805 2 2 2 2 1827 2 2 2 2 1752 2 2 2 2 1746 2 2 2 2 1666 2 2 2 2 187 2 2 2 2 1718 2 2 2 2 1483 2 2 2 2 1632 2 2 2 2 1759 2 2 2 2 1723 2 2 2 2 1684 2 2 2 2 1887 2 2 2 2 1756 2 2 2 2 1816 2 2 2 2 1792 2 2 2 2 1614 2 2 2 2 1566 2 2 2 2 1741 2 2 2 2 1752 2 2 2 2 1697 2 2 2 2 1762 2 2 2 2 1801 2 2 2 2 1766 2 2 2 2 1747 2 2 2 2 1744 2 2 2 2 1692 2 2 2 2 186 2 2 2 2 1756 2 2 2 2 1896 2 2 2 2 1814 2 2 2 2 1703 2 2 2 2 1722 2 2 2 2 1688 2 2 2 2 1821 2 2 2 2 1758 2 2 2 2 1787 2 2 2 2

1784 2 2 2 2 179 2 2 2 2 1681 2 2 2 2 1685 2 2 2 2 1798 2 2 2 2 171 2 2 2 2 1638 2 2 2 2 1931 2 2 2 2 1786 2 2 2 2 1768 2 2 2 2 1764 2 2 2 2 1819 2 2 2 2 1666 1 2 3 2 1626 1 2 3 2 188 1 2 3 2 1653 1 2 3 2 1733 1 2 3 2 1551 1 2 3 2 1884 1 2 3 2 1718 1 2 3 2 1549 1 2 3 2 1723 1 2 3 2 194 1 2 3 2 164 1 2 3 2 1565 1 2 3 2 1787 1 2 3 2 1655 2 2 3 2 167 2 2 3 2 1673 2 2 3 2 1705 2 2 3 2 1755 2 2 3 2 1625 2 2 3 2 1708 2 2 3 2 1705 2 2 3 2 1862 2 2 3 2 1686 2 2 3 2 1723 2 2 3 2 1909 2 2 3 2 1859 2 2 3 2 1897 2 2 3 2 1972 2 2 3 2 END

Table C-3 Example OpenBUGS Certification Summary for Component B

mean sd MC_error val25pc median val975pc start sample b 09337 03387 0007504 02695 09371 1592 100001 100000 mu[1] 1764 03101 0007483 1704 1763 1825 100001 100000 mu[2] 1748 008014 4719E-4 1733 1748 1764 100001 100000 mu[3] 1735 02338 0003093 1688 1736 1781 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

62

Certification data for Component C list(Ntot=280) lamb[] asy[] cmp[] typ[] pip[] 163 1 3 1 1 1596 2 3 1 1 1574 3 3 1 1 159 4 3 1 1 161 5 3 1 1 1589 6 3 1 1 1569 7 3 1 1 1565 8 3 1 1 1564 9 3 1 1 1548 10 3 1 1 1706 1 3 1 1 1576 2 3 1 1 1538 3 3 1 1 1555 4 3 1 1 1607 5 3 1 1 1548 6 3 1 1 1572 7 3 1 1 1681 8 3 1 1 1492 9 3 1 1 1655 10 3 1 1 1499 1 3 1 1 1576 2 3 1 1 1545 3 3 1 1 1584 4 3 1 1 1577 5 3 1 1 1599 6 3 1 1 1587 7 3 1 1 152 8 3 1 1 1479 9 3 1 1 1512 10 3 1 1 1497 1 3 1 1 1611 2 3 1 1 156 3 3 1 1 1546 4 3 1 1 1529 5 3 1 1 1478 6 3 1 1 1504 7 3 1 1 1551 8 3 1 1 1438 9 3 1 1 1494 10 3 1 1 1452 1 3 1 1 1347 2 3 1 1 1395 3 3 1 1 1442 4 3 1 1 1448 5 3 1 1 1463 6 3 1 1 1402 7 3 1 1 1474 8 3 1 1 1466 9 3 1 1 1439 10 3 1 1 1424 1 3 1 1 1397 2 3 1 1 135 3 3 1 1 1459 4 3 1 1 1602 5 3 1 1 1435 6 3 1 1 1433 7 3 1 1 1481 8 3 1 1 1482 9 3 1 1

1568 10 3 1 1 147 1 3 1 1 1411 2 3 1 1 1512 3 3 1 1 1452 4 3 1 1 1567 5 3 1 1 148 6 3 1 1 1442 7 3 1 1 143 8 3 1 1 1413 9 3 1 1 1486 10 3 1 1 1444 1 3 1 1 1427 2 3 1 1 1439 3 3 1 1 1468 4 3 1 1 1524 5 3 1 1 1424 6 3 1 1 1459 7 3 1 1 1467 8 3 1 1 1411 9 3 1 1 148 10 3 1 1 1485 1 3 1 1 1422 2 3 1 1 1438 3 3 1 1 1439 4 3 1 1 1515 5 3 1 1 1409 6 3 1 1 1469 7 3 1 1 1453 8 3 1 1 1465 9 3 1 1 1363 10 3 1 1 1458 1 3 1 1 1459 2 3 1 1 1465 3 3 1 1 1468 4 3 1 1 1457 5 3 1 1 1412 6 3 1 1 1474 7 3 1 1 1428 8 3 1 1 1425 9 3 1 1 1393 10 3 1 1 1357 1 3 1 2 1447 2 3 1 2 1448 3 3 1 2 1409 4 3 1 2 1389 5 3 1 2 1415 6 3 1 2 1431 7 3 1 2 1381 8 3 1 2 1403 9 3 1 2 1435 10 3 1 2 1364 1 3 1 2 1361 2 3 1 2 1404 3 3 1 2 1411 4 3 1 2 1429 5 3 1 2 1417 6 3 1 2 138 7 3 1 2 1336 8 3 1 2 1387 9 3 1 2

1372 10 3 1 2 1493 1 3 2 2 1516 1 3 2 2 1451 1 3 2 2 144 1 3 2 2 1523 1 3 2 2 1437 1 3 2 2 1485 1 3 2 2 1433 1 3 2 2 1463 1 3 2 2 1534 1 3 2 2 1533 1 3 2 2 145 1 3 2 2 1537 1 3 2 2 1515 1 3 2 2 1511 1 3 2 2 1449 1 3 2 2 1388 1 3 2 2 1475 1 3 2 2 1468 1 3 2 2 1549 1 3 2 2 153 1 3 2 2 1442 1 3 2 2 1385 1 3 2 2 1513 1 3 2 2 1528 1 3 2 2 1486 1 3 2 2 1472 1 3 2 2 1493 1 3 2 2 1477 1 3 2 2 1405 1 3 2 2 1499 1 3 2 2 1458 1 3 2 2 1514 1 3 2 2 1525 1 3 2 2 1527 1 3 2 2 149 1 3 2 2 1568 1 3 2 2 1452 1 3 2 2 1421 1 3 2 2 1455 1 3 2 2 1525 1 3 2 2 1531 1 3 2 2 144 1 3 2 2 1487 1 3 2 2 151 1 3 2 2 1577 1 3 2 2 1506 1 3 2 2 152 1 3 2 2 1483 1 3 2 2 1522 1 3 2 2 1463 1 3 2 2 1478 1 3 2 2 1505 1 3 2 2 1454 1 3 2 2 1555 1 3 2 2 1497 1 3 2 2 1515 1 3 2 2 1494 1 3 2 2 1594 1 3 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

63

1471 1 3 2 2 1394 1 3 2 2 1445 1 3 2 2 1446 1 3 2 2 1499 1 3 2 2 1484 1 3 2 2 1436 1 3 2 2 1565 2 3 2 2 149 2 3 2 2 1485 2 3 2 2 1529 2 3 2 2 1513 2 3 2 2 1492 2 3 2 2 1475 2 3 2 2 1497 2 3 2 2 1503 2 3 2 2 1448 2 3 2 2 1488 2 3 2 2 1496 2 3 2 2 1499 2 3 2 2 1469 2 3 2 2 1515 2 3 2 2 1543 2 3 2 2 1452 2 3 2 2 1458 2 3 2 2 1435 2 3 2 2 1499 2 3 2 2 1499 2 3 2 2 1589 2 3 2 2 1562 2 3 2 2 1547 2 3 2 2 1505 2 3 2 2 1508 2 3 2 2 152 2 3 2 2

1481 2 3 2 2 1547 2 3 2 2 1501 2 3 2 2 1491 2 3 2 2 1465 2 3 2 2 147 2 3 2 2 1483 2 3 2 2 1479 2 3 2 2 148 2 3 2 2 1543 2 3 2 2 1448 2 3 2 2 1531 2 3 2 2 1478 2 3 2 2 1465 2 3 2 2 1454 2 3 2 2 1591 2 3 2 2 1466 2 3 2 2 157 2 3 2 2 1471 2 3 2 2 1506 2 3 2 2 1491 2 3 2 2 1461 2 3 2 2 1485 2 3 2 2 1495 2 3 2 2 1516 2 3 2 2 1474 2 3 2 2 1483 2 3 2 2 1519 2 3 2 2 1475 2 3 2 2 1466 2 3 2 2 1565 2 3 2 2 1491 2 3 2 2 1435 2 3 2 2 1459 2 3 2 2

1421 2 3 2 2 1468 2 3 2 2 1556 2 3 2 2 1424 1 3 3 2 1427 1 3 3 2 1381 1 3 3 2 1394 1 3 3 2 1473 1 3 3 2 133 1 3 3 2 1446 1 3 3 2 1424 1 3 3 2 133 1 3 3 2 1452 1 3 3 2 1512 1 3 3 2 1314 1 3 3 2 1525 1 3 3 2 1475 1 3 3 2 1542 1 3 3 2 1461 2 3 3 2 1443 2 3 3 2 1547 2 3 3 2 1394 2 3 3 2 1364 2 3 3 2 1395 2 3 3 2 1522 2 3 3 2 1483 2 3 3 2 1452 2 3 3 2 1482 2 3 3 2 1439 2 3 3 2 1458 2 3 3 2 155 2 3 3 2 1502 2 3 3 2 152 2 3 3 2 END

Table C-4 Example OpenBUGS Certification Summary for Component C

mean sd MC_error val25pc median val975pc start sample b 09626 01701 0002428 06285 09633 1293 100001 100000 mu[1] 1399 01566 0002413 1369 1399 143 100001 100000 mu[2] 1491 005701 2386E-4 148 1491 1503 100001 100000 mu[3] 1449 0132 0001181 1423 1449 1475 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

64

34 Results Table C-5 lists the results in CopiesnL for the three components

Table C-5 Parameter Values for Components A B and C

Posterior Distribution Summary Statistics Component Parameter Mean SD 25 50 975

A

β 260 029 203 260 318 μ1 1415 027 1362 1415 1467 μ2 1559 010 1539 1559 158 μ3 1533 016 1501 1533 1565

B

β 095 034 029 095 162 μ1 1762 031 1701 1762 1822 μ2 1748 010 1728 1748 1769 μ3 1735 024 1687 1735 1783

C

β 097 017 064 097 130 μ1 1399 016 1368 1399 1429 μ2 1491 008 1476 1491 1507 μ3 1449 014 1421 1449 1477

Since in each case the three data set types represent independent measurements the three estimates can be combined to obtain a consensus value In this way a between data set type variability is included in the final standard uncertainty Table C-6 lists the results in Copies per Nanoliter obtained using the NIST Consensus Builder (NICOB)C3

Table C-6 Certified Values for Components A B and C

Component Mean u(Mean) Uk=2(Mean) A 151 07 14 B 175 03 06 C 145 05 10

4 Mito-to-nuclear DNA ratios (mtDNAnDNA)

41 Data Measurements were also obtained on the same samples using three mitochondrial detection assays mtND1 mtBatz and mtKav and a single nDNA assay (genomic) to compute the ratios mtDNAnDNA For the complete mtDNAnDNA data see

Table 16 mtDNAnDNA Ratios

42 Model For each component the data was analyzed using the following random effects model to account for between assay variability

119903119903119894119894~119873119873120572120572119894119894 1205901205901198861198861198861198861198861198861198861198861198861198862

where 119894119894 denotes assay (for A and B 1 = mtND1 2 = mtBatz 3 = mtKav for C 1 = mtBatz 2 = mtKav) and 120572120572119894119894~119873119873micro120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119886119886119886119886119886119886119886119886119886119886

2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

65

43 OpenBUGS code inits and data Two versions of the same model are required one for the results from the three mtDNA assays available for components A and B mtND1 mtBatz and mtKav and the other for the results from the two mtDNA assays available for component C mtBatz and mtKav mtDNAnDNA code for components A and B ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 13)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for components A and B list(sigt=c(111)) Data for Component A list(Ndata=102) set[] rep[] asy[] rat[] 1 1 1 1931 1 2 1 187 1 1 2 1845 1 2 2 1764 2 1 1 181 2 2 1 1906 2 1 2 1785 2 2 2 1837 3 1 1 1743 3 2 1 1761 3 1 2 1829 3 2 2 1847 4 1 1 1716 4 2 1 180 4 1 2 174 4 2 2 1813 5 1 1 1923 5 2 1 1835 5 1 2 1935 5 2 2 1851 6 1 1 1757 6 2 1 1724 6 3 1 1768 6 1 2 1743 6 2 2 1671 6 3 2 1766 6 1 3 1692 6 2 3 1662 6 3 3 1767 7 1 1 1924 7 2 1 1944 7 3 1 1893 7 1 2 1786

7 2 2 1752 7 3 2 1743 7 1 3 153 7 2 3 1542 7 3 3 1523 7 1 2 158 7 2 2 1557 7 3 2 1592 7 1 3 2103 7 2 3 210 7 3 3 2049 7 1 1 1749 7 2 1 1769 7 3 1 1777 7 1 2 1965 7 2 2 1889 7 3 2 1947 7 1 3 1813 7 2 3 1864 7 3 3 1864 8 1 1 178 8 2 1 1776 8 3 1 1706 8 1 2 1804 8 2 2 1751 8 3 2 1738 8 1 3 1769 8 2 3 1786 8 3 3 1761 9 1 2 1584 9 2 2 1544 10 1 1 1947 10 2 1 1928 10 3 1 1922 10 1 1 1829

10 2 1 1895 10 3 1 1851 10 1 1 1674 10 2 1 1617 11 1 2 1667 11 2 2 1813 11 3 2 1708 11 4 2 1598 11 1 2 1566 11 2 2 1687 11 3 2 1663 11 4 2 1589 11 1 2 1643 11 2 2 1668 11 3 2 1604 11 4 2 1573 12 1 2 1719 12 2 2 1716 12 3 2 167 12 4 2 1666 12 1 2 1581 12 2 2 1573 12 3 2 1601 12 4 2 1612 12 1 2 1546 12 2 2 1542 12 3 2 1538 12 4 2 1587 13 1 1 1617 13 2 1 1599 13 1 1 1622 13 2 1 1579 13 1 1 1575 13 2 1 1528 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

66

Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A

mean sd MC_error val25pc median val975pc start sample mu 1730 1443 004419 1703 1730 1759 100001 100000

data for Component B list(Ndata=87) set[] rep[] asy[] rat[] 1 1 1 2461 1 2 1 2401 1 1 2 2555 1 2 2 2608 2 1 1 2229 2 2 1 2200 2 1 2 2108 2 2 2 2192 3 1 1 2127 3 2 1 2198 3 1 2 2455 3 2 2 2506 4 1 1 2211 4 2 1 2199 4 1 2 2380 4 2 2 2249 5 1 1 2105 5 2 1 1943 5 1 2 2136 5 2 2 2119 6 1 1 1990 6 2 1 1921 6 3 1 1982 6 1 2 1990 6 2 2 1924 6 3 2 1958 6 1 3 1903 6 2 3 1831

6 3 3 1919 7 1 1 2304 7 2 1 2341 7 3 1 2332 7 1 2 2048 7 2 2 2078 7 3 2 2067 7 1 3 2399 7 2 3 2479 7 3 3 2493 7 1 1 2065 7 2 1 2166 7 3 1 2164 7 1 2 2314 7 2 2 2280 7 3 2 2258 7 1 3 1867 7 2 3 1987 7 3 3 1971 8 1 1 2046 8 2 1 2057 8 3 1 2024 8 1 2 1998 8 2 2 2043 8 3 2 1952 8 1 3 2077 8 2 3 2117 8 3 3 2041 9 1 2 1890 9 2 2 1886

10 1 2 1928 10 2 2 1916 10 3 2 1915 10 1 2 1947 10 2 2 1946 10 3 2 1961 10 4 2 1909 10 1 2 1953 10 2 2 1952 10 3 2 1941 10 4 2 1881 11 1 2 1865 11 2 2 1863 11 3 2 1828 11 4 2 1828 11 1 2 1944 11 2 2 1964 11 3 2 1947 11 4 2 1895 11 1 2 1908 11 2 2 1909 11 3 2 1940 11 4 2 1839 12 1 1 1688 12 2 1 1783 12 1 1 1823 12 2 1 1631 12 1 1 1920 12 2 1 1887 END

Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B

mean sd MC_error val25pc median val975pc start sample mu 2045 2246 0111 1999 2045 2088 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

67

mtDNAnDNA code for component C ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 12)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for component C list(sigt=c(11)) mtDNAnDNA data for component C list(Ndata=66) set[] rep[] asy[] rat[] 1 1 1 3201 1 2 1 3299 2 1 1 3347 2 2 1 3287 3 1 1 3217 3 2 1 3384 4 1 1 3200 4 2 1 3363 5 1 1 3294 5 2 1 3013 6 1 1 2805 6 2 1 2825 6 3 1 2829 6 1 2 2709 6 2 2 2738 6 3 2 2727 7 1 1 2797 7 2 1 2642 7 3 1 2792 7 1 2 2804 7 2 2 2789

7 3 2 2842 7 1 1 2715 7 2 1 2606 7 3 1 2687 7 1 2 2799 7 2 2 2628 7 3 2 2559 7 1 1 2639 7 2 1 2535 7 3 1 2657 7 1 2 2953 7 2 2 2757 7 3 2 2802 8 1 1 2927 8 2 1 2776 8 3 1 2762 8 1 2 3044 8 2 2 2830 8 3 2 2825 9 1 1 2501 9 2 1 2522 10 1 1 2832 10 2 1 2840

10 3 1 2788 10 4 1 2725 10 1 1 2835 10 2 1 2790 10 3 1 2736 10 4 1 2679 10 1 1 2803 10 2 1 2773 10 3 1 2757 10 4 1 2752 11 1 1 2599 11 2 1 2508 11 3 1 2551 11 4 1 2609 11 1 1 2617 11 2 1 2483 11 3 1 2626 11 4 1 2602 11 1 1 2665 11 2 1 2535 11 3 1 2570 11 4 1 2650 END

Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C

mean sd MC_error val25pc median val975pc start sample mu 2774 2367 01176 2729 2774 2821 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

68

Table C-10 Results for mtDNAnDNA Posterior Distribution Summary Statistics

Component Mean Std Dev 25 50 975 A 1730 14 1703 1730 1758 B 2045 23 1994 2047 2087 C 2774 23 2729 2774 2821

5 ReferencesC1 Lunn DJ Spiegelhalter D Thomas A Best N Statistics in Medicine 2009283049ndash3082C2 Polson NG Scott JG Bayesian Analysis 20127887ndash902C3 NIST Consensus Builder Software and userrsquos guide freely available at

httpsconsensusnistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

69

Appendix D Performance in Commercial qPCR Chemistries The performance of the SRM 2372a materials in commercial qPCR chemistries was assessed using manufacturerrsquos recommended protocols Each panel in Figures D-1 to D-4 displays the average cycle threshold (Ct) as a function of DNA concentration for the 5-point dilutions of SRM 2372a components In all cases the external calibrant was SRM 2372 component A Error bars represent approximate 95 confidence intervals Table D-1 lists the estimated [DNA] for the three SRM 2372a components for all commercial assays evaluated This Table also lists the Component BComponent A (BA) and Component CComponent A (CA) ratios calculated from the [DNA] estimates

Figure D-1 Results for the Innogenomics Commercial qPCR assays

20

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantLong(IQ_h)

ABCStd

[A] = (384 plusmn 26) ngμL[B] = (441 plusmn 24) ngμL[C] = (516 plusmn 24) ngμL

Rsup2 = 0999 se(Ct) = 01016

17

18

19

20

21

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantShort(IQ_d)

ABCStd

[A] = (435 plusmn 28) ngμL[B] = (543 plusmn 12) ngμL[C] = (510 plusmn 28) ngμL

Rsup2 = 0999 se(Ct) = 007

24

25

26

27

28

29

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYLong(IQHY_h)

ABCStd

[A] = (387 plusmn 24) ngμL[B] = (448 plusmn 21) ngμL[C] = (559 plusmn 31) ngμL

Rsup2 = 0998 se(Ct) = 01314

16

18

20

22

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYShort(IQHY_d)

ABCStd

[A] = (353 plusmn 23) ngμL[B] = (267 plusmn 15) ngμL[C] = (390 plusmn 18) ngμL

Rsup2 = 1000 se(Ct) = 00420

22

24

26

28

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYY(IQHY_y)

ABCStd

[A] = (572 plusmn 13) ngμL[B] = 0 ngμL[C] = (137 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

70

Figure D-2 Results for the Thermo Fisher Commercial qPCR assays

21

22

23

24

25

26

27

28

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HumanHuman(QFH_h)

ABCStd

[A] = (464 plusmn 13) ngμL[B] = (529 plusmn 19) ngμL[C] = (472 plusmn 22) ngμL

Rsup2 = 1000 se(Ct) = 004

22

23

24

25

26

27

28

29

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoHuman(QFD_h)

ABCStd

[A] = (701 plusmn 53) ngμL[B] = (561 plusmn 11) ngμL[C] = (472 plusmn 11) ngμL

Rsup2 = 0994 se(Ct) = 01922

24

26

28

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoMale(QFD_y)

ABCStd

[A] = (706 plusmn 49) ngμL[B] = 0 ngμL[C] = (122 plusmn 02) ngμL

Rsup2 = 0999 se(Ct) = 007

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPLarge(QFHP_h)

ABCStd

[A] = (340 plusmn 18) ngμL[B] = (432 plusmn 05) ngμL[C] = (342 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 00720

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPSmall(QFHP_d)

ABCStd

[A] = (434 plusmn 15) ngμL[B] = (514 plusmn 10) ngμL[C] = (466 plusmn 07) ngμL

Rsup2 = 1000 se(Ct) = 005

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioLarge(QFT_h)

ABCStd

[A] = (362 plusmn 23) ngμL[B] = (430 plusmn 06) ngμL[C] = (378 plusmn 07) ngμL

Rsup2 = 0999 se(Ct) = 00920

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioSmall(QFT_d)

ABCStd

[A] = (468 plusmn 08) ngμL[B] = (505 plusmn 08) ngμL[C] = (523 plusmn 14) ngμL

Rsup2 = 1000 se(Ct) = 00318

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioY(QFT_y)

ABCStd

[A] = (372 plusmn 11) ngμL[B] = 0 ngμL[C] = (110 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 003

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

71

Figure D-3 Results for the Qiagen Commercial qPCR assays

Figure D-4 Results for the Promega Commercial qPCR assays

19

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresAutosomal(QPH_h)

ABCStd

[A] = (554 plusmn 21) ngμL[B] = (479 plusmn 05) ngμL[C] = (381 plusmn 09) ngμL

Rsup2 = 1000 se(Ct) = 00418

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresY(QPH_y)

ABCStd

[A] = (358 plusmn 10) ngμL[B] = 0 ngμL[C] = (118 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 007

16

17

18

19

20

21

22

23

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProHuman(QPP_h)

ABCStd

[A] = (525 plusmn 49) ngμL[B] = (480 plusmn 42) ngμL[C] = (467 plusmn 45) ngμL

Rsup2 = 0999 se(Ct) = 01019

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProDegradation(QPP_d)

ABCStd

[A] = (567 plusmn 18) ngμL[B] = (558 plusmn 08) ngμL[C] = (484 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 00410

15

20

25

30

35

-50 -40 -30 -20 -10 00 10

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProY(QPP_y)

ABCStd

[A] = (538 plusmn 27) ngμL[B] = (00 plusmn 00) ngμL[C] = (162 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 003

16

18

20

22

24

-05 00 05 10 15 20

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorAutosomal(P_h)

ABCStd

[A] = (759 plusmn 64) ngμL[B] = (775 plusmn 24) ngμL[C] = (1056 plusmn 17) ngμL

Rsup2 = 0995 se(Ct) = 01818

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorY(P_y)

ABCStd

[A] = (596 plusmn 51) ngμL[B] = 0 ngμL[C] = (117 plusmn 01) ngμL

Rsup2 = 0994 se(Ct) = 020

18

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantAutosomal(PQ_h)

ABCStd

[A] = (436 plusmn 13) ngμL[B] = (606 plusmn 20) ngμL[C] = (504 plusmn 10) ngμL

Rsup2 = 1000 se(Ct) = 00218

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantDegradation(PQ_d)

ABCStd

[A] = (420 plusmn 18) ngμL[B] = (529 plusmn 16) ngμL[C] = (530 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 00516

18

20

22

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantY(PQ_y)

ABCStd

[A] = (410 plusmn 20) ngμL[B] = 0 ngμL[C] = (139 plusmn 05) ngμL

Rsup2 = 0999 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

72

Table D-1 Summary [DNA] Results for the Commercial qPCR Assays Estimated [DNA] ngμL Ratios

A B C BA CA Assaya xb u(x)c xb u(x)c xb u(x)c yd u(y)e yd u(y)e

IQd 435 28 543 12 510 28 125 008 117 010 IQh 384 26 441 24 516 24 115 010 134 011

IQHYd 353 23 267 15 390 18 076 006 111 009 IQHYh 387 24 448 21 559 31 116 009 144 012

Ph 759 64 775 24 1056 16 102 009 139 012 PQd 420 18 529 16 530 06 126 007 126 006 PQh 436 13 606 20 504 10 139 006 116 004

QFDh 701 53 561 11 472 11 080 006 067 005 QFHh 464 13 529 19 472 22 114 005 102 006

QFHPd 434 15 514 10 466 07 118 005 107 004 QFHPh 340 18 432 05 342 03 127 007 101 005

QFTd 468 08 505 08 523 14 108 003 112 004 QFTh 362 23 430 06 378 07 118 008 104 007 QPHh 554 21 479 05 380 09 086 003 069 003 QPPd 567 18 558 08 484 11 098 003 085 003 QPPh 525 49 480 42 467 45 092 012 089 012

IQHYy 572 13 NDf 137 06

NRg

024 001 Py 596 51 NDf 117 01 020 002

PQy 410 20 NDf 139 05 034 002 QFDy 706 49 NDf 122 02 017 001 QFTy 372 11 NDf 110 06 030 002 QPHy 358 10 NDf 118 03 033 001 QPPy 538 27 0002 0001 162 11 030 003

a Assay codes are listed in Table 8 b Mean of dilution-adjusted results from five dilutions of the SRM 2372a component c Standard uncertainty of the mean estimated from the stdanrd deviation of the five dilutions d Ratio of the estimated [DNA] of component B or C relative to that of component A e Standard uncertainty of the ratio estimated from the standard uncertainties of the individual components f Not detected Component B was prepared using tissue from only female donors it is not expected to

respond to male-specific assays g Not relevant

  • Abstract
  • Keywords
  • Technical Information Contact for this SRM
  • Table of Contents
  • Table of Tables
  • Table of Figures
  • Purpose and Description
  • Warning SRM 2372a is a Human Source Material
  • Storage and Use
  • History and Background
    • Figure 1 Sales History of SRM 2372
      • Experimental Methods
        • Polymerase Chain Reaction Assays
          • Table 1 NIST-Developed Human Nuclear DNA Assays
          • Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays
          • Table 2 NIST-Optimized Human Mitochondrial DNA Assays
          • Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays
            • Droplet Digital Polymerase Chain Reaction Measurements
              • Nuclear DNA (nDNA)
                • Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays
                  • Mitochondrial DNA (mtDNA)
                    • Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays
                    • Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays
                        • Sample Preparation
                        • Absorbance Measurements
                        • SRM Unit Production
                        • Component Homogeneity
                          • Figure 4 Plate Layout Design for Homogeneity Measurements
                            • Component Stability
                              • Figure 5 Plate Layout Design for Stability Measurements
                                • Component Certification Measurements
                                  • Table 6 Tubes used for Certification Measurements
                                  • Figure 6 Basic Plate Layout Design for Certification Measurements
                                    • Mitochondrial DNA Measurements
                                      • Table 7 Run Parameters for Mitochondrial DNA
                                        • Performance in Commercial qPCR Chemistries
                                          • Table 8 Commercial qPCR Chemistries Tested with SRM 2372a
                                          • Figure 7 Plate Layout Design for Performance Assessments
                                              • Measurement Results
                                                • Absorbance Spectrophotometry
                                                  • Table 9 Absorbance at Selected Wavelengths
                                                  • Figure 8 Absorbance Spectra of SRM 2372a Components A B and C
                                                    • Droplet Volume
                                                    • Certification Measurements
                                                      • Homogeneity Results
                                                        • Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component
                                                        • Table 11 Within- and Between-Tube Relative Standard Deviations
                                                          • Stability Results
                                                            • Figure 9 ddPCR Stability Measurements as Function of Time
                                                            • Table 12 Stability Data as λacute Copy Number per Nanoliter of Component
                                                              • Ten-Assay Certification Results
                                                                • Figure 10 Comparison of Assay Performance
                                                                • Table 13 Certification Data as λacute Copy Number per Nanoliter of Component
                                                                  • Estimation of Copy Number Concentration
                                                                    • Table 14 Measured Copy Number Concentrations
                                                                      • Conversion of Copy Number to Mass Concentration
                                                                        • Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution
                                                                            • Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)
                                                                              • Figure 11 mtDNAnDNA for Components A B and C
                                                                              • Table 16 mtDNAnDNA Ratios
                                                                              • Table 17 mtDNAnDNA Summary Results
                                                                                • Performance in Commercial qPCR Chemistries
                                                                                  • Figure 12 qPCR Autosomal Quantitation Performance Summary
                                                                                  • Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary
                                                                                      • Qualification of ddPCR as Fit-For-Purpose
                                                                                        • Figure 14 Absorbance Spectra of SRM 2372 Components A and B
                                                                                        • Digital Polymerase Chain Reaction Assays
                                                                                        • Heat Treatments (Melting of the DNA into Single Strands)
                                                                                        • Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a
                                                                                          • Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a
                                                                                            • Quantitative Evaluation of ssDNA Content
                                                                                              • ddPCR Measurement of ssDNA Proportion
                                                                                              • Proof of Principle
                                                                                                • Table 19 Comparison of Snap-Cooling to Slow-Cooling
                                                                                                  • Time at Denaturation Temperature
                                                                                                    • Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures
                                                                                                      • Temperature
                                                                                                        • Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a
                                                                                                          • DNA Concentration in the ddPCR Solution
                                                                                                            • Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution
                                                                                                              • Equivalence of Assays
                                                                                                                • Table 23 Comparison of Assays
                                                                                                                    • ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials
                                                                                                                      • Table 24 Evaluation of Limiting Ratio
                                                                                                                          • Conclusions and Recommendations
                                                                                                                            • Recommended Certification Values
                                                                                                                              • Table 25 Recommended Values and 95 Uncertainties for Certification
                                                                                                                                • Use of SRM 2372a for Value-Assigning In-House Reference Materials
                                                                                                                                  • Certificate of Analysis
                                                                                                                                  • References
                                                                                                                                  • Appendix A DNA Extraction Method
                                                                                                                                    • Procedure
                                                                                                                                    • Reagents
                                                                                                                                    • Reference
                                                                                                                                      • Appendix B Spectrophotometric Measurement Protocol
                                                                                                                                        • Spectrophotometer Calibration
                                                                                                                                          • Establish Baseline
                                                                                                                                          • Enable Verification of Wavelength Calibration
                                                                                                                                          • Enable Verification of Absorbance Calibration
                                                                                                                                            • For ldquonativerdquo dsDNA Samples
                                                                                                                                              • Verify Optical Balance of Cuvettes
                                                                                                                                                • For NaOH denatured ssDNA Samples
                                                                                                                                                  • Prepare a 04 molL NaOH Stock Solution
                                                                                                                                                  • Prepare the Sample Solutions
                                                                                                                                                  • Measure Sample Solutions
                                                                                                                                                    • After All Sample Measurements are Complete
                                                                                                                                                      • Enable Verification of Spectrophotometer Stability
                                                                                                                                                      • Clean Up
                                                                                                                                                        • Evaluation
                                                                                                                                                        • Reference
                                                                                                                                                          • Appendix C SRM 2372a Statisticianrsquos report
                                                                                                                                                            • 1 Introduction
                                                                                                                                                            • 11 OpenBUGS Analysis
                                                                                                                                                            • 12 OpenBUGS Exemplar Output
                                                                                                                                                            • 2 Homogeneity assessment
                                                                                                                                                            • 21 Data
                                                                                                                                                            • 22 Model
                                                                                                                                                            • 23 OpenBUGS code and data
                                                                                                                                                              • Table C-1 Example OpenBUGS Homogeneity Summary
                                                                                                                                                                • 24 Results
                                                                                                                                                                  • Figure C-1 Homogeneity Assessment
                                                                                                                                                                    • 3 Certification measurements
                                                                                                                                                                    • 31 Data
                                                                                                                                                                    • 32 Model
                                                                                                                                                                    • 33 OpenBUGS code inits and data
                                                                                                                                                                      • Table C-2 Example OpenBUGS Certification Summary for Component A
                                                                                                                                                                      • Table C-3 Example OpenBUGS Certification Summary for Component B
                                                                                                                                                                      • Table C-4 Example OpenBUGS Certification Summary for Component C
                                                                                                                                                                        • 34 Results
                                                                                                                                                                          • Table C-5 Parameter Values for Components A B and C
                                                                                                                                                                          • Table C-6 Certified Values for Components A B and C
                                                                                                                                                                            • 4 Mito-to-nuclear DNA ratios (mtDNAnDNA)
                                                                                                                                                                            • 41 Data
                                                                                                                                                                            • 42 Model
                                                                                                                                                                            • 43 OpenBUGS code inits and data
                                                                                                                                                                              • Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A
                                                                                                                                                                              • Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B
                                                                                                                                                                              • Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C
                                                                                                                                                                              • Table C-10 Results for mtDNAnDNA
                                                                                                                                                                                • 5 References
                                                                                                                                                                                  • Appendix D Performance in Commercial qPCR Chemistries
                                                                                                                                                                                    • Table D-1 Summary [DNA] Results for the Commercial qPCR Assays
Page 3: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .

Certain commercial entities equipment or materials may be identified in this document to describe an experimental procedure or concept adequately

Such identification is not intended to imply recommendation or endorsement by the National Institute of Standards and Technology nor is it intended to imply that the entities materials or equipment are necessarily the best available for the purpose

National Institute of Standards and Technology Special Publication 260-189 Natl Inst Stand Technol Spec Publ 260-189 83 pages (March 2018)

CODEN NSPUE2

This publication is available free of charge from httpsdoiorg106028NISTSP260-189

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

i

Abstract Standard Reference Material (SRM) 2372a is intended for use in the value assignment of human genomic deoxyribonucleic acid (DNA) forensic quantitation materials A unit of SRM 2372a consists of three well-characterized human genomic DNA materials solubilized in 10 mmolL 2-amino-2-(hydroxymethyl)-13 propanediol hydrochloride (Tris HCl) and 01 mmolL ethylenediaminetetraacetic acid disodium salt (disodium EDTA) pH 80 buffer (TE-4) The three component genomic DNA materials labeled A B and C are respectively derived from a single male donor a single female donor and 13 mixture of a male and a female donor A unit of the SRM consists of one 05 mL tube of each component each tube containing approximately 55 microL of DNA solution Each of these tubes is labeled and is sealed with a color-coded screw cap This publication documents the production analytical methods and statistical evaluations involved in production of this SRM

Keywords Human mitochondrial DNA (mtDNA)

Human nuclear DNA (nDNA) Standard Reference Material (SRM)

Technical Information Contact for this SRM Please address technical questions about this SRM to srmsnistgov where they will be assigned to the appropriate Technical Project Leader responsible for support of this material For sales and customer service inquiries please contact srminfonistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

ii

Table of Contents Purpose and Description 1 Warning SRM 2372a is a Human Source Material 1 Storage and Use 1 History and Background 2 Experimental Methods 4

Polymerase Chain Reaction Assays 4 Droplet Digital Polymerase Chain Reaction Measurements 7

Nuclear DNA (nDNA) 7 Mitochondrial DNA (mtDNA) 8

Sample Preparation 9 Absorbance Measurements 10 SRM Unit Production 10 Component Homogeneity 10 Component Stability 11 Component Certification Measurements 12 Mitochondrial DNA Measurements 13 Performance in Commercial qPCR Chemistries 13

Measurement Results 15 Absorbance Spectrophotometry 15 Droplet Volume 16 Certification Measurements 16

Homogeneity Results 16 Stability Results 18 Ten-Assay Certification Results 20 Estimation of Copy Number Concentration 22 Conversion of Copy Number to Mass Concentration 22

Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA) 23 Performance in Commercial qPCR Chemistries 25

Qualification of ddPCR as Fit-For-Purpose 27 Digital Polymerase Chain Reaction Assays 28 Heat Treatments (Melting of the DNA into Single Strands) 28 Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a 29 Quantitative Evaluation of ssDNA Content 30

ddPCR Measurement of ssDNA Proportion 30 Proof of Principle 31 Time at Denaturation Temperature 33 Temperature 35 DNA Concentration in the ddPCR Solution 36 Equivalence of Assays 38

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials 40 Conclusions and Recommendations 43

Recommended Certification Values 43 Use of SRM 2372a for Value-Assigning In-House Reference Materials 44

Certificate of Analysis 44 References 44

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

iii

Appendix A DNA Extraction Method 46 Procedure 46 Reagents 46 Reference 47

Appendix B Spectrophotometric Measurement Protocol 48 Spectrophotometer Calibration 48

Establish Baseline 48 Enable Verification of Wavelength Calibration 48 Enable Verification of Absorbance Calibration 48

For ldquonativerdquo dsDNA Samples 48 Verify Optical Balance of Cuvettes 48

For NaOH denatured ssDNA Samples 49 Prepare a 04 molL NaOH Stock Solution 49 Prepare the Sample Solutions 49 Measure Sample Solutions 50

After All Sample Measurements are Complete 50 Enable Verification of Spectrophotometer Stability 50 Clean Up 50

Evaluation 50 Appendix C SRM 2372a Statisticianrsquos report 51

1 Introduction 51 11 OpenBUGS Analysis 51 12 OpenBUGS Exemplar Output 51 2 Homogeneity assessment 52 21 Data 52 22 Model 52 23 OpenBUGS code and data 52 24 Results 55 3 Certification measurements 57 31 Data 57 32 Model 57 33 OpenBUGS code inits and data 58 34 Results 64 4 Mito-to-nuclear DNA ratios (mtDNAnDNA) 64 41 Data 64 42 Model 64 43 OpenBUGS code inits and data 65 5 References 68

Appendix D Performance in Commercial qPCR Chemistries 69

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

iv

Table of Tables Table 1 NIST-Developed Human Nuclear DNA Assays 5 Table 2 NIST-Optimized Human Mitochondrial DNA Assays 6 Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays 8 Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays 8 Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays 8 Table 6 Tubes used for Certification Measurements 12 Table 7 Run Parameters for Mitochondrial DNA 13 Table 8 Commercial qPCR Chemistries Tested with SRM 2372a 14 Table 9 Absorbance at Selected Wavelengths 15 Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component 17 Table 11 Within- and Between-Tube Relative Standard Deviations 18 Table 12 Stability Data as λacute Copy Number per Nanoliter of Component 19 Table 13 Certification Data as λacute Copy Number per Nanoliter of Component 21 Table 14 Measured Copy Number Concentrations 22 Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution 23 Table 16 mtDNAnDNA Ratios 24 Table 17 mtDNAnDNA Summary Results 25 Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a 29 Table 19 Comparison of Snap-Cooling to Slow-Cooling 32 Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures 34 Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 36 Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution 37 Table 23 Comparison of Assays 39 Table 24 Evaluation of Limiting Ratio 41 Table 25 Recommended Values and 95 Uncertainties for Certification 43 Table C-1 Example OpenBUGS Homogeneity Summary 55 Table C-2 Example OpenBUGS Certification Summary for Component A 60 Table C-3 Example OpenBUGS Certification Summary for Component B 61 Table C-4 Example OpenBUGS Certification Summary for Component C 63 Table C-5 Parameter Values for Components A B and C 64 Table C-6 Certified Values for Components A B and C 64 Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A 66 Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B 66 Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C 67 Table C-10 Results for mtDNAnDNA 68 Table D-1 Summary [DNA] Results for the Commercial qPCR Assays 72

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

i

Table of Figures Figure 1 Sales History of SRM 2372 3 Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays 6 Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays 7 Figure 4 Plate Layout Design for Homogeneity Measurements 11 Figure 5 Plate Layout Design for Stability Measurements 11 Figure 6 Basic Plate Layout Design for Certification Measurements 12 Figure 7 Plate Layout Design for Performance Assessments 14 Figure 8 Absorbance Spectra of SRM 2372a Components A B and C 16 Figure 9 ddPCR Stability Measurements as Function of Time 18 Figure 10 Comparison of Assay Performance 20 Figure 11 mtDNAnDNA for Components A B and C 24 Figure 12 qPCR Autosomal Quantitation Performance Summary 26 Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary 26 Figure 14 Absorbance Spectra of SRM 2372 Components A and B 27 Figure 15 Comparison of Snap-Cooling to Slow-Cooling 32 Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures 33 Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 35 Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution 37 Figure 19 Comparison of Assays 38 Figure 20 Evaluation of Limiting Ratio 40 Figure 21 Degrees of Equivalence 42 Figure C-1 Homogeneity Assessment 56 Figure D-1 Results for the Innogenomics Commercial qPCR assays 69 Figure D-2 Results for the Thermo Fisher Commercial qPCR assays 70 Figure D-3 Results for the Qiagen Commercial qPCR assays 70 Figure D-4 Results for the Promega Commercial qPCR assays 71

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

1

Purpose and Description Standard Reference Material (SRM) 2372a is intended for use in the value assignment of human genomic deoxyribonucleic acid (DNA) forensic quantitation materials A unit of SRM 2372a consists of three well-characterized human genomic DNA materials solubilized in 10 mmolL 2-amino-2-(hydroxymethyl)-13 propanediol hydrochloride (Tris HCl) and 01 mmolL ethylenediaminetetraacetic acid disodium salt (disodium EDTA) pH 80 buffer (TE-4) The three component genomic DNA materials labeled A B and C are respectively derived from a single male donor a single female donor and 13 mixture of a male and a female donor A unit of the SRM consists of one 05 mL tube of each component each tube containing approximately 55 microL of DNA solution Each of these tubes is labeled and is sealed with a color-coded screw cap NIST is guided by and adheres to the ethical principles set forth in the Belmont Report1 SRM 2372a was developed after an appropriate human subject research determination by the NIST Human Subjects Protections Office

Warning SRM 2372a is a Human Source Material The three SRM 2372a components are human genomic DNA extracts prepared at NIST from buffy coat white blood cells from single-source anonymous donors All buffy coats were purchased from Interstate Blood Bank (Memphis TN) The supplier reported that each donor unit used in the preparation of this product was tested by FDA-licensed tests and found to be negative for human immunodeficiency virus (HIV-1 RNA and Anti-HIV 12) hepatitis B (HBV DNA and HBsAg) hepatitis C (HCV RNA and Anti-HCV) West Nile virus (WNV RNA) and syphilis Additionally each donor was tested according to FDA guidelines for Trypanosoma cruzi (Chagas) and found to be negativenon-reactive However no known test method can offer complete assurance that infectious agents are absent from this material Accordingly this human blood-based product should be handled at the Biosafety Level 12 SRM 2372a components and derived solutions should be disposed of in accordance with local state and federal regulations

Storage and Use Until required for use SRM 2372a should be stored in the dark between 2 degC and 8 degC The SRM 2372a component tubes should be mixed briefly and centrifuged (without opening the tube cap) prior to removing sample aliquots for analysis For the certified values to be applicable materials should be withdrawn immediately after opening the tubes and processed without delay Certified values do not apply to any material remaining in recapped tubes The certification only applies to the initial use and the same results are not guaranteed if the remaining material is used later

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

2

History and Background SRM 2372a is the second member of the SRM 2372 Human DNA Quantitation Standard series The supply of the initial material SRM 2372 was exhausted in May 2017 SRM 2372a is derived from commercially obtained buffy coat cells from different donors than those used in SRM 2372 A series of interlaboratory studies coordinated by what was then the Biotechnology Division of the National Institute of Standards and Technology (NIST) during the 1990s demonstrated that the DNA quantitation technologies then in use by the forensic human identification community adequately met the needs of the DNA typing tools then available34 However our 2004 study demonstrated that the increasingly sensitive and informative polymerase chain reaction (PCR)-based multiplex typing tools required more reliable determinations of DNA input quantity for best performance5 Quantitative PCR (qPCR) methods for DNA quantitation require calibration to an external standard of defined DNA concentration ([DNA]) expressed in units of nanograms of DNA per microliter of extract solution The true [DNA] of a commercial secondary standard may vary from its nominal concentration depending on supplier and lot number To enable providers of secondary standards to more reliably evaluate their materials and forensic testing laboratories to validate their use of commercial quantitation methods and supplies in 2007 we delivered the human DNA calibration standard SRM 2372 Human DNA Quantitation Standard6 The quantity of DNA in each of the three components of SRM 2372 was originally assigned based on the absorbance of double-stranded DNA (dsDNA) at 260 nm7 However by 2012 the absorbance properties of the materials had increased beyond the certified uncertainty limits due to slow but progressive changes in DNA tertiary structure Sales of the SRM were restricted until the DNA quantities were reassigned based on measurements of single-stranded DNA (ssDNA) in early 20138-10 Figure 1 displays the sales history of this relatively high-selling SRM While SRM 2372 was fit for the practical purpose of harmonizing DNA quantitation measurements to an accessible and reproducible reference material the conversion of both dsDNA and ssDNA absorbance to mass [DNA] is through conventional proportionality constants that are not metrologically traceable to the modern reference system for measurements the International System of Units (SI) Responding to requests from the forensic and clinical communities for calibration standards that provide traceability to the SI we and others have established droplet digital polymerase chain reaction (ddPCR) as a potentially primary measurement process that can provide traceable results11-18 The copy numbers of human nuclear DNA (nDNA) in the reaction mixture per partition symbolized as λ were determined by ddPCR using ten PCR assays optimized for use with our ddPCR platform These assays target loci on eight chromosomes with three assays on widely separated locations on one chromosome (see Table 1)

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

3

Figure 1 Sales History of SRM 2372

Each small black dot represents the sale of one unit of SRM 2372 The red lines summarize the average rate of sales as originally certified (2007 to 2012) and after re-certification (2013 to 2017) The magenta labels state the average yearly sales rate during these two periods

The nDNA copy number concentration human haploid genome equivalents (HHGE) per nanoliter component solution and symbolized here as λacute is calculated as

copy number [DNA] = λ = λ(VF) HHGEnL

where V is the average droplet volume and F is the total dilution factor the volume fraction of the component solution in the ddPCR reaction mixture The manufacturer of the ddPCR platform used a Bio-Rad QX200 Droplet Digital PCR System (Hercules CA) system does not provide metrologically traceable droplet volume information This lack of information prevented metrologically traceable conversion of the number per partition measurements to the desired number per reaction volume units In consequence NIST developed a measurement method that provides metrologically traceable droplet volumes19 We show elsewhere that the mass [DNA] nanograms of human DNA per microliter component can be traceably estimated as

mass [DNA] = 3031timesλacute ngmicroL

where the constant is the product of the number of nucleotide basepairs (bp) in the human reference genome the average bp molecular weight Avogadrorsquos number of bp per mole and

2372

2372(Recertif ied)

050

010

00

1713 unitsyear

1416 unitsyear

2007 2009 2011 2013 2015 2017

Uni

ts S

old

Calendar Year

Human DNA Quantitation StandardSRM 2372

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

4

unit scaling factors18 The three components of SRM 2372a are certified for both copy number and mass [DNA] Mitochondrial DNA (mtDNA) analysis within the forensic community is useful when the nDNA may be degraded or is in low-to-undetectable quantity The challenges of mtDNA testing include contamination of the standard due to routine laboratory use degradation of plasmids or synthetic oligomers or improper commutability of the standard relative to casework samples Commonly cell line materials are used as standards for qPCR with human genomic DNA which may lead to an incorrect mtDNA-to-nDNA ratio (mtDNAnDNA) for unknown samples Non-certified values of mtDNAnDNA for the three components are provided to bridge the current gap between well characterized mitochondrial DNA quantification assays and the availability of commercial standards With the understanding that the forensic DNA community primarily uses qPCR as a method for DNA quantification the performance of the three SRM 2372a materials was evaluated in eleven commercial qPCR chemistries currently used employed within the forensic DNA community

Experimental Methods Production and value-assignment of any certified reference material (CRM) is a complex process In addition to acquiring processing and packaging the materials included in the CRM the materials must be sufficiently homogenous across the units stable over time within its packaging and well-characterized to deliver results that are fit-for-purpose within the intended measurement community20 The process has been particularly complicated for SRM 2372a because we needed to determine the metrological properties of the required ddPCR measurement systems before we could trust the results they provide The following sections describe the methods and processes used to produce and characterize the SRM 2372a materials including sample preparation SRM unit production PCR assays ddPCR measurement processes spectrophotometric absorbance measurements homogeneity testing stability testing ddPCR nDNA and mtDNAnDNA measurements and performance in commercial qPCR assessments Many of these processes were mutually dependent therefore the sections are presented in a ldquodefinitionalrdquo order that sequentially introduces terminology Polymerase Chain Reaction Assays

Ten PCR assays were developed for characterization of SRM 2372a Each assay was confirmed to target one locus per HHGE using the National Center for Biotechnology Informationrsquos BLASTblastn system21 Primers were purchased from Eurofins Operon (Huntsville AL) TaqMan probes labeled with 6-carboxyfluorescein (FAM) were purchased from Thermo Fisher (Waltham MA) FAM-labeled Blackhole Quencher and Blackhole Quencher+ probes were purchased from LGC Biosearch Technologies (Novato CA) Table 1 details the ten nDNA assays used to certify copy number and mass [DNA] of the SRM 2372a components FAM-labeled TaqMan probes were purchased from Thermo Fisher

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

5

and used in monoplex reactions VIC (a proprietary dye)-labeled probes were purchased from the same source and were used when nDNA and mtDNA were analyzed in the same reaction mixture Figure 2 displays exemplar droplet patterns for the assays All assays were developed at NIST and have been optimized for ddPCR18 Table 2 details three mitochondrial DNA (mtDNA) assays used to estimate mtDNAnDNA in the SRM 2372a components Figure 3 displays exemplar droplet patterns for the assays These assays were optimized for ddPCR from real-time quantitative PCR (qPCR) assays from the literature All primers probes were diluted with TE-4 buffer purchased from Affymetrix-USB Products (Affymetrix Inc Cleveland OH) to a nominal 5 micromol working solution for all single-plex assays (individual nDNA and mtDNA assays) For duplex assays (combined nDNA and mtDNA) primer and probe working solutions were diluted to a nominal 10 micromol

Table 1 NIST-Developed Human Nuclear DNA Assays

Assay Target

Chromosome Band Accession Primers and Probe a

Amplicon Length bp

NEIF Gene EIF5B

Chr 2 p111-q111 NC_00000212

F gccaaacttcagccttctcttc R ctctggcaacatttcacactaca PB+ tcatgcagttgtcagaagctg

67

2PR4 Gene RPS27A

Chr 2 p16 NC_00000212

F cgggtttgggttcaggtctt R tgctacaatgaaaacattcagaagtct PB tttgtctaccacttgcaaagctggccttt

97

POTP STR TPOX

Chr 2 p253 NC_00000212

F ccaccttcctctgcttcacttt R acatgggtttttgcctttgg PT caccaactgaaatatg

60

NR4Q Gene DCK

Chr 4 q133-q211 NC_00000412

F tggtgggaatgttcttcagatga R tcgactgagacaggcatatgtt PB+ tgtatgagaaacctgaacgatggt

83

D5 STR D5S2500

Chr 5 q112 NC_00000510

F ttcatacaggcaagcaatgcat R cttaaagggtaaatgtttgcagtaatagat PT ataatatcagggtaaacaggg

75

ND6 STR D6S474

Chr 6 q21-22 NC_00000612

F gcatggctgagtctaaagttcaaag R gcagcctcagggttctcaa PB+ cccagaaccaaggaagatggt

82

D9 STR D9S2157

Chr 9 q342 NC_00000912

F ggctttgctgggtactgctt R ggaccacagcacatcagtcact PT cagggcacatgaat

60

HBB1 Gene HBB

Chr 11 p155 NC_00001110

F gctgagggtttgaagtccaactc R ggtctaagtgatgacagccgtacct PT agccagtgccagaagagccaagga

76

ND14 STR D14S1434

Chr 14 q3213C NC_0000149

F tccaccactgggttctatagttc R ggctgggaagtcccacaatc PB+ tcagactgaatcacaccatcag

109

22C3 Gene PMM1

Chr 22 q132 NC_00002210

F cccctaagaggtctgttgtgttg R aggtctggtggcttctccaat PB caaatcacctgaggtcaaggccagaaca

78

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

6

a F Forward primer R Reverse primer PB Blackhole quencher probe (FAM labeled) PB+ Blackhole Plus quencher probe (FAM labeled) PT TaqMan MGB probe (FAM labeled)

Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays

Table 2 NIST-Optimized Human Mitochondrial DNA Assays

Assay Location Primers and Probe a Amplicon Length bp Ref

mtND1 34853504 F ccctaaaacccgccacatct

69 22 35333553 R gagcgatggtgagagctaaggt 35063522 PT ccatcaccctctacatc

mtBatz 84468473 F aatattaaacacaaactaccacctacct

79 23 85038524 R tggttctcagggtttgttataa 84758491 PT cctcaccaaagcccata

mtKav 1318813308 F ggcatcaaccaaccacaccta

105 24 1336913392 R attgttaaggttgtggatgatgga 1331013325 PT cattcctgcacatctg

a F Forward primer R Reverse primer PT TaqMan MGB probe (FAM labeled)

2PR4

POTP

NEI

F

R4Q

5

D5

ND6

D9

HBB1

ND1

4

22C3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

7

Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays

Droplet Digital Polymerase Chain Reaction Measurements

Nuclear DNA (nDNA) All nDNA ddPCR measurements used in this study followed the same measurement protocol

bull Droplets were generated on the Auto Droplet Generator using the Bio-Rad ldquoddPCR ldquoSupermix for Probes (No dUTP)rdquo (Bio-Rad cat 186-3024 Control 64079080) and ldquoDroplet Generation Oil for Probesrdquo (Bio-Rad cat 186-4110 Control 64046051) Note desoxyuridine 5-triphosphatase (dUTP) is an enzyme sometimes included in PCR reaction mixtures to minimize PCR carryover contamination

bull Table 3 details the composition of the ddPCR mastermix setup for a given assay and sample where ldquomastermix refers to reaction mixture of manufacturerrsquos proprietary PCR reagent ldquoSupermixrdquo forward and reverse PCR primers for a given assay probe for that assay the sample DNA and sometimes additional water to produce the desired value dilution factor F This setup was used for all assays and samples

bull Non-Template Controls (NTCs) for each assay were included in every analysis bull Droplets were thermal cycled on a ProFlex thermal cycler (Applied Biosystems) for

95 degC for 10 min followed by 60 cycles of 94 degC for 05 min and 61 degC for 1 min then 98 degC for 10 min before a 4 degC hold until the plate was removed

bull Droplets were read on the QX200 Droplet Reader and analysis was performed using the QuantaSoft Analysis Software version 1740917

mtND1 mtBatz mtKav

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

8

bull The numbers of positive and negative droplets at the end of 60 cycles were determined and were exported into a spreadsheet for further analysis Assay-specific intensity thresholds were determined by visual inspection for each assay for each measurement session

Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays

Concentration nDNA Mastermix Microliter

per Reaction 2X ldquoSupermix for Probes (no dUTPs)rdquo 125

5 μmolL Forward Primer 1875 5 μmolL Reverse Primer 1875 5 μmolL Probe (FAM) 125

PCR Grade Water 50 14 Diluted Sample 25 Total Volume 250

Mitochondrial DNA (mtDNA) All the mtDNA measurements used in this study followed the nDNA protocol described above but with different mastermix setups The duplex mtDNA measurements used the setup described in Table 4 The monoplex mtDNA measurements used the setup described in Table 5

Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 125

10 μmolL Forward Primer 09375 10 μmolL Reverse Primer 09375 10 μmolL Probe (VIC) 0625

PCR Grade Water 95 nDNA Mastermix from Table 3 05 Total Volume 250

Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 120

10 μmolL Forward Primer 09 10 μmolL Reverse Primer 09 10 μmolL Probe (FAM) 06

PCR Grade Water 72 Diluted Sample (See Table 7) 24

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

9

Total Volume 240 Sample Preparation

The buffy coat white blood cells used to prepare SRM 2372a materials were purchased from Interstate Blood Bank These materials were from four single-source anonymous individuals supplied to us labeled as Female1 Female2 Male1 and Male2 In our laboratories each donor buffy coat bag was aliquoted (5 mL per aliquot) into sterile 50 mL conical tubes and stored at 4 degC prior to extraction The modified salt-out manual extraction protocol described in Appendix A was performed for each aliquot with rehydration of the DNA in TE-4 buffer From 150 mL to 250 mL of solubilized DNA for each component was accumulated through the extraction of the 5 mL aliquots The DNA was stored in perfluoroalkoxy polymer (PFA) containers that had been bleach sterilized thoroughly rinsed with deionized water rinsed with ethanol and air-dried The minimum volume for production of 2000 units was 120 mL Preliminary estimates of the mass [DNA] of the re-solubilized components were obtained using a Nanodrop microvolume spectrophotometer Dilutions of the component materials were made until the absorbance at 260 nm was approximately 1 corresponding to a nominal mass [DNA] of 50 ngμL of dsDNA7 After addition of the TE-4 buffer the containers were placed on an orbital shaker and gently rotated for several hours prior to refrigerated storage at 4 ordmC The four single-source components were initially screened with the nDNA ddPCR measurement process using four nDNA assays 2PR4 NR4Q ND6 and 22C3 Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate The gravimetric 1rarr4 dilution was prepared by weighing an empty PFA vial and recording the weight then pipetting 30 microL TE-4 into the vial and recording the weight then pipetting 10 microL DNA and recording the final weight The dilution factor was calculated in a spreadsheet These screening measurements verified that the individual component materials were fit-for-purpose The four single-source components were evaluated with the nDNA ddPCR measurement process using the ten nDNA assays Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate Quantitative information was derived from the average of the ten nDNA assays to prepare the component C mixture Thirty milliliters of sample Male2 (estimated at 452 ngμL mass [DNA]) were combined with 80 mL of sample Female2 (estimated at 507 ngμL mass [DNA]) and 10 mL of TE-4 buffer to create a 1 male nDNA to 3 female nDNA producing a 14 malehuman mixture After addition of the TE-4 buffer the component C container was gently rotated on the orbital shaker for several hours prior to refrigerated storage at 4 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

10

Absorbance Measurements

In February 2017 absorbance spectra of the recently prepared fully diluted nominally dsDNA and ssDNA solutions of the three components were acquired with a BioCary 100 spectrophotometer using the measurement protocol described in Appendix B In December 2017 additional spectra were acquired for the remaining bulk solution of components A and B and for SRM 2372 components A and B These solutions had been stored at 4 degC in well-sealed PFA containers There was no remaining bulk component C solution SRM Unit Production

In the morning of the day a component was transferred into the sterile component tubes (Sarstedt Biosphere SC micro tube-05 mL Newton NC) used to package the SRM the container holding the diluted component was removed from the refrigerator and placed on a slow orbital shaker for 2 h The tubes were opened and placed under the laminar flow hood in rows of 5 within 80-hole tube racks for a total of 30 tubes per rack Tubes were placed in every second row of the racks to facilitate filling with an Eppendorf Repeater Xstream pipette (Eppendorf North America Inc Hauppauge NY) fitted with a 1-mL positive displacement tip set to dispense 55 microL per tube The component container was removed from the shaker a magnetic stir-bar was added the container was placed on a magnetic stir plate in the laminar flow hood and the material was stirred gently Filling proceeded until 2200 vials were filled with pipet tip replacement after filling every 200 tubes The filling process consisted of two individuals manually filling each vial within a biosafety cabinet (one physically pipetting each sample the other verifying that each tube was filled) and four individuals tightly closing the lids to each tube on a sterilized and covered bench When all tubes in a rack were filled the rack was moved out of the hood the tubes checked for proper filling their lids closed and the tubes transferred to 100-unit pre-labeled storage boxes As the 100-unit boxes were filled they were transferred to the labeling room where SRM labels were applied by hand and colored inserts were added to the lids red for component A white for component B and blue for component C After all tubes for a given component were labeled the 22 boxes were placed in a locked refrigerator Filling of each of the components took 60 min to 90 min All units were transferred to the refrigerator within 4 hours of beginning the filling process All units of the three materials were allowed to equilibrate for a minimum of two weeks prior to beginning homogeneity analysis Component Homogeneity

The ddPCR measurement protocol described above was used to assess the nDNA homogeneity of the three components One tube from the approximate center of each of the 22 storage boxes was assayed using the NEIF and ND6 assays (see Table 1) A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Three technical replicates were obtained for every tube A total

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

11

of six plates were used two plates per component one with the NEIF assay and one with the ND6 assay Figure 4 details the layout of the samples

Figure 4 Plate Layout Design for Homogeneity Measurements

Component Stability

Stability of the materials was evaluated at 4 degC room temperature (RT asymp 22 degC) and 37 degC Tubes from the homogeneity study were divided among the three temperatures for the stability study Additionally two never-opened tubes (one from box 8 and one from box 12 of each component) were held at each temperature Two tubes at each temperature for each component were opened sampled and resealed 4 times over the course of 8 weeks Measurements were made at week 1 week 3 week 5 and week 8 The measurement on week 8 was from an unopened tube (from box 8) for each of the components at all temperatures The data from the homogeneity study was used as timepoint 0 in the stability study anchored to 4 degC by the storage recommendations A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used for both the NEIF and ND6 assays All temperatures components and both assays fit onto one plate for timepoints 1 to 5 Twelve NTCs placed in rows G and H were included for each assay Figure 5 details the sample layout

Figure 5 Plate Layout Design for Stability Measurements

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

ND6 ND6 ND6

NEIFNEIFNEIFComponent A

Component A

Component B

Component B

Component C

Component C

1 2 3 4 5 6 7 8 9 10 11 12A A-7 A-7 B-7 B-7 C-7 C-7 A-7 A-7 B-7 B-7 C-7 C-7B A-10 A-10 B-10 B-10 C-10 C-10 A-10 A-10 B-10 B-10 C-10 C-10C A-9 A-9 B-9 B-9 C-9 C-9 A-9 A-9 B-9 B-9 C-9 C-9D A-12 A-12 B-12 B-12 C-12 C-12 A-12 A-12 B-12 B-12 C-12 C-12E A-8 A-8 B-8 B-8 C-8 C-8 A-8 A-8 B-8 B-8 C-8 C-8F A-11 A-11 B-11 B-11 C-11 C-11 A-11 A-11 B-11 B-11 C-11 C-11G NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTCH NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

NEIF

4C

RT

37C

ND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

12

Component Certification Measurements

After establishing homogeneity and stability measurements intended to be used to certify nDNA values and inform mtDNA values were started using randomly selected SRM tubes from the boxes listed in Table 7 The ten nDNA assays described in Table 1 and two of the three mtDNA assays described in Table 2 were run simultaneously on the same sample plate A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used in each of the six measurement campaigns

Table 6 Tubes used for Certification Measurements

Component Plate A B C

dd20170711 Box 1 Box 1 Box 1 dd20170712a Box 10 tube a Box 10 tube a Box 10 tube a dd20170712b Box 10 tube b Box 10 tube b Box 10 tube b dd20170713 Box 20 Box 20 Box 20 dd20170714 Box 5 Box 5 Box 5 dd20170801 Box 5 Box 5 Box 5

Every sample was analyzed in duplicate with every assay in all six plates Component A was always located in rows B and C component B in rows D and E component C in rows F and G and NTCs in rows A and H Figure 6 displays the basic layout of the ten nDNA and two mtDNA assays In addition to the plate column assignment scheme shown the three assay groupings were also ordered as Group 2 Group 3 Group 1 and Group 3 Group 1 Group 2 to address potential positional bias

Group 1 Group 2 Group 3 2PR4 POTP NEIF NR4Q D5 ND6 D9 HBB1 ND14 22C3 mtND1 mtBatz 1 2 3 4 5 6 7 8 9 10 11 12 A NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Comp A B Comp A C Comp B D Comp B E Comp C F Comp C G

H NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Figure 6 Basic Plate Layout Design for Certification Measurements

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

13

Mitochondrial DNA Measurements

A ldquoduplexed reactionrdquo was used to determine mtDNAnDNA where 1 μL of one of the nDNA mastermix was added to 49 μL of the mtDNA mastermix resulting in a 1rarr50 volumetric dilution of the nDNA The nDNA was targeted with a FAM-labeled probe while the mtDNA was targeted with a VIC labeled probe The primer and probe concentrations were increased from 5 micromol to 10 micromol in the duplexed reactions Additional testing of the mtDNA assays were performed in monoplex with FAM-labeled probes to determine the additional sources of bias between sampling assay setup and dilution factors Table 7 outlines the different methods examined to determine mtDNAnDNA for the three components

Table 7 Run Parameters for Mitochondrial DNA Droplet

Generation nDNA mtDNA

Plate Component Assay Dilution Reps Assay Dilution Reps dd20170824 A manualDG HBB1 1rarr4 3 mtND1 1rarr200 9 dd20170825 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170829 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170831 AB autoDG HBB1 1rarr3 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr6 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr12 2 mtND1 1rarr10rarr100 3 dd20170907 ABC manualDG HBB1 1rarr3 2 mtBatz 1rarr10rarr100 3 dd20170913 AB manualDG NEIF 1rarr4 3 mtND1 mtBatz mtKav 1rarr10rarr100 3 dd20170913 C manualDG NEIF 1rarr4 3 mtBatz mtKav 1rarr10rarr100 3 Performance in Commercial qPCR Chemistries

Table 8 lists the 11 qPCR chemistries that were commercially available in late 2017 All three of the SRM 2372a components were evaluated with these chemistries using component A of SRM 2372 as the reference standard For each component of SRM 2372a four tubes were pulled from box 22 and combined into one tube A 1rarr5 dilution of all samples was prepared followed by three serial 1rarr2 dilutions A five-point serial dilution of each SRM 2372a component was amplified in triplicate for each chemistry This consisted of the neat material the 1rarr5 dilution then the three serial 1rarr2 dilutions A six-point serial dilution of SRM 2372 Component A was also created for the standard curve for the commercial qPCR chemistries This consisted of the neat material followed by a 1rarr5 dilution then followed by four 1rarr2 dilutions serially The six-point serial dilution series of SRM 2372 Component A was also amplified in triplicate These commercial chemistries were amplified in individual plates on an Applied Biosystems 7500 HID Real-Time PCR Instrument (Foster City CA) following the manufacturersrsquo recommended protocols for reaction mix setup and cycling conditions Figure 7 details the plate layout for the qPCR plates Data was analyzed with the HID Real-Time PCR Analysis Software v12 and further analyzed in a spreadsheet

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

14

Table 8 Commercial qPCR Chemistries Tested with SRM 2372a

Manufacturer Commercial Kit Target Codea

Innogenomics

InnoQuant Long IQh Short IQd

InnoQuantHY Long IQHYh Short IQHYd Y IQHYy

Thermo Fisher Quantifiler Duo Human QFDh Male QFDy

Quantifiler HP Large QFHPh Small QFHPd

Quantifiler Human Human QFHh

Quantifiler Trio Large QFTh Small QFTd Y QFTy

Qiagen

Quantiplex Autosomal QPh

Quantiplex Hyres Autosomal QPHh Y QPHy

Quantiplex Pro Human QPPh Degradation QPPd Y QPPy

Promega

Plexor Autosomal Ph Y Py

PowerQuant Autosomal PQh Degradation PQd Y PQy

a Assays are coded with a short acronym derived from the kit name followed by the following codes for the type of target ldquohrdquo autosomal non-degradation described as ldquoAutosomalrdquo ldquoHumanrdquo or ldquoLongrdquo ldquodrdquo autosomal degradation described as ldquoDegradationrdquo or ldquoShortrdquo ldquoyrdquo Y-chromosome described as ldquoMalerdquo or ldquoYrdquo

1 2 3 4 5 6 7 8 9 10 11 12

A

B 2372-A_57 2372-A_57 2372-A_57 2372a-neat_A

2372a-neat_A

2372a-neat_A

2372a-neat_B

2372a-neat_B

2372a-neat_B

2372a-neat_C

2372a-neat_C

2372a-neat_C

C 2372-A_11

4 2372-A_11

4 2372-A_11

4 2372a-15_A

2372a-15_A

2372a-15_A

2372a-15_B

2372a-15_B

2372a-15_B

2372a-15_C

2372a-15_C

2372a-15_C

D 2372-A_5

7 2372-A_5

7 2372-A_5

7 2372a-110_A

2372a-110_A

2372a-110_A

2372a-110_B

2372a-110_B

2372a-110_B

2372a-110_C

2372a-110_C

2372a-110_C

E 2372-A_2

85 2372-A_2

85 2372-A_2

85 2372a-120_A

2372a-120_A

2372a-120_A

2372a-120_B

2372a-120_B

2372a-120_B

2372a-120_C

2372a-120_C

2372a-120_C

F 2372-A_1

425 2372-A_1

425 2372-A_1

425 2372a-140_A

2372a-140_A

2372a-140_A

2372a-140_B

2372a-140_B

2372a-140_B

2372a-140_C

2372a-140_C

2372a-140_C

G 2372-A_0

7125 2372-A_0

7125 2372-A_0

7125

H NTC NTC NTC

Figure 7 Plate Layout Design for Performance Assessments

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

15

Measurement Results After preparing the SRM 2372a bulk solutions their approximate 50 ngmicroL mass [DNA] were confirmed spectrophotometrically The bulk solutions were then packaged into individual tubes The materials contained in the tubes were characterized for homogeneity stability nDNA copy number [DNA] and mtDNAnDNA ratio The following sections detail the results of these measurements Absorbance Spectrophotometry

By widely accepted convention an aqueous DNA solution having an absorbance of 10 at 260 nm corresponds to a mass [DNA] of 50 ngmicroL for dsDNA and (37 to 40) ngmicroL for ssDNA7 Absorbance at (230 270 280 and 330) nm are traditionally used in the assessment of DNA quality725 Due to minor baseline shifts during acquisition of some spectra it was necessary to bias-adjusted all spectra to have absorbance of 00 at 330 nm Table 9 lists the resulting bias-adjusted values for selected wavelengths and the conventional mass [DNA] conversion of the absorbance at 260 nm The close agreement between the mass concentration values for the nominally dsDNA and the converted-to-ssDNA spectra confirm that the bulk solutions were correctly prepared to be asymp 50 ngμL dsDNA The slight increase in absorbance over the 10 months between acquiring the spectra suggests that like SRM 2372 the tertiary structure of the dsDNA in the SRM 2372a materials may be changing slowly with time Figure 8 displays the spectra

Table 9 Absorbance at Selected Wavelengths

Wavelength nm Form Date Component 230 260 270 280 ngμL dsDNA 2142017 A 0394 0931 0758 0488 466

B 0480 1112 0905 0583 556 C 0451 1006 0815 0524 503

ssDNA 2142017 A 0390 0614 0540 0316 454 B 0457 0723 0634 0372 535 C 0411 0641 0563 0330 474

dsDNA 12142017 A 0402 0956 0775 0503 478 B 0484 1128 0914 0589 564

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

16

Figure 8 Absorbance Spectra of SRM 2372a Components A B and C

dsDNA is denatured to ssDNA by diluting with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to DNA mass concentration

Droplet Volume

Droplet volumes for the two lots of ldquoSupermixrdquo used in the studies described in this report were determined using NISTrsquos Special Test Method 11050S-D as described in Dagata et al19 The lot used for most of the ddPCR analyses produced (07349 plusmn 00085) nL droplets where the ldquoplusmnrdquo value is the combined standard uncertainty The lot used for analyses performed after October 11 2017 produced (07376 plusmn 00085) nL droplets NIST measurements have demonstrated that droplet volumes are not influenced by the presence or absence of DNA in the sample solution Measurements repeated over time on an earlier lot of the Supermix product used in these studies suggest that droplet volumes remain constant over the shelf-life of the lot However droplet volumes for different lots of this product as well as for different Supermix products have been observed to differ by several percent We intend to document these measurements in a future report Other researchers have demonstrated that non-constant droplet volume distributions may contribute to ddPCR measurement imprecision and bias2627 However these effects become significant only with distributions wider than observed using Special Test Method 11050S-D19 and at copydroplet values larger than the λ le 05 copydroplet typically used in our measurements Certification Measurements

Homogeneity Results Table 10 lists the λacute nDNA copy number values adjusted for both droplet volume and sample dilution from the homogeneity measurements The data was analyzed using the three-level Gaussian hierarchical ANOVA model described in Appendix C Table 11 lists the values of the within- and between-tube variance components expressed as percent relative standard deviations At the 95 confidence level all components are homogeneous

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372a-A2141721417-NaOH121417

220 260 300 340

Wavelength nm

2372a-B2141721417-NaOH121517

220 260 300 340

Wavelength nm

2372a-C21417

21417-NaOH

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

17

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 NEIF A-1-01 1632 1529 1511 B-1-01 1790 1909 1885 C-1-01 1493 1385 1510 ND6 1616 1640 1582 NAe 1870 1860 1565 1562 1466

NEIF A-1-02 1531 1525 1549 B-1-02 1712 1719 1704 C-1-02 1516 1513 1577 ND6 1615 1589 1601 1775 1718 1756 1490 1547 1570

NEIF A-1-03 1524 1540 1533 B-1-03 1693 1755 1749 C-1-03 1451 1528 1506 ND6 1582 1152 1555 1824 1483 1896 1485 1505 1471

NEIF A-1-04 1454 1458 1427 B-1-04 1754 1745 1739 C-1-04 1440 1486 1520 ND6 1627 1605 1607 1670 1632 1814 1529 1508 1506

NEIF A-1-05 1465 1540 1504 B-1-05 1757 1687 1735 C-1-05 1523 1472 1483 ND6 1607 1633 1586 1683 1759 1703 1513 1520 1491

NEIF A-1-06 1512 1491 1491 B-1-06 1697 1709 1714 C-1-06 1437 1493 1522 ND6 1511 1483 1524 1693 1723 1722 1492 1481 1461

NEIF A-1-07 1418 1480 1455 B-1-07 1722 1625 1703 C-1-07 1485 1477 1463 ND6 1479 1557 1656 1690 1684 1688 1475 1547 1485

NEIF A-1-08 1496 1560 1501 B-1-08 1884 1802 1772 C-1-08 1433 1405 1478 ND6 1514 1615 1540 1864 1887 1821 1497 1501 1495

NEIF A-1-09 1588 1581 1637 B-1-09 1713 1734 1727 C-1-09 1463 1499 1505 ND6 1474 1438 1454 1709 1756 1758 1503 1491 1516

NEIF A-1-10 1574 1513 1570 B-1-10 1759 1809 1738 C-1-10 1534 1458 1454 ND6 1594 1713 1669 1764 1816 1787 1448 1465 1474

NEIF A-1-11 1557 1508 1633 B-1-11 1749 1782 1740 C-1-11 1533 1514 1555 ND6 1571 1611 1631 1741 1792 1784 1488 1470 1483

NEIF A-1-12 1500 1476 1515 B-1-12 1748 1712 1718 C-1-12 1450 1525 1497 ND6 1626 1586 1608 1797 1614 1790 1496 1483 1519

NEIF A-1-13 1672 1634 1804 B-1-13 1632 1668 1701 C-1-13 1537 1527 1515 ND6 1490 1516 1497 1626 1566 1681 1499 1479 1475

NEIF A-1-14 1551 1568 1538 B-1-14 1752 1535 1673 C-1-14 1515 1490 1494 ND6 1617 1614 1639 1724 1741 1685 1469 1480 1466

NEIF A-1-15b 1793 1722 1733 B-1-15 1871 1835 1887 C-1-15 1511 1568 1594 ND6 1818 1753 1762 1770 1752 1798 1515 1543 1565

NEIF A-1-16 1573 1520 1617 B-1-16 1688 1725 1743 C-1-16 1449 1452 1471 ND6 1634 1627 1575 1744 1697 1710 1543 1448 1491

NEIF A-1-17 1516 1516 1628 B-1-17 1789 1769 1825 C-1-17 1388 1421 1394 ND6 1613 1567 1567 1716 1762 1638 1452 NAe 1435

NEIF A-1-18 1612 1535 1672 B-1-18 1826 1780 1849 C-1-18 1475 1455 1445 ND6 1529 1495 1528 1805 1801 1931 1458 1531 1459

NEIF A-1-19 1598 1514 1596 B-1-19 1751 1759 1880 C-1-19 1468 1525 1446 ND6 1571 1535 1511 1827 1766 1786 1435 1478 1421

NEIF A-1-20 1575 1568 1586 B-1-20 1738 1678 1768 C-1-20 1549 1531 1499 ND6 1652 1545 1627 1752 1747 1768 1499 1465 1468

NEIF A-1-21 1658 1611 1578 B-1-21 1742 1836 1713 C-1-21 1530 1440 1484 ND6 1647 1607 1612 1746 1744 1764 1499 1454 NAe

NEIF A-1-22 1547 1528 1532 B-1-22 1761 1743 1660 C-1-22 1442 1487 1436 ND6 1537 1513 1536 1666 1692 1819 1589 1591 1556

NEIF A-1-15c 1456 1448 1523 ND6 1525 1402 1436

NEIF A-2-15d 1679 1599 1674 ND6 1581 1550 1584

a Tube coded as Component-Tube-Box b These results were identified as a probable error in the preparation or volumetric delivery of the sample

solution into the microfluidic device and were not used in the homogeneity evaluations c Repeat assessment of the solution in sample A-1-15 d Results for a second tube from box 15 analyzed at the same time as the repeat assessment of A-1-15 e Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

18

Table 11 Within- and Between-Tube Relative Standard Deviations

Within-Tube Component NEIF ND6 Between-Tube

A 28 40 33 B 27 38 25 C 23 18 19

Stability Results Table 12 lists the λacute volume- and dilution-adjusted nDNA copy number values from the stability measurements Figure 9 summarizes these results The results for the tubes held at 4 degC room temperature (RT asymp 22 degC) and 37 degC are very similar for all three components The absence of between-temperature differences identifies the between-week differences as attributable to variation in the measurement processes rather than sample instability These results indicate that the SRM 2372a genomic copy number is thermally stable from 4 degC to 37 degC over an extended period

Figure 9 ddPCR Stability Measurements as Function of Time

Each dot-and-bar symbol represents the mean λacute and the approximate 95 expanded uncertainty for all results for a given component at a given storage temperature The thick black horizontal lines represent an approximate 95 confidence interval on the λacute values estimated from the homogeneity data

13

14

15

16

17

18

19

0 1 2 3 4 5 6 7 8

λacute H

aplo

id n

DNA

Copi

es μL

Weeks

A

0 1 2 3 4 5 6 7 8

Weeks

B

4 degC 22 degC 37 degC

0 1 2 3 4 5 6 7 8

Weeks

C

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

19

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Week Temp Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3

1 4 degC NEIF A-10 1568 1441 B-10 1666 1653 C-10 1424 1394 ND6 A-10 1474 1494 B-10 1655 1705 C-10 1461 1394

NEIF A-07 1500 1463 B-07 1718 1640 C-07 1446 1452 ND6 A-07 1479 1479 B-07 1708 1686 C-07 1522 1482

RT (22 degC) NEIF A-12 1674 1608 B-12 1805 1777 C-12 1506 1495 ND6 A-12 1522 1507 B-12 1759 1788 C-12 1478 1461 NEIF A-09 1635 1539 B-09 1791 1793 C-09 1411 NAb ND6 A-09 1543 1519 B-09 1661 1682 C-09 1482 1496

37 degC NEIF A-11 1614 1637 B-11 1748 1744 C-11 1515 1464 ND6 A-11 1494 1502 B-11 1740 1697 C-11 1495 1535 NEIF A-08 1592 1614 B-08 1691 1725 C-08 1528 1500 ND6 A-08 1581 1595 B-08 1727 1699 C-08 1458 1483 3 4 degC NEIF A-10 1520 1706 B-10 NAb 1733 C-10 1427 1473 ND6 A-10 1543 1624 B-10 1670 1755 C-10 1443 1364 NEIF A-07 1537 1570 B-07 1549 1565 C-07 1424 1512 ND6 A-07 1517 1479 B-07 1705 1723 C-07 1483 1439

RT (22 degC) NEIF A-12 1521 1607 B-12 1817 1842 C-12 1566 1471 ND6 A-12 1644 1597 B-12 1845 1786 C-12 1504 1511 NEIF A-09 1432 1626 B-09 1825 1856 C-09 1392 1387 ND6 A-09 1539 1528 B-09 1734 1737 C-09 1535 1465

37 degC NEIF A-11 1651 1685 B-11 1759 1779 C-11 1437 1440 ND6 A-11 1680 1594 B-11 1806 1806 C-11 1471 1509 NEIF A-08 1580 1613 B-08 1740 1791 C-08 1582 1590 ND6 A-08 1592 1524 B-08 1769 1694 C-08 1468 1520 5 4 degC NEIF A-10 1486 1440 B-10 1626 1551 C-10 1381 1330 ND6 A-10 1555 1577 B-10 1673 1625 C-10 1547 1395 NEIF A-07 1387 1429 B-07 1723 1787 C-07 1330 1314 ND6 A-07 1538 1522 B-07 1862 1909 C-07 1452 1458

RT (22 degC) NEIF A-12 1601 1631 B-12 1718 1752 C-12 1546 1541 ND6 A-12 1699 1677 B-12 1772 1721 C-12 1490 1471 NEIF A-09 1381 1484 B-09 1675 1694 C-09 1371 1376 ND6 A-09 1625 1615 B-09 1904 1815 C-09 1468 1472

37 degC NEIF A-11 1456 1517 B-11 1720 1704 C-11 1505 1493 ND6 A-11 1650 1658 B-11 1148 1152 C-11 1527 1600 NEIF A-08 1590 1666 B-08 1745 1723 C-08 1469 1540 ND6 A-08 1235 1244 B-08 1809 1749 C-08 1508 1529 8 4 degC NEIF A-08a 1667 1632 1589 B-08a 1880 1884 1940 C-08a 1525 1475 1542 ND6 A-08a 1599 1593 1590 B-08a 1859 1897 1972 C-08a 1550 1502 1520

RT (22 degC) NEIF A-08b 1693 1666 1634 B-08b 1769 1884 1771 C-08b 1595 1557 1545 ND6 A-08b 1644 1620 1647 B-08b 1862 1738 1811 C-08b 1611 1585 1535

37 degC NEIF A-08c 1620 1576 1608 B-08c 1934 1952 1956 C-08c 1595 1600 1569 ND6 A-08c 1576 1553 1532 B-08c 1889 1844 1889 C-08c 1569 1514 1581

a Tube coded as Component-Tube-Box b Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

20

Ten-Assay Certification Results The original intention was to certify the copy number [DNA] of the three components with results from five independently prepared sets of samples analyzed over four days Each sample of the three components was from a new tube from different boxes and was evaluated with 10 nDNA assays The assay positions were alternated across these five ldquo10-assay certificationrdquo plates The λ results for these five plates are systematically somewhat larger than those for the homogeneity and stability data On review we realized that a different workspace and its suite of pipettes had been used to prepare the 1rarr10 dilution of the 1rarr4 diluted components into the reaction mastermix The ldquohighrdquo workstation was used for stability and homogeneity measurements while the ldquolowrdquo workstation was used for certification measurements Gravimetric testing of the pipettes established that the volumes delivered by these alternate pipettes was sufficiently biased to account for the observed λ differences A sixth ldquocertificationrdquo plate was prepared using the original (and now re-verified) pipettes The λ results or this plate agreed well with the homogeneity and stability data Table 13 lists the λacute volume- and dilution-adjusted nDNA copy number values for all six of the ten-assay datasets Figure 10 displays the extent of agreement among the ten assays While there are small-but-consistent differences between the assays there is only a 4 difference between the assays providing smallest and largest λ

Figure 10 Comparison of Assay Performance

The blue dot-and-bars represent the mean and its combined standard uncertainty for the A B and C components for each assay relative to the over-all grand mean for each component The black dot-and-bars represent combination of three component results for each assay The solid horizontal line represents the median difference of the assay means from the grand mean the dotted horizontal lines bound an approximate 95 confidence region about the median

2PR4POTPNEIFR4Q5 D5ND6 D9 HBB1ND14 22C3-10

-8

-6

-4

-2

0

2

4

6

8

10

Rela

tive

Diffe

renc

e fr

om M

ean

Assays

Component MeansAssay MeansMedianU95

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

21

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

7112017 2PR4 1667 1698 1869 1960 1630 1706 POTP 1655 1685 1850 1781 1596 1576 NEIF 1647 1665 1709 1707 1574 1538 R4Q5 1594 1658 1856 1787 1590 1555 D5 1655 1683 1848 1829 1610 1607 ND6 1629 1647 1828 1826 1589 1548 D9 1655 1652 1861 1858 1569 1572 HBB1 1644 1611 1825 1869 1565 1681 ND14 1645 1707 1871 1773 1564 1492 22C3 1545 1603 1776 1902 1548 1655

7122017 2PR4 1668 1739 2174 2231 1499 1497 POTP 1587 1702 2202 2292 1576 1611 NEIF 1678 1614 2172 2216 1545 1560 R4Q5 1628 1762 NA 2078 1584 1546 D5 1738 1649 2230 2192 1577 1529 ND6 1667 1678 2094 2034 1599 1478 D9 1695 1653 1949 2110 1587 1504 HBB1 1757 1750 1971 2123 1520 1551 ND14 1771 1764 2123 2066 1479 1438 22C3 1762 1759 2107 2046 1512 1494

7122017 2PR4 1638 1667 1810 1806 1452 1424 POTP 1676 1638 1865 1801 1347 1397 NEIF 1616 1624 1806 1786 1395 1350 R4Q5 1559 1587 1878 1768 1442 1459 D5 1608 1652 1824 1790 1448 1602 ND6 1611 1589 1768 1670 1463 1435 D9 1517 1607 1811 1749 1402 1433 HBB1 1557 1582 1763 1741 1474 1481 ND14 1604 1625 1865 1781 1466 1482 22C3 1675 1616 1815 1866 1439 1568

7132017 2PR4 1888 1939 1784 1821 1470 1444 POTP 1868 1815 1735 1813 1411 1427 NEIF 1953 1902 1778 1813 1512 1439 R4Q5 1779 1871 1835 1749 1452 1468 D5 1924 2124 1794 1848 1567 1524 ND6 1861 1787 1700 1711 1480 1424 D9 1824 1844 1799 1779 1442 1459 HBB1 1809 1799 1765 1769 1430 1467 ND14 1864 1857 1754 1725 1413 1411 22C3 1930 1942 1792 1814 1486 1480

7142017 2PR4 1597 1590 1718 1768 1485 1458 POTP 1533 1404 1753 1698 1422 1459 NEIF 1462 1533 1773 1792 1438 1465 R4Q5 1493 1519 1777 1740 1439 1468 D5 1538 1586 1803 1795 1515 1457 ND6 1492 1516 1718 1693 1409 1412 D9 1528 1538 1802 1743 1469 1474 HBB1 1548 1581 1834 1745 1453 1428 ND14 1513 1514 1731 1714 1465 1425 22C3 1562 NA 1842 NA 1363 1393

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

22

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

101217 2PR4 1431 1411 1791 1770 1357 1364 POTP 1407 1395 1746 1820 1447 1361 NEIF 1487 1398 1803 1803 1448 1404 R4Q5 1450 1439 1767 1739 1409 1411 D5 1394 1422 1745 1773 1389 1429 ND6 1396 1393 1824 1729 1415 1417 D9 1414 1390 1752 1744 1431 1380 HBB1 1401 1382 1813 1754 1381 1336 ND14 1412 1460 1648 1671 1403 1387 22C3 1435 1410 1818 1741 1435 1372

Estimation of Copy Number Concentration It was decided that all three sets of results (homogeneity stability and certification) would be used to derive the certified values with the λ of the first five ldquoten-assayrdquo plates adjusted by a data-driven estimate of the volumetric bias This bias was incorporated into the statistical model providing better estimates of the λ means while capturing all of the variance components (homogeneity stability between-assay and between data sets) as random effects using Bayesian MCMC programmed in OpenBUGS28 Table 14 lists the estimated values and their uncertainties See Appendix C for details

Table 14 Measured Copy Number Concentrations

Genomic Copies per Nanoliter of Solution Component Mean a u(Mean) b U95(Mean) c CV95

d A 151 07 14 93 B 175 03 06 34 C 145 05 10 69

a Mean the experimentally determined most probable estimate of the true value b u(Mean) the standard uncertainty of the Mean c Mean plusmnU95(Mean) the interval which is believed with about 95 confidence to contain the true value d CV95 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the counting unit one29 2) the validity of the Poisson endpoint transformation for digital PCR endpoint assays when applied to samples providing le 05 copies per droplet 3) the determination that the copies are dispersed as dsDNA and 4) calibrated mean droplet volume measurements made at NIST during sample dilution and mastermix preparation Conversion of Copy Number to Mass Concentration As described in Reference 18 metrologically valid conversion from copies per nanoliter to nanograms per microliter requires quantitative estimates for the number of nucleotide bases per reference human genome and the average nucleotide mass Using the values estimated in that report and asserting that each target copy corresponds to one HHGE the required transformation is

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

23

[DNA] ngmicroL

= [DNA] HHGE

nLtimes

3301 ngHHGE

The combined standard uncertainty is

119906119906 [DNA] ng

microL = 1199061199062 [DNA] HHGE

nL + 0531100

times[DNA] HHGE

nL 2

Both the measured and the conversion factor uncertainties are associated with ldquolargerdquo degrees of freedom Therefor the approximate 95 confidence uncertainty is estimated

11988011988095 [DNA] ng

microL = 2 times 119906119906

[DNA] ngmicroL

Table 15 lists the transformation of the Table 14 copy number values These values agree well with the conventional 260 nm absorbance values averaging (4 plusmn 5) larger

Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution

ngmicroL Component Value a u(Measurement)b u(Conversion)c U95(Value)d CV95

e A 498 23 03 47 93 B 578 10 03 21 36 C 479 17 03 33 70

a Estimated true value b Standard uncertainty component from the ddPCR measurements c Standard uncertainty estimated from the relative uncertainty of the conversion factor d Value plusmnU95(Value) is believed with about 95 confidence to contain the true value e 100U95(Value)Value the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the measured estimates of copies per nanoliter and 2) determinations made by us from internationally recognized literature resources of the average number of nucleotide bases per HHGE and the mean molecular weights of the bases18 Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)

While not intended to provide certified values mtDNAnDNA was measured as part of 13 experiments over 5 months Measurements were made using dilution-adjusted λ for the mtND1 mtBatz andor mtKav mitochondrial assays referenced against the dilution-adjusted λ of either the NEIF or HBB1 nDNA assays By chance both DNAs in the component C mixture have a binding site mutation that silences the mtND1 probe Figure 11 graphically summarizes the results by assay and component Table 16 lists the data Table 17 summarizes results from a random effects model that accounts for between-assay variability See Appendix C for details

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

24

Due to the limited number of mtDNA assays used in these measurements these results are metrologically traceable only to the assays used

Figure 11 mtDNAnDNA for Components A B and C

Each dot-and-bar symbol represents a mean mtDNAnDNA and its approximate 95 expanded uncertainty The dotted horizontal lines represent approximate 95 confidence intervals on the ratios

Table 16 mtDNAnDNA Ratios

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

20170711a 1 mtND1 1931 1870 2461 2401 1 mtBatz 1845 1764 2555 2608 3201 3299

20170711b 1 mtND1 1810 1906 2229 2200 1 mtBatz 1785 1837 2108 2192 3347 3287

20170712 1 mtND1 1743 1761 2127 2198 1 mtBatz 1829 1847 2455 2506 3217 3384

20170713 1 mtND1 1716 1800 2211 2199 1 mtBatz 1740 1813 2380 2249 3200 3363

20170714 1 mtND1 1923 1835 2105 1943 1 mtBatz 1935 1851 2136 2119 3294 3013

20170814 1 mtND1 1780 1776 1706 2046 2057 2024 1 mtBatz 1804 1751 1738 1998 2043 1952 2927 2776 2762 1 mtKav 1769 1786 1761 2077 2117 2041 3044 2830 2825

20170824 1 mtND1 1947 1928 1922 2 mtND1 1829 1895 1851 3 mtND1 1674 1617

20170825 1 mtBatz 1667 1813 1708 1598 1928 1916 1915 2832 2840 2788 2725 2 mtBatz 1566 1687 1663 1589 1947 1946 1961 1909 2835 2790 2736 2679 3 mtBatz 1643 1668 1604 1573 1953 1952 1941 1881 2803 2773 2757 2752

20170829 1 mtBatz 1719 1716 1670 1666 1865 1863 1828 1828 2599 2508 2551 2609 2 mtBatz 1581 1573 1601 1612 1944 1964 1947 1895 2617 2483 2626 2602 3 mtBatz 1546 1542 1538 1587 1908 1909 1940 1839 2665 2535 2570 2650

20170831 1 mtND1 1617 1599 1688 1783

ND1

ND1

ND1Ka

v

Kav

Kav

Batz

Batz

Batz

A

B

C

160

210

260

mtD

NAn

DNA

Ratio

Assay

Bind

ing s

ite m

utat

ion

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

25

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

2 mtND1 1622 1579 1823 1631 3 mtND1 1575 1528 1920 1887

20170907 1 mtBatz 1584 1544 1890 1886 2501 2522 20170913 1 mtND1 1757 1724 1768 1990 1921 1982

1 mtBatz 1743 1671 1766 1990 1924 1958 2805 2825 2829 1 mtKav 1692 1662 1767 1903 1831 1919 2709 2738 2727

20171016 1 mtND1 1924 1944 1893 2304 2341 2332 1 mtBatz 1786 1752 1743 2048 2078 2067 2797 2642 2792 1 mtKav 1530 1542 1523 2399 2479 2493 2804 2789 2842 2 mtBatz 1580 1557 1592 2715 2606 2687 2 mtKav 2103 2100 2049 2799 2628 2559 3 mtND1 1749 1769 1777 2065 2166 2164 3 mtBatz 1965 1889 1947 2314 2280 2258 2639 2535 2657 3 mtKav 1813 1864 1864 1867 1987 1971 2953 2757 2802

Table 17 mtDNAnDNA Summary Results

mtDNAnDNA Component Value a u(Value)b U95(Value)c CV95

d A 1737 16 38 17 B 2058 22 44 21 C 2791 23 47 17

a Estimated true value b Standard uncertainty c Value plusmnU95(Value) is believed with about 95 confidence to contain the true value d 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) Performance in Commercial qPCR Chemistries

Traditional qPCR assays only quantify relative to a reference standard Thus qPCR assays do not provide absolute [DNA] for the individual materials but they can provide [DNA] of one component relative to another Figure 12 summarizes the BA and CA ratios for the commercial autosomal (Human) targets from the eleven assays evaluated (see Table 8) The Y-chromosome (Male) assays cannot be summarized in this manner as the (all female) component B does not include a Y chromosome Figure 13 summaries the mass [DNA] results for the (all male) component A and (male-female mixture) component C Since this material was prepared as a nominal 13 malefemale mixture the expected malehuman (total) ratio is 14 sjc As displayed in these figures there is intrinsic variability between the SRM 2372a components and between the commercial qPCR chemistries examined This reflects the difference in the targets and amplification chemistries for each of the commercial kits Many of the commercial chemistries on the market to date contain multi-copy targets for both autosomal and Y-chromosomal quantification to increase the sensitivity of the assays

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

26

Figure 12 qPCR Autosomal Quantitation Performance Summary

Open circles denote the BA CA ratios the 17 commercial qPCR assays evaluated each labeled with the code listed in Table 8 The solid black circle represents ratios as estimated by ddPCR The red lines bound the approximate 95 confidence intervals on the ddPCR-estimated ratios

Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary

Open squares denote the ngμL values for SRM 2372a Components A (male) and C (mixture of male and female) for the seven commercial Y-chromosome qPCR assays evaluated each labeled with the code listed in Table 8 The dotted red lines represent malehuman ratios of 13 14 and 15

Appendix D provides a graphical analysis of all the data generated in the performance assessments for each of the targets amplified in the eleven commercial qPCR chemistries and the [DNA] values for all of the assays

dd

IQ_h

IQ_d

IQHY_h

IQHY_d

QFD_h

QFH_h QFHP_h

QFHP_dQFT_h

QFT_d

QPH_h

QPP_hQPP_d

P_h

PQ_h

PQ_d

06

08

10

12

14

07 09 11 13

CA

BA

IQHY_y

QFD_y

QFT_y

QPH_y

QPP_y

P_y

PQ_y

1413

15

10

12

14

16

18

30 40 50 60 70

[DNA

] C n

gμL

[DNA]A nguL

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

27

Qualification of ddPCR as Fit-For-Purpose We have documented the requirements for establishing metrological traceability of ddPCR human genomic assay results expressed as copies per nanoliter of sample in a series of three peer-reviewed papers151617 Through the experimental evidence they are based upon these papers demonstrate that ddPCR can accurately enumerate the number of independently migrating genomic copies in an aqueous extract Our most recent paper establishes the metrological traceability of these ddPCR results when transformed into the nanogram per microliter units used by the forensic and clinical communities18 However our previous studies have not addressed a fundamental requirement for correctly interpreting digital PCR measurements In broad simplification end-point assay systems count the number of independently partitioning entities regardless of the number of copies the entities contain With assays that target just one site per HHGE entities that migrate as ssDNA will yield counts that are twice the value had they migrated as dsDNA30 Should some originally dsDNA be converted over time to ssDNA ddPCR measurements of the number of independently migrating entities would increase although the number of HHGE would be unchanged Measurements of mass [DNA] provided by absorbance spectrophotometry and ddPCR reflect different properties absorbance at 260 nm is sensitive to the spatial configuration of the nucleotides as well as their number while ddPCR counts independently migrating entities The SRM 2372a materials have absorbance spectra shown in Figure 8 consistent with being predominantly dsDNA The SRM 2372 materials now have the absorbance spectra shown in Figure 14 which are consistent with them being entirely ssDNA

Figure 14 Absorbance Spectra of SRM 2372 Components A and B dsDNA is denatured to ssDNA by diluting the source material with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to mass [DNA]

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372-A200720122012-NaOH121417

0

4

6

0

4

6

220 260 300 340

Wavelength nm

2372-B200720122012-NaOH121517

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

28

To ensure the validity of ddPCR-based value assignments it is thus necessary to evaluate the nature of the partitioning entities Should a solution of what is thought to be dsDNA contain an appreciable fraction of ssDNA then the ddPCR results would require adjustment31 To avoid further delay in making SRM 2372a available to the forensic community the following sections detail a series of experiments that have enabled us to conclude with high confidence that the SRM 2372a components have a dsDNA conformation and that for this material ldquoentitiesrdquo can be considered dsDNA ldquocopiesrdquo The remaining stock of the SRM 2372 components also have a dsDNA conformation however the strands in the old materials are much less robustly connected than they are in the new materials Digital Polymerase Chain Reaction Assays

Three nDNA assays were used HBB1 NEIF and ND6 The HBB1 and NEIF assays reproducibly yield very similar results for DNA extracts from many different donors The HBB1 target is at the tip of the short (p) arm of chromosome 11 The NEIF target is in the centromere region of chromosome 2 The ND6 assay was selected for use because it typically yields the lowest results of the ten nDNA assays used to certify the SRM 2372a components (see Figure 10 below) The ND6 target is about halfway down the long (q) arm of chromosome 6 Heat Treatments (Melting of the DNA into Single Strands)

In all experiments the non-heated control and the heat-treated diluted solutions were stored in the same type of PCR reaction tube The starting point for the heat treatment was to determine if there was a difference between snap-cooling on ice and slower cooling at 25 degC Two tubes of the diluted material were placed in a GeneAmp 9700 thermal cycler (Applied Biosystems) and cycled up to 98 degC and held for 15 min After 15 min one tube was removed and placed on ice while the remaining tube was cooled at 1 degCs in the thermal cycler to 25 degC then placed on ice until prepared for ddPCR Additional heat treatments included cycling samples on a GeneAmp 9700 thermal cycler

a) from 98 degC for 5 min then cooling to 25 degC for 2 min During the 25 degC hold a randomly selected tube was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 45 min at 98 degC

b) from 94 degC for 1 min then cooling to 25 degC for 1 min During the 1 min hold at 25 degC a random sample was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 9 min at 94 degC

c) from 90 degC for 1 min then cooling to 25 degC for 1 min for a total of 20 cycles Samples were pulled at 25 degC after 1 min 3 min 5 min 10 min and 20 min and placed on ice

After the above experiments heat treatments were performed using a ProFlex thermal cycler This thermal cycler has six temperature zones that can be le 5 degC apart This enables heating

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

29

tubes at the same time but at different temperatures After treatment tubes were cooled at 5 degCs to 4 degC in the thermal cycler While designed for use with PCR plates a MicroAmp 96-well tray for the VeriFlex block enabled its use with single PCR reaction tubes Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a

As a first step in evaluating whether the DNA in the SRM materials disperse as dsDNA or ssDNA a direct comparison of the five materials was made using the HBB1 and NEIF assays Table 18 reports summary statistics for the ddPCR measurements reported as λ copiesdroplet and the estimated mass [DNA] values from the ddPCR results18 from the conventional dsDNA relationship7 and from the conventional ssDNA relationship after denaturing with 04 molL NaOH89 While the mass [DNA] values differ somewhat between the three methods there is no evidence for the SRM materials segregating exclusively or even mostly as ssDNA The extent of the hypochromic interactions among the nucleotides does not accurately reflect the extent to which the two strands of dsDNA have separated

Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a

λ copies per droplet λ copies per droplet ngmicroL Sample Assay Nrep

a Mean CVb ncmbc Mean CVb ddPCRd dsDNAe ssDNAf

2372-A HBB1 4 03645 19 8 03644 20 566g 524 57 2372-A NEIF 4 03644 23 2372-B HBB1 4 03513 23 8 03487 35 580h 536 61 2372-B NEIF 4 03461 47 2372a-A HBB1 4 02603 27 8 02649 30 474 465 454 2372a-A NEIF 4 02694 24 2372a-B HBB1 4 02908 34 8 02891 37 517 556 534 2372a-B NEIF 4 02873 45 2372a-C HBB1 4 02757 28 8 02711 28 485 502 474 2372a-C NEIF 4 02665 16

a Number of technical replicates for the given assay b CV = 100(standard deviation)mean c Combined number of technical replicates ignoring potential differences between the assays d Estimated from the ddPCR λ measurements listed in this Table18 e Estimated from the conventional relationship between the mass [DNA] of dsDNA and absorbance at

260 nm7 using absorbance spectra acquired shortly after solutions were produced f Estimated from the relationship between the mass [DNA] of ssDNA and absorbance at 260 nm after 11

dilution with 04 molL NaOH89 g Values adjusted to measured (132 plusmn19) volume loss of archived units due to evaporation h Values adjusted to measured (71 plusmn25) volume loss of archived due to evaporation

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

30

Quantitative Evaluation of ssDNA Content

Absorbance spectrophotometry is routinely used to quantitatively assess DNA tertiary structure3233 On storage ldquoat physiological pH the intrastrand repulsion between the negatively charged phosphate groups forces the double helix into more rigid rodlike conformationrdquo34 suggesting a mechanism for the observed slow changes in the SRM 2372 materials As demonstrated in Table 18 absorbance at 260 nm does not necessarily correlate with ddPCR behavior Yet ldquofurthermore the repulsion between phosphate groups on opposite strands tends to separate the complementary strandsrdquo34 Therefore a method that is independent of hypochromism is needed to adequately evaluate the proportion (if any) of ldquotruerdquo (completely disjoint) ssDNA in the SRM materials Circular dichroism spectroscopy is sensitive to the conformational structures of DNA in solution35 and different DNA configurations migrate differentially in agarose gel electrophoresis7 Both techniques have been used to qualitatively discriminate ssDNA from dsDNA in ldquopurerdquo single-sequence materials but neither technique appears to be able to quantify small proportions of ssDNA in a complex multi-chromosome genomic material Wilson and Ellison have proposed a real-time chamber digital PCR (cdPCR) technique that can estimate excess proportions of ssDNA copies from the crossing-threshold distribution31 However their method is insufficiently sensitive for low ssDNA proportions It is well known that heating DNA beyond the melting temperature separates mammalian dsDNA into ssDNA36 albeit we have shown using cdPCR that boiling dsDNA for some minutes also renders some target sites non-amplifiable15 In the following section we document a ddPCR-based approach for estimating the proportion if any of ssDNA in a sample that involves less destructive heating While the method is not yet optimized the results from these preliminary studies demonstrate that the SRM 2372 and 2372a components disperse into the ddPCR droplets as dsDNA ddPCR Measurement of ssDNA Proportion Given two otherwise identical aliquots of a given DNA material one subjected to some treatment and the other untreated the ratio between the average number of copies per droplet of treated material and that of its untreated sibling λtλu quantitatively estimates the change in the number of independently dispersing entities Should the treatment reduce the number of accessible amplifiable targets (AAT) λtλu will be less than 1 Reduction in AAT may result from damage to the target itself andor conformational changes to the fragment containing the target that render the target inaccessible or non-amplifiable16 Assuming no denaturation of dsDNA to ssDNA either because of an ineffective treatment or the material having started as ldquopurerdquo ssDNA if the treatment does not itself damage targets then λtλu will equal 1 Should treatment convert at least some dsDNA to ssDNA without damaging targets the number of AAT will increase and λtλu will be greater than 1 Assuming complete conversion of ldquopurerdquo dsDNA to ldquopurerdquo ssDNA without other damage then λtλu will equal 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

31

Values greater than 1 but less than 2 indicate some conversion of dsDNA to ssDNA but also some combination of (a) incomplete denaturation of dsDNA to ssDNA (b) partial renaturation of ssDNA to dsDNA after treatment but before ddPCR analysis (c) damage to targets during the process or ndash of greatest interest ndash (d) the original sample material was an impure mixture of dsDNA and ssDNA The following discussions graphically summarize sets of λtλu followed by tabular listings of the relevant experimental conditions ddPCR measurements and the numerical λtλu values The information in these tables includes

Plate Index Experiment identifier indicates date performed in YYYYMMDD format Material Sample material SRM 2372-A 2372-B 2372a-A 2372a-B or 2372a-C Assay ddPCR assay HBB1 NEIF or ND6 degC Denaturation temperature in degC Min Length of time the treated material was held at denaturation temperature in

minutes Cool Method used to cool treated material from denaturation temperature Snap

Slow R1 or R5 where ldquoSnaprdquo indicates cooling on ice ldquoSlowrdquo cooling at room temperature to 25 degC ldquoR1rdquo cooling in a thermal cycler to 25 degC at 1 degCs and ldquoR5rdquo cooling in a thermal cycler to 4 degC at 5 degCs

F Dilution factor expressed as (volume Material) per (final volume of diluted solution) 12 14 or 18

n Number of technical replicates Mean Mean of technical replicates CV Coefficient of Variation expressed as a percentage 100 Standard Deviation

Mean

λtλu the copies per droplet result of the treated solution λt divided by the copies per droplet result of its untreated sibling λu

u(λtλu) Standard uncertainty of λtλu 119906119906(120582120582t 120582120582ufrasl ) = 120582120582t 120582120582ufrasl CV(120582120582t)100

2

+ CV(120582120582u)100

2

Proof of Principle Figure 15 summarizes two replicate sets of measurements performed to assess whether the SRM 2372 and 2372a materials respond similarly whether the denatured materials renature if allowed to cool slowly and whether results are (reasonably) repeatable Table 19 details the experimental conditions The λtλu values for the four materials are similar all much greater than 1 but not approaching 2 Most of the ratios are on or close to the ldquoslow cooling is the same as snap coolingrdquo line suggesting that renaturation is not a significant concern While the replicate results for the SRM 2372-A and 2372a-A materials do not overlap as closely as those for SRM 2372-B and 2372a-B similar treatment conditions produce similar results

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

32

Figure 15 Comparison of Snap-Cooling to Slow-Cooling

Symbols mark the means of replicate measurements made using the same dilutions on successive days with bars representing plusmn one standard combined uncertainty Results are labeled by material ldquoArdquo and ldquoBrdquo for SRM 2372 components and ldquoaArdquo ldquoaBrdquo and ldquoaCrdquo for SRM 2372a components The diagonal black line represents equality between the snap- and slow-cooling treatments

Table 19 Comparison of Snap-Cooling to Slow-Cooling

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171218 2372-A HBB1 14 4 03830 29

dd20171218 2372-A HBB1 98 15 Slow 14 2 05644 18 1474 0051 dd20171218 2372-A HBB1 98 15 Snap 14 4 05852 15 1528 0051

dd20171218 2372a-A HBB1 14 4 03029 18

dd20171218 2372a-A HBB1 98 15 Slow 14 4 04430 21 1463 0040 dd20171218 2372a-A HBB1 98 15 Snap 14 4 04454 33 1471 0055

dd20171218 2372-B HBB1 14 4 03594 31

dd20171218 2372-B HBB1 98 15 Slow 14 4 05806 18 1615 0058 dd20171218 2372-B HBB1 98 15 Snap 14 4 05825 13 1620 0055

dd20171218 2372a-B HBB1 14 4 03383 15

dd20171218 2372a-B HBB1 98 15 Slow 14 4 05043 38 1491 0061 dd20171218 2372a-B HBB1 98 15 Snap 14 4 05254 20 1553 0039

dd20171219 2372-A HBB1 14 4 03357 23

dd20171219 2372-A HBB1 98 15 Slow 14 4 05340 16 1591 0046 dd20171219 2372-A HBB1 98 15 Snap 14 4 05340 12 1591 0042

A

aA

B

aB

A

aA

B

aB

14

15

16

17

14 15 16 17

λ tλ

u Sn

ap C

oole

d

λtλu Slow Cooled

Rep 1Rep 2Snap = Slow

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

33

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171219 2372a-A HBB1 14 4 02735 24

dd20171219 2372a-A HBB1 98 15 Slow 14 4 04151 03 1518 0037 dd20171219 2372a-A HBB1 98 15 Snap 14 4 03918 21 1433 0045 dd20171219 2372-B HBB1 14 4 03346 22

dd20171219 2372-B HBB1 98 15 Slow 14 4 05476 17 1637 0045 dd20171219 2372-B HBB1 98 15 Snap 14 4 05480 29 1638 0059 dd20171219 2372a-B HBB1 14 4 03158 19

dd20171219 2372a-B HBB1 98 15 Slow 14 4 04798 16 1519 0038 dd20171219 2372a-B HBB1 98 15 Snap 14 4 04856 17 1538 0039

Time at Denaturation Temperature Figure 16 summarizes λtλu for the SRM 2372-B material as a function of denaturation temperature and the cumulative time the solution is held at that temperature Table 20 details the experimental conditions

Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures

The symbols represent the ratio of the mean result of three technical replicates of the untreated and each of the treated solutions the bars represent combined standard uncertainties The lines denote regression fits to the function λtλu = atimese-bt where t is the cumulative time Values and the standard uncertainties of the coefficients are listed in the legend

12

14

16

18

20

0 5 10 15 20

λ t λ

u

Cummulative Treatment Time min

degC a u (a ) b u (b )90 1804 0039 00198 0000494 1788 0029 00099 0002998 1855 0019 00067 00019

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

34

For all three of the denaturation temperatures studied the relationship between λtλu and cumulative time t can be modeled as

λtλu = ae-bt

where a is the extrapolated value of λtλu at t = 0 and b is the rate of change in AAT per minute As expected the rate of loss due to the treatment increases with increasing treatment temperature and cumulative time at that temperature37 Given the much longer treatment time for the 98 degC series the agreement among the three t = 0 parameter (a) values is remarkable This suggests that the first 1 min of treatment efficiently separates dsDNA into ssDNA The pooled relative standard uncertainty of these regression parameter is 17 somewhat smaller than the 25 pooled relative standard uncertainties of the individual λtλu However the plusmnstandard uncertainty interval of the 1-min λtλu results for the 94 degC and 90 degC both overlap the extrapolated t = 0 value suggesting that reliable results can be obtained using a single treatment of short duration

Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171228 2372-B HBB1 14 3 03650 17 dd20171228 2372-B HBB1 98 5 R1 14 3 06157 12 1687 0035 dd20171228 2372-B HBB1 98 10 R1 14 3 05493 08 1505 0028 dd20171228 2372-B HBB1 98 15 R1 14 3 05019 13 1375 0029 dd20171228 2372-B HBB1 98 20 R1 14 3 04583 21 1256 0034 dd20171228 2372-B HBB1 98 25 R1 14 3 04154 32 1138 0041 dd20171228 2372-B HBB1 98 30 R1 14 3 03784 14 1037 0023 dd20171228 2372-B HBB1 98 35 R1 14 3 03408 10 0934 0019 dd20171228 2372-B HBB1 98 40 R1 14 3 02980 01 0816 0014 dd20171228 2372-B HBB1 98 40 R1 14 3 02815 08 0771 0015

dd20171229 2372-B HBB1 14 3 03708 27

dd20171229 2372-B HBB1 94 1 R1 14 3 06598 26 1779 0056 dd20171229 2372-B HBB1 94 2 R1 14 3 06553 32 1767 0063 dd20171229 2372-B HBB1 94 3 R1 14 3 06301 20 1699 0045 dd20171229 2372-B HBB1 94 4 R1 14 3 06252 14 1686 0037 dd20171229 2372-B HBB1 94 5 R1 14 3 06577 14 1773 0039 dd20171229 2372-B HBB1 94 6 R1 14 3 06156 07 1660 0030 dd20171229 2372-B HBB1 94 7 R1 14 3 06118 10 1650 0033 dd20171229 2372-B HBB1 94 8 R1 14 3 06256 12 1687 0035 dd20171229 2372-B HBB1 94 9 R1 14 3 05999 10 1618 0032

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

35

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180104 2372-B HBB1 14 4 03713 25 dd20180104 2372-B HBB1 90 1 R1 14 3 06916 22 1863 0052 dd20180104 2372-B HBB1 90 3 R1 14 3 06465 06 1741 0032 dd20180104 2372-B HBB1 90 5 R1 14 3 06431 20 1732 0045 dd20180104 2372-B HBB1 90 10 R1 14 3 06014 25 1620 0049 dd20180104 2372-B HBB1 90 15 R1 14 3 06024 10 1623 0032 dd20180104 2372-B HBB1 90 20 R1 14 3 06003 14 1617 0035

Temperature Figure 17 summarizes the effect of denaturation temperature using a single treatment of 05 min on the 2372-B and 2372a-B materials Table 21 details the experimental conditions

Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

The symbols represent the ratio of the mean result of four technical replicates of the untreated and three replicates of the treated solutions the bars represent combined standard uncertainties In all cases the duration of the treatment was 05 min The lines denote regression fits to the linear function λtλu = a +bT where Tis the denaturation temperature

The SRM 2372 and 2372a materials respond very differently to changes in denaturation temperature The older spectrophotometrically ssDNA material suffers somewhat greater loss in λtλu as temperature increases The new spectrophotometrically dsDNA material suffers much greater loss in λtλu as temperature decreases

14

15

16

17

18

19

20

82 86 90 94 98

λ tλ

u

Temperature degC

2372-B

2372a-B

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

36

Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180109 2372-B HBB1 14 4 03577 25 dd20180109 2372-B HBB1 100 05 R5 14 3 06314 16 1765 0051

dd20180104a 2372-B HBB1 14 4 03577 05

dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032 dd20180104a 2372-B HBB1 94 05 R5 14 3 06459 25 1806 0046 dd20180104a 2372-B HBB1 90 05 R5 14 2 06591 08 1842 0017 dd20180104a 2372-B HBB1 88 05 R5 14 3 06589 05 1842 0013 dd20180104a 2372-B HBB1 84 05 R5 14 3 06729 10 1881 0022 dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

We hypothesize that in the older SRM 2372 material the between-strand bond is relatively fragile and can be completely severed at relatively low temperature while avoiding much temperature-dependent AAT loss In contrast the between-strand bond in the SRM 2372a material requires higher temperature to completely disassociate the strands This suggests that there is not a sample-independent ldquooptimumrdquo combination of denaturation temperature and treatment duration However 98 degC for 05 min appears to provide comparable results for these two materials DNA Concentration in the ddPCR Solution Figure 18 summarizes the effect of [DNA] in the ddPCR solution with SRM 2372-B at a denaturation temperature of 98 degC and treatment duration of 05 min Table 22 details the experimental conditions

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

37

Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution

The solid black symbols represent the ratio of the mean result of four technical replicates of the untreated and treated solutions with the λtλu values shown on the left-hand vertical axis the bars represent combined standard uncertainties The solid blue circles represent λu and the open red circles λt with values shown on the right-hand vertical axis at this graphical scale the standard deviations of the technical replicates are covered by the symbol The diagonal blue and red lines represent the log-linear relationships between just the 18 and 14 λ values

The primary effect of [DNA] appears to be the loss of linearity in λt with copiesdroplet values greater than about 1 copiesdroplet There may be some advantage to use somewhat greater dilutions than the routine 14 however the small increase in λtλu may be offset by the increased variability with small λu This suggests the use of dilutions that give λu values of (01 to 04) copiesdroplet

Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180117 2372-B HBB1 12 4 07127 11 dd20180117 2372-B HBB1 98 05 R5 12 4 12331 14 1730 0031

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 3 06151 15 1843 0040

dd20180117 2372-B HBB1 18 4 01549 18 dd20180117 2372-B HBB1 98 05 R5 18 4 02897 20 1870 0051

12 14 18

170

175

180

185

190

02

04

06

08

101214

λ T

arge

t D

ropl

et

λ tλ

u

Dilution Factor

λ₊λᵤλᵤλ₊

λtλu

λu

λt

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

38

Equivalence of Assays Figure 19 summarizes the agreement between the HBB1 NEIF and ND6 assays at a denaturation temperature of 98 degC and treatment duration of 05 min Table 23 details the experimental conditions

Figure 19 Comparison of Assays

The symbols compared the λtλu results from the HBB1 and either the NEIF or ND6 assays Each λtλu value is the ratio of the mean result of four technical replicates of an untreated and three or four replicates of the treated solutions The bars represent combined standard uncertainties The diagonal line represents equality between the assays

There is little or no difference between the λtλu results from the HBB1 NEIF and ND6 assays suggesting that the chromosomal location of the assay target has little influence Since the NEIFHBB1 comparison includes results for both the SRM 2372 and 2372a materials this equivalence is not material-specific

14

16

18

20

14 16 18 20

λ tλ

u NE

IF o

r ND6

λtλu HBB1

NEIFHBB1

ND6HBB1

2nd Assay = HBB1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

39

Table 23 Comparison of Assays

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu)

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045 dd20180105 2372a-B ND6 94 05 R5 14 3 05526 36 1711 0074 dd20180105 2372a-B ND6 90 05 R5 14 3 05326 18 1649 0051 dd20180105 2372a-B ND6 88 05 R5 14 3 05077 22 1572 0052 dd20180105 2372a-B ND6 86 05 R5 14 3 05038 23 1560 0053 dd20180105 2372a-B ND6 84 05 R5 14 3 04906 10 1519 0041

dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

40

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials

Figure 20 summarizes the currently available λtλu results obtained by denaturing at 98 degC for 05 min rapidly cooling in the thermal cycler with dilutions that provide asymp 03 copiesdroplet The figure also displays the consensus result and its 95 confidence region estimated using the Hierarchical Bayes method provided in the NIST Consensus Builder (NICOB) system3839 Table 24 details the experimental conditions

Figure 20 Evaluation of Limiting Ratio

The symbols represent the ratio of the mean result of three or four technical replicates of the untreated and treated solutions the bars represent combined standard uncertainties The thick horizontal line represents a consensus value covering all displayed results the dotted horizontal lines bracket the 95 confidence interval about the consensus value

The Hierarchical Bayes consensus result for these values is 183 with a 95 confidence interval of plusmn003 All but three of the 16 λtλu values are within one standard uncertainty of this consensus value with most of the λtλu means lying within the 95 confidence interval This suggests that there are no statistically significant differences between the SRM 2372 and 2372a materials with regard to ssDNA proportion

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

17

18

19

20

λ tλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

41

Table 24 Evaluation of Limiting Ratio

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180104a 2372-B HBB1 14 4 03577 05 dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032

dd20180108 2372-B HBB1 14 3 03463 09 dd20180108 2372-B HBB1 98 05 R5 14 3 06562 14 1895 0031

dd20180109 2372-B HBB1 14 3 03577 25 dd20180109 2372-B HBB1 98 05 R5 14 3 06495 21 1816 0059

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 4 06151 15 1843 0040

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

42

The agreement among the measured λtλu is more directly visualized using estimated unilateral ldquoDegrees of Equivalencerdquo d Each d is the absolute difference between the measured and consensus values

119889119889 = |120582120582119905119905 120582120582119906119906frasl minus consensus|

and is associated with a 95 confidence interval U95(d) that combines the uncertainty estimates of the measured and consensus values The U95(d) calculation details are specific to the consensus estimation method39 Values of d plusmn U95(d) that include 0 indicate that the λtλu result does not differ from the majority at about a 95 level of confidence Figure 21 displays the d for the λtλu results for the Hierarchical Bayes consensus value39 By this figure of merit all the λtλu can be considered as members of a single population

Figure 21 Degrees of Equivalence

The symbols represent the Degree of Equivalence d for the λtλu results relative to the Hierarchical Bayes consensus value The bars represent 95 level of confidence uncertainty intervals d plusmn U95(d) The horizontal line marks the d = 0 ldquono differencerdquo value Results with d plusmn U95(d) that include this line are not significantly different from the majority

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

-024

-018

-012

-006

000

006

012

018

024

Degr

ees o

f Equ

ival

ence

λtλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

43

Conclusions and Recommendations With high confidence all tubes of the SRM 2372a solution have the same nDNA copy number content within measurement repeatability Additionally within measurement repeatability all tubes of the SRM 2372a have the same mtDNAnDNA With high confidence the nDNA copy number content of the SRM 2372a solution is thermally stable from 4 degC to 37 degC over an extended period The solution should not be shipped or stored below 4 degC With high confidence the ddPCR measurements used to establish the entity and mass concentrations and the mtDNAnDNA ratios are adequately precise and unbiased With high confidence the DNA in the SRM 2372a solutions is now dsDNA and can be expected to remain dsDNA for at least ten years when stored at 4 degC Recommended Certification Values

Since the component materials are believed to be homogeneous are of very similar composition and were prepared and processed similarly the observed differences in the measured entity concentration uncertainties are likely artifacts of the measurement process To limit over-interpretation of these differences we recommend that the copy per nanoliter values of the three components be assigned same 95 relative uncertainty interval (CV95) based on the largest of the measured CV95 values (93 ) rounded up to 10 Since the ngμL transformation does not appreciably inflate this relative uncertainty we recommend that the mass concentration values also be assigned CV95 of 10 Similarly we recommend that the mtDNAnDNA ratios be rounded to integer values all three components be assigned the same CV95 and that this uncertainty be based on the largest (21 ) of the measured CV95 rounded up to 25 Table 25 lists the values that we recommend be used in the Certificate of Analysis

Table 25 Recommended Values and 95 Uncertainties for Certification

Component Units Value U95(Value) a A Copies per nanoliter 151 15 B Copies per nanoliter 175 18 C Copies per nanoliter 145 15 A ngμL 498 50

B ngμL 578 58 C ngμL 479 48 A mtDNAnDNA 174 4

B mtDNAnDNA 206 5 C mtDNAnDNA 279 7

a The interval Value plusmn U95(Value) is believed with about 95 confidence to contain the true value of the measurand

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

44

Use of SRM 2372a for Value-Assigning In-House Reference Materials

The within-component and between-assay variability observed for the commercial qPCR assays is not unexpected since different qPCR methods exploit qualitatively different molecular targets Given the intrinsic variability of the human genome the prospect of developing a human DNA source material that responds equally to all current and future qPCR assays is improbable All three components should be used to elucidate potential bias when using these materials to value assign in-house DNA reference and control materials for qPCR assays

Certificate of Analysis In accordance with ISO Guide 31 2000 a NIST SRM certificate is a document containing the name description and intended purpose of the material the logo of the US Department of Commerce the name of NIST as a certifying body instructions for proper use and storage of the material certified property value(s) with associated uncertainty(ies) method(s) used to obtain property values the period of validity if appropriate and any other technical information deemed necessary for its proper use A Certificate is issued for an SRM certified for one or more specific physical or engineering performance properties and may contain NIST reference information or both values in addition to certified values A Certificate of Analysis is issued for an SRM certified for one or more specific chemical properties Note ISO Guide 31 is updated periodically check with ISO for the latest version [httpswwwnistgovsrmsrm-definitions] For the most current version of the Certificate of Analysis for NIST SRM 2372a Human DNA Quantification Standard please visit httpswww-snistgovsrmorsview_detailcfmsrm=2372a

References

1 National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research Department of Health Education and Welfare (1978) Belmont Report Ethical Principles and Guidelines for the Protection of Human Subjects of Research DHEW Publication No (OS) 78-0012 httpvideocastnihgovpdfohrp_belmont_reportpdf

2 CDCNIH Biosafety in Microbiological and Biomedical Laboratories 5th ed HHS publication No (CDC) 21-1112 Chosewood LC Wilson DE Eds US Government Printing Office Washington DC (2009) httpwwwcdcgovbiosafetypublicationsbmbl5indexhtm

3 Duewer DL Kline MC Redman JW Newall PJ Reeder DJ NIST Mixed Stain Studies 1 and 2 Interlaboratory Comparison of DNA Quantification Practice and Short Tandem Repeat Multiplex Performance with Multiple-Source Samples J Forensic Sci 200146(5)1199-1210 PMID 11569565

4 Kline MC Duewer DL Redman JW Butler JM NIST Mixed Stain Study 3 DNA quantitation accuracy and its influence on short tandem repeat multiplex signal intensity Anal Chem 200375(10)2463-2469 httpsdoiorg101021ac026410i

5 Kline MC Duewer DL Redman JW Butler JM Results from the NIST 2004 DNA Quantitation Study J Forensic Sci 200550(3)571-578 PMID 15932088

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

45

6 Kline MC Duewer DL Travis JC Smith MV Redman JW Vallone PM Decker AE Butler JM Production and Certification of Standard Reference Material 2372 Human DNA Quantitation Standard Anal Bioanal Chem 2009394(4)1183-1192 httpsdoiorg101007s00216-009-2782-0

7 Sambrook J and Russell DW (2001) Molecular Cloning a Laboratory Manual Cold Spring Harbor Laboratory Press Cold Spring Harbor New York

8 EU-JRC Protocol NK603 ndash Community Reference Laboratory Event-specific method for the quantification of maize line NK603 using real-time PCR Section 33 Spectrophotometric Measurement of DNA Concentration httpgmo-crljrceceuropaeusummariesNK603-WEB-ProtocolValidationpdf

9 ISO 215712005(E) Foodstuffs ndash Methods of analysis for the detection of genetically modified organisms and derived products ndash Nucleic acid extraction Annex B Methods for quantitation of extracted DNA pp 34-36 International Organization for Standardization Geneva (2005)

10 Certificate of Analysis Standard Reference Material 2372 Human DNA Quantitation Standard (2013) Gaithersburg MD USA Standard Reference Materials Program National Institute of Standards and Technology httpswww-snistgovsrmorsview_certcfmsrm=2372

11 Vogelstein B Kinzler KW Digital PCR Proc Natl Acad Sci USA 1999969236-9241 12 Hindson BJ Ness KD Masquelier DA Belgrader P Heredia NJ Makarewicz AJ Bright IJ Lucero

MY Hiddessen AL Legler TC Kitano TK Hodel MR Petersen JF Wyatt PW Steenblock ER Shah PH Bousse LJ Troup CB Mellen JC Wittmann DK Erndt NG Cauley TH Koehler RT So AP Dube S Rose KA Montesclaros L Wang S Stumbo DP Hodges SP Romine S Milanovich FP White HE Regan JF Karlin-Neumann GA Hindson CM Saxonov S Colston BW High-Throughput Droplet Digital PCR System for Absolute Quantitation of DNA Copy Number Anal Chem 2011838604-8610 httpsdoiorg101021ac202028g

13 Pinheiro LB Coleman VA Hindson CM Herrmann J Hindson BJ Bhat S Emslie KR Evaluation of a Droplet Digital Polymerase Chain Reaction Format for DNA Copy Number Quantification AnalChem 2012841003-1011 httpsdoiorg101021ac202578x

14 Corbisier P Pinheiro L Mazoua S Kortekaas AM Chung PY Gerganova T Roebben G Emons H Emslie K DNA copy number concentration measured by digital and droplet digital quantitative PCR using certified reference materials Anal Bioanal Chem 2015407(7)1831-1840 httpsdoiorg101007s00216-015-8458-z

15 Duewer DL Kline MC Romsos EL Real-time cdPCR opens a window into events occurring in the first few PCR amplification cycles Anal Bioanal Chem 2015407(30)9061-9069 httpsdoiorg101007s00216-015-9073-8

16 Kline MC Romsos EL Duewer DL Evaluating Digital PCR for the Quantification of Human Genomic DNA Accessible Amplifiable Targets Anal Chem 201688(4)2132-2139 httpsdoiorg101021acsanalchem5b03692

17 Kline MC Duewer DL Evaluating Droplet Digital Polymerase Chain Reaction for the Quantification of Human Genomic DNA Lifting the Traceability Fog Anal Chem 201789(8)4648-4654 httpsdoiorg101021acsanalchem7b00240

18 Duewer DL Kline MC Romsos EL Toman B Evaluating Droplet Digital PCR for the quantification of human genomic DNA Conversion from Copies per Nanoliter to Nanogram Genomic DNA per Microliter Anal Bioanal Chem 2018in press httpsdoiorg101007s00216-018-0982-1

19 Dagata JA Farkas N Kramer JA Method for measuring the volume of nominally 100 μm diameter spherical water in oil emulsion droplets NIST Special Publication 260-184 2016 httpsdoiorg106028NISTSP260-184

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

20 ISO 170342016 General requirements for the competence of reference material producers International Organization for Standardization Geneva (2016)

21 NCBI Standard Nucleotide BLAST httpsblastncbinlmnihgovBlastcgiPAGE_TYPE=BlastSearch

22 Timken MD Swango KL Orrego C Buoncristiani MR A Duplex Real-Time qPCR Assay for the Quantification of Human Nuclear and Mitochondrial DNA in Forensic Samples Implications for Quantifying DNA in Degraded Samples J Forensic Sci 200550(5)1044-60 PMID 16225209

23 Walker JA Hedges DJ Perodeau BP Landry KE Stoilova N Laborde ME Shewale J Sinha SK Batzer MA Multiplex polymerase chain reaction for simultaneous quantitation of human nuclear mitochondrial and male Y-chromosome DNA application in human identification Anal Biochem 200533789ndash97 httpsdoiorg101016jab200409036

24 Kavlick MF Lawrence HS Merritt RT Fisher C Isenberg A Robertson JM Budowle B Quantification of Human Mitochondrial DNA Using Synthesized DNA Standards J Forensic Sci 201156(6) 1457ndash1463 httpsdoiorg101111j1556-4029201101871x

25 Stulnig TM Amberger A Exposing contaminating phenol in nucleic acid preparations BioTechniques 199416(3) 403-404

26 Majumdar N Wessel T Marks J Digital PCR Modeling for Maximal Sensitivity Dynamic Range and Measurement Precision PLoS ONE 201510(3)e0118833 httpsdoi101371journalpone0118833

27 Tellinghuisen J Partition Volume Variability in Digital Polymerase Chain Reaction Methods Polydispersity Causes Bias but Can Improve Precision Anal Chem 2016881283-12187 httpsdoiorg101021acsanalchem6b03139

28 Lunn DJ Spiegelhalter D Thomas A Best N (2009) Statistics in Medicine 283049ndash3082 29 De Biegravevre P Dybkaer R Fajgelj A Hibbert DB Metrological traceability of measurement results in

chemistry Concepts and implementation Pure Appl Chem 201183(10)1873-1935 httpsdoiorg101007s00769-011-0805-y

30 Huggett JF Foy CA Benes V Emslie K Garson JA Haynes R Hellemans J Kubista M Mueller RD Nolan T Pfaffl MW Shipley GL Vandesompele J Wittwer CT Bustin SA The Digital MIQE Guidelines Minimum Information for Publication of Quantitative Digital PCR Experiments Clin Chem 201359(6)892-902 httpsdoiorg101373clinchem2013206375

31 Wilson PJ Ellison SLR Extending digital PCR analysis by modelling quantification cycle data BMC Bioinfo 201617421 httpsdoiorg101186s12859-016-1275-3

32 Wang X Son A Effects of pretreatment on the denaturation and fragmentation of genomic DNA for DNA hybridization Environ Sci Processes Impacts 2013152204 httpsdoiorg101039c3em00457k

33 Nwokeoji AO Kilby PM Portwood DE Dickman MJ Accurate Quantification of Nucleic Acids Using Hypochromicity Measurements in Conjunction with UV Spectrophotometry Anal Chem 201789(24)13567-13574 https101021acsanalchem7b04000

34 Farkas DH Kaul KL Wiedbrauk DL Kiechle FL Specimen collection and storage for diagnostic molecular pathology investigation Arch Pathol Lab Med 1996120(6)591-6

35 Kypr J Kejnovskaacute I Renčiuk D Vorličkovaacute M Circular dichroism and conformational polymorphism of DNA Nucleic Acids Res 200937(6)1713-1725 httpsdoiorg101093nargkp026

36 Strachan T Read AP (1999) Human Molecular Genetics 2nd Ed John Wiley amp Sons NY

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

37 Bhat S McLaughlin JL Emslie KR Effect of sustained elevated temperature prior to amplification on template copy number estimation using digital polymerase chain reaction Analyst 2011136724ndash32 httpsdoiorg101039c0an00484g

38 NIST Consensus Builder (NICOB) httpsconsensusnistgov 39 Koepke A Lafarge T Toman B Possolo A NIST Consensus Builder Userrsquos Manual

httpsconsensusnistgovNISTConsensusBuilder-UserManualpdf

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

Appendix A DNA Extraction Method The following method has been adapted from Miller et al (1988)A1 Procedure

1 Transfer approximately 5 mL of buffy coat into a 50 mL sterile polypropylene centrifuge tube and add 30 mL Red Blood Cell Lysis Buffer

2 Mix by gentle inversion and let stand at room temperature for 15 min

3 Centrifuge at approximately 209 rads (2000 rpm) for 20 min

4 Discard supernatant

5 Add 9 mL of Nucleic Lysis Buffer and re-suspend the pelleted cells by agitating with a plastic transfer pipette

6 Add 300 microL of 20 sodium dodecyl sulfate (SDS) and 18 mL of Protease K Solution (freshly prepared 2 mg to 5 mg Protease K per 1 mL Protease K Buffer)

7 Mix by inversion followed by incubation overnight or longer at 37 ordmC to 40 ordmC in a shaking water bath The length of incubation is determined by the time required to solubilize the cells present in the tube

8 Add 4 mL saturated ammonium acetate solution (in deionized water stored at 2 ordmC to 8 ordmC) Shake vigorously for 15 s

9 Centrifuge at approximately 471 rads (4500 rpm) for 10 min

10 Pipette supernatant into a 50 mL sterile polypropylene centrifuge tube and add two volumes of room temperature absolute ethanol Mix by gentle inversion until the DNA aggregates and floats to the top of the tube

11 Spool the aggregated DNA from the current tube into a fresh tube of 80 ethanol using a sterile inoculation loop At this step multiple tubes of precipitated DNA may be pooled to make a large lot of the purified material Mix by gentle inversion

12 Transfer the aggregated DNA to a fresh dry 50 mL polypropylene centrifuge tube Remove as much ethanol as possible by squeezing the DNA against the wall of the tube with a sterile inoculation loop Air-dry the material in a laminar flow hood

13 Resolubilize DNA in TE-4 buffer and store material at 4 ordmC Reagents

Red Blood Cell Lysis Buffer 144 mmolL ammonium chloride 1 mmolL sodium bicarbonate in deionized (DI) water Store at 2 ordmC to 8 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

Nucleic Lysis Buffer 10 mmolL Tris-HCl 400 mmolL sodium chloride 20 mmolL disodium ethylenediaminetetraacetic acid in DI Water adjust pH to 82 Store at 2 ordmC to 8 ordmC

Protease K Buffer

20 mmolL disodium ethylenediaminetetraacetic acid 10 (mass per volume) Sodium Dodecyl Sulfate (SDS) in DI water Store at 2 ordmC to 8 ordmC

Protease K

Sigma-Aldrich cat P6556 Tritirachium album lyophilized powder Store at ndash20 ordmC

Ammonium Acetate

192 molL ammonium acetate (saturated) in DI water Store at 2 ordmC to 8 ordmC

Absolute Ethanol (200 proof)

Store at room temperature Reference

A1 Miller SA Dykes DD Polesky HF A simple salting out procedure for extracting DNA from human nucleated cells Nucleic Acids Res 198816(3)1215 httpsdoiorg101093nar1631215

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

48

Appendix B Spectrophotometric Measurement Protocol The following is for use with a scanning recording spectrophotometer The procedures include the spectrophotometer calibration measurement of the ldquonativerdquo dsDNA and measurement of the NaOH denatured ssDNA The procedure for evaluating solutions after treatment with NaOH is adapted from references89 Spectrophotometer Calibration

Establish Baseline Set the spectrophotometer to acquire data from 220 nm to 345 nm in steps of 01 nm Zero the instrument with nothing (air) in the sample and reference cells Acquire an air-air baseline of the instrument with nothing (air) in the sample and reference cells If the trace has significant structure or excessive noise take corrective action If the trace is acceptable proceed with data collection Enable Verification of Wavelength Calibration SRM 2034 is a solution of holmium oxide in 10 perchloric acid contained in a sealed quartz cuvette delivering an intrinsic UVVis wavelength standard for absorption spectrophotometryB1 Evaluation of the location of peak maxima in the spectra of this solution enables verification of the spectrophotometers wavelength calibration Place SRM 2034 in the sample cell Create and name a file to hold the verification spectra Perform three replicate SRM 2034-to-air scans saving each scan to the file Remove SRM 2034 from the spectrophotometer and store appropriately Enable Verification of Absorbance Calibration SRM 2031 consists of three metal-on-fused silica filters held in cuvette-sized filter holders plus an empty filter holder The three filters are certified for transmittance and transmittance density (-log10(transmittance) effectively absorbance) at several wavelengths with nominal wavelength-independent transmittances of 10 30 and 90 Evaluation of the absorbance of each of these three filters at the certified 250 nm 280 nm and 340 nm wavelengths enables verification of the spectrophotometers absorbance calibration Place the empty SRM 2031 filter holder in the reference cell Remove the protective covers from the 10 transmittance filter and place in the sample cell Perform three replicate scans saving each scan to file Remove the filter from the spectrophotometer replace the protective covers and store appropriately Repeat for the 30 and 90 filters Close the file containing the verification spectra For ldquonativerdquo dsDNA Samples

Verify Optical Balance of Cuvettes The measurements of the DNA-containing samples are made using a 100 microL quartz cuvette in the sample cell compared to an optically matched cuvette in the reference cell The optical equivalence of these two cuvettes can be verified by filling both cuvettes with appropriate volumes of a sample solution matrix (TE-4 buffer) that does not contain DNA Fill two dry

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

49

clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes followed by a methanol rinse in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquonativerdquo dsDNA remove and clean both the reference cuvette and the sample cuvette For NaOH denatured ssDNA Samples

Prepare a 04 molL NaOH Stock Solution To a 15 mL disposable centrifuge tube add 016 g NaOH pellets add deionized (DI) water to the 10 mL line and mix until all the NaOH pellets have dissolved Centrifuge at 524 rads (5000 rpm) for 5 min For every 9 samples to be analyzed aliquot 1 mL of this 04 molL NaOH stock solution into a 15 mL screw top tube Use these tubes to prepare the NaOH blanks and DNA samples In a 15 mL screw cap vial mix 200 microL TE-4 buffer with 200 microL 04 molL sodium hydroxide solution to achieve a final NaOH concentration of 02 molL Vortex 5 s then centrifuge at 1257 rads (12 000 rpm) for 1 min Fill two dry clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Prepare the Sample Solutions For each unit of material to be measured (Tube or vial)

a) Vortex the unit for 5 s b) Centrifuge at 1257 rads (12 000 rpm) for 1 min c) Remove a 100 microL aliquot and place in a new labeled vial d) Add 100 microL of the freshly prepared 04 molL NaOH solution (with the same pipet) e) Vortex 5 s f) Centrifuge at 1257 rads (12 000 rpm) for 1 min

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

50

Measure Sample Solutions Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes rinse with methanol in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the NaOH denatured DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquodenaturedrdquo ssDNA close the file holding the spectra for this series Remove and clean both the reference cuvette and the sample cuvette After All Sample Measurements are Complete

Enable Verification of Spectrophotometer Stability The stability of the spectrophotometer over the course of the data collection can be evaluated by comparing spectra for SRM 2034 and 2031 taken at the start of data collection with spectra taken at the end of data collection Repeat the wavelength and absorbance verification measurements described above collecting the spectra in a suitably named file Clean Up Transfer all relevant data files from the spectrophotometers storage to a transfer device Clean the spectrophotometer work area as necessary and restore the SRMs and cuvettes to their appropriate storage locations Evaluation

Export the files to a spreadsheet Average the replicate spectra Subtract the average absorbance at A320 from the absorbance at A260 resulting in the corrected absorbance at A260 The conventional spectrophotometric estimate of the mass concentration of the dsDNA in the sample is achieved by multiplying the corrected absorbance by 50 The conventional estimate for the NaOH-treated ssDNA is achieved by multiplying the corrected absorbance by 37times2 where the ldquotimes2rdquo adjusts for the 1rarr2 volumetric dilution of the sample Note reliable spectrophotometric measurements require corrected A260 to be greater than 005 Reference

B1 Travis JC Acosta JC Andor G Bastie J Blattner P Chunnilall CJ Crosson SC Duewer DL Early EA Hengstberger F Kim C-S Liedquist L Manoocheri F Mercader F Metzdorf J Mito A Monardh LAG Nevask S Nilsson M Noeumll M Rodriguez AC Ruiacutez A Schirmacher A Smith MV Valencia G van Tonder N Zwinkels J Intrinsic Wavelength Standard Absorption Bands in Holmium Oxide Solution for UVvisible Molecular Absorption Spectrophotometry J Phys Chem Ref Data 200534(1)41-56

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

51

Appendix C SRM 2372a Statisticianrsquos report Extracted from Blaza Toman

November 30 2017 1 Introduction Candidate material for SRM 2372a is composed of three components labeled A B and C All components are human genomic DNA and prepared at NIST from buffy coat white blood cells from single-source anonymous donors Component A and B are single source (malefemale respectively) and Component C is a 1 male to 3 female mixture of two single-source donors Droplet digital PCR (ddPCR) assays were used to obtain measurements of the concentration of copies per microliter of sample Ten NIST-developed human nuclear DNA and three literature-derived human mitochondrial DNA PCR assays were used in this study

11 OpenBUGS Analysis All parameters in the following sections were estimated using Markov Chain Monte Carlo (MCMC) implemented in the OpenBUGS freewareC1 OpenBUGS is a software application for the Bayesian analysis of complex statistical models

12 OpenBUGS Exemplar Output All following sections list the OpenBUGs model (ldquocoderdquo) initialization values (ldquoinitsrdquo) if they are needed and the data used to evaluate the various data sets discussed in the body of this report homogeneity stability nDNA certification and mtDNAnDNA value assignment The provided information in each of the following sections is sufficient to produce estimates but the estimates may not exactly reproduce the exemplar results provided The MCMC paradigm does not produce ldquoanalyticalrdquo values but approaches them asymptotically as the number o model updates increases All exemplar results presented here are based upon the following bull 10-fold ldquothinningrdquo to minimize autocorrelation there were 10 MCMC estimates for every

model update used in parameter estimation bull a ldquoburn inrdquo of 5 000 updates before beginning parameter estimation and bull 10 000 model updates

The following statistics are provided in the tables listing the exemplar results

mean arithmetic mean of the parameterrsquos model update values sd standard deviation of the parameterrsquos model update values MC_error Monte Carlo standard deviation of the mean meanradic(number of model updates) val25pc 25th percentile of the model update values median 50th percentile of the model update values val975pc 975th percentile of the model update values startnumber of ldquoburn inrdquo updates sample number of model update values

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

52

2 Homogeneity assessment 21 Data

Samples of the material of all three components were drawn from 22 different boxes Two assays (NEIF ND6) were used to obtain 3 replicates of the measurement in copies per nanoliter for each sample and component For the complete homogeneity data see

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

22 Model The data were analyzed using a three-level Gaussian hierarchical ANOVA model

119910119910119894119894119894119894119894119894~119873119873120572120572119894119894119894119894119894119894 120590120590119908119908119894119894119905119905ℎ119894119894119894119894 119887119887119887119887119887119887 1198941198941198941198942

where ~N() denotes ldquodistributed as a normal distribution of some mean μ and variancerdquo 119894119894 denotes component (1 for A 2 for B 3 for C) 119895119895 denotes assay (1 for NEIF 2 for ND6) 119896119896 denotes box (1hellip22) and 120572120572119894119894119894119894119894119894~119873119873micro119894119894119894119894120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119887119887119887119887119887119887119894119894

2

All variance components were given a Half Cauchy prior as described in Reference C2 Note The Half Cauchy prior is specified by ldquo~dnorm(0 00016)I(0001)rdquo The ldquoIrdquo notation is for censoring It cuts off any random draws from the Normal distribution that are lower than 0001 An informative prior is needed when there are less than four data per distribution and the half Cauchy is a good general-purpose prior

23 OpenBUGS code and data Homogeneity code ModelBegin for(i in 13)xiNsb[i] ~ dnorm(0 00016)I(0001) chSqNsb[i] ~ dgamma(0505) sigb[i] lt- xiNsb[i]sqrt(chSqNsb[i]) for(i in 13)for(j in 12)mu[ij]~dnorm(010E-5) for(i in 13) muA[i]lt-mean(mu[i])sigbP[i]lt-100(muA[i]sqrt(sigb[i])) for(i in 13)for(j in 12)xiNs[ij] ~ dnorm(0 00016)I(0001) chSqNs[ij]~dgamma(0505) sigt[ij]lt-xiNs[ij]sqrt(chSqNs[ij]) sigtP[ij]lt-100(mu[ij]sqrt(sigt[ij])) for(i in 13)for(j in 12)for(k in 122)alpha[ijk]~dnorm(mu[ij]sigb[i]) for(idx in 1393)lamb[idx]~dnorm(alpha[cmp[idx]asy[idx]box[idx]]sigt[cmp[idx]asy[idx]]) ModelEnd Inits None required

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

53

Homogeneity Data lamb[] asy[] box[] cmp[] rep[] 1632 1 1 1 1 1529 1 1 1 2 1511 1 1 1 3 1616 2 1 1 1 1640 2 1 1 2 1582 2 1 1 3 1531 1 2 1 1 1525 1 2 1 2 1549 1 2 1 3 1615 2 2 1 1 1589 2 2 1 2 1601 2 2 1 3 1524 1 3 1 1 1540 1 3 1 2 1533 1 3 1 3 1582 2 3 1 1 1152 2 3 1 2 1555 2 3 1 3 1454 1 4 1 1 1458 1 4 1 2 1427 1 4 1 3 1627 2 4 1 1 1605 2 4 1 2 1607 2 4 1 3 1465 1 5 1 1 1540 1 5 1 2 1504 1 5 1 3 1607 2 5 1 1 1633 2 5 1 2 1586 2 5 1 3 1512 1 6 1 1 1491 1 6 1 2 1491 1 6 1 3 1511 2 6 1 1 1483 2 6 1 2 1524 2 6 1 3 1418 1 7 1 1 1480 1 7 1 2 1455 1 7 1 3 1479 2 7 1 1 1557 2 7 1 2 1656 2 7 1 3 1496 1 8 1 1 1560 1 8 1 2 1501 1 8 1 3 1514 2 8 1 1 1615 2 8 1 2 1540 2 8 1 3 1588 1 9 1 1 1581 1 9 1 2 1637 1 9 1 3 1474 2 9 1 1 1438 2 9 1 2 1454 2 9 1 3 1574 1 10 1 1 1513 1 10 1 2 1570 1 10 1 3 1594 2 10 1 1 1713 2 10 1 2 1669 2 10 1 3

1557 1 11 1 1 1508 1 11 1 2 1633 1 11 1 3 1571 2 11 1 1 1611 2 11 1 2 1631 2 11 1 3 1500 1 12 1 1 1476 1 12 1 2 1515 1 12 1 3 1626 2 12 1 1 1586 2 12 1 2 1608 2 12 1 3 1672 1 13 1 1 1634 1 13 1 2 1804 1 13 1 3 1490 2 13 1 1 1516 2 13 1 2 1497 2 13 1 3 1551 1 14 1 1 1568 1 14 1 2 1538 1 14 1 3 1617 2 14 1 1 1614 2 14 1 2 1639 2 14 1 3 1456 1 15 1 1 1448 1 15 1 2 1523 1 15 1 3 1525 2 15 1 1 1402 2 15 1 2 1436 2 15 1 3 1573 1 16 1 1 1520 1 16 1 2 1617 1 16 1 3 1634 2 16 1 1 1627 2 16 1 2 1575 2 16 1 3 1516 1 17 1 1 1516 1 17 1 2 1628 1 17 1 3 1613 2 17 1 1 1567 2 17 1 2 1567 2 17 1 3 1612 1 18 1 1 1535 1 18 1 2 1672 1 18 1 3 1529 2 18 1 1 1495 2 18 1 2 1528 2 18 1 3 1598 1 19 1 1 1514 1 19 1 2 1596 1 19 1 3 1571 2 19 1 1 1535 2 19 1 2 1511 2 19 1 3 1575 1 20 1 1 1568 1 20 1 2 1586 1 20 1 3 1652 2 20 1 1 1545 2 20 1 2 1627 2 20 1 3 1658 1 21 1 1

1611 1 21 1 2 1578 1 21 1 3 1647 2 21 1 1 1607 2 21 1 2 1612 2 21 1 3 1547 1 22 1 1 1528 1 22 1 2 1532 1 22 1 3 1537 2 22 1 1 1513 2 22 1 2 1536 2 22 1 3 1790 1 1 2 1 1909 1 1 2 2 1885 1 1 2 3 1870 2 1 2 2 1860 2 1 2 3 1712 1 2 2 1 1719 1 2 2 2 1704 1 2 2 3 1775 2 2 2 1 1718 2 2 2 2 1756 2 2 2 3 1693 1 3 2 1 1755 1 3 2 2 1749 1 3 2 3 1824 2 3 2 1 1483 2 3 2 2 1896 2 3 2 3 1754 1 4 2 1 1745 1 4 2 2 1739 1 4 2 3 1670 2 4 2 1 1632 2 4 2 2 1814 2 4 2 3 1757 1 5 2 1 1687 1 5 2 2 1735 1 5 2 3 1683 2 5 2 1 1759 2 5 2 2 1703 2 5 2 3 1697 1 6 2 1 1709 1 6 2 2 1714 1 6 2 3 1693 2 6 2 1 1723 2 6 2 2 1722 2 6 2 3 1722 1 7 2 1 1625 1 7 2 2 1703 1 7 2 3 1690 2 7 2 1 1684 2 7 2 2 1688 2 7 2 3 1884 1 8 2 1 1802 1 8 2 2 1772 1 8 2 3 1864 2 8 2 1 1887 2 8 2 2 1821 2 8 2 3 1713 1 9 2 1 1734 1 9 2 2 1727 1 9 2 3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

54

1709 2 9 2 1 1756 2 9 2 2 1758 2 9 2 3 1759 1 10 2 1 1809 1 10 2 2 1738 1 10 2 3 1764 2 10 2 1 1816 2 10 2 2 1787 2 10 2 3 1749 1 11 2 1 1782 1 11 2 2 1740 1 11 2 3 1741 2 11 2 1 1792 2 11 2 2 1784 2 11 2 3 1748 1 12 2 1 1712 1 12 2 2 1718 1 12 2 3 1797 2 12 2 1 1614 2 12 2 2 1790 2 12 2 3 1632 1 13 2 1 1668 1 13 2 2 1701 1 13 2 3 1626 2 13 2 1 1566 2 13 2 2 1681 2 13 2 3 1752 1 14 2 1 1535 1 14 2 2 1673 1 14 2 3 1724 2 14 2 1 1741 2 14 2 2 1685 2 14 2 3 1871 1 15 2 1 1835 1 15 2 2 1887 1 15 2 3 1770 2 15 2 1 1752 2 15 2 2 1798 2 15 2 3 1688 1 16 2 1 1725 1 16 2 2 1743 1 16 2 3 1744 2 16 2 1 1697 2 16 2 2 1710 2 16 2 3 1789 1 17 2 1 1769 1 17 2 2 1825 1 17 2 3 1716 2 17 2 1 1762 2 17 2 2 1638 2 17 2 3 1826 1 18 2 1 1780 1 18 2 2 1849 1 18 2 3 1805 2 18 2 1 1801 2 18 2 2 1931 2 18 2 3 1751 1 19 2 1 1759 1 19 2 2 1880 1 19 2 3 1827 2 19 2 1 1766 2 19 2 2

1786 2 19 2 3 1738 1 20 2 1 1678 1 20 2 2 1768 1 20 2 3 1752 2 20 2 1 1747 2 20 2 2 1768 2 20 2 3 1742 1 21 2 1 1836 1 21 2 2 1713 1 21 2 3 1746 2 21 2 1 1744 2 21 2 2 1764 2 21 2 3 1761 1 22 2 1 1743 1 22 2 2 1660 1 22 2 3 1666 2 22 2 1 1692 2 22 2 2 1819 2 22 2 3 1493 1 1 3 1 1385 1 1 3 2 1510 1 1 3 3 1565 2 1 3 1 1562 2 1 3 2 1466 2 1 3 3 1516 1 2 3 1 1513 1 2 3 2 1577 1 2 3 3 1490 2 2 3 1 1547 2 2 3 2 1570 2 2 3 3 1451 1 3 3 1 1528 1 3 3 2 1506 1 3 3 3 1485 2 3 3 1 1505 2 3 3 2 1471 2 3 3 3 1440 1 4 3 1 1486 1 4 3 2 1520 1 4 3 3 1529 2 4 3 1 1508 2 4 3 2 1506 2 4 3 3 1523 1 5 3 1 1472 1 5 3 2 1483 1 5 3 3 1513 2 5 3 1 1520 2 5 3 2 1491 2 5 3 3 1437 1 6 3 1 1493 1 6 3 2 1522 1 6 3 3 1492 2 6 3 1 1481 2 6 3 2 1461 2 6 3 3 1485 1 7 3 1 1477 1 7 3 2 1463 1 7 3 3 1475 2 7 3 1 1547 2 7 3 2 1485 2 7 3 3 1433 1 8 3 1

1405 1 8 3 2 1478 1 8 3 3 1497 2 8 3 1 1501 2 8 3 2 1495 2 8 3 3 1463 1 9 3 1 1499 1 9 3 2 1505 1 9 3 3 1503 2 9 3 1 1491 2 9 3 2 1516 2 9 3 3 1534 1 10 3 1 1458 1 10 3 2 1454 1 10 3 3 1448 2 10 3 1 1465 2 10 3 2 1474 2 10 3 3 1533 1 11 3 1 1514 1 11 3 2 1555 1 11 3 3 1488 2 11 3 1 1470 2 11 3 2 1483 2 11 3 3 1450 1 12 3 1 1525 1 12 3 2 1497 1 12 3 3 1496 2 12 3 1 1483 2 12 3 2 1519 2 12 3 3 1537 1 13 3 1 1527 1 13 3 2 1515 1 13 3 3 1499 2 13 3 1 1479 2 13 3 2 1475 2 13 3 3 1515 1 14 3 1 1490 1 14 3 2 1494 1 14 3 3 1469 2 14 3 1 1480 2 14 3 2 1466 2 14 3 3 1511 1 15 3 1 1568 1 15 3 2 1594 1 15 3 3 1515 2 15 3 1 1543 2 15 3 2 1565 2 15 3 3 1449 1 16 3 1 1452 1 16 3 2 1471 1 16 3 3 1543 2 16 3 1 1448 2 16 3 2 1491 2 16 3 3 1388 1 17 3 1 1421 1 17 3 2 1394 1 17 3 3 1452 2 17 3 1 1435 2 17 3 3 1475 1 18 3 1 1455 1 18 3 2 1445 1 18 3 3 1458 2 18 3 1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

55

1531 2 18 3 2 1459 2 18 3 3 1468 1 19 3 1 1525 1 19 3 2 1446 1 19 3 3 1435 2 19 3 1 1478 2 19 3 2 1421 2 19 3 3 1549 1 20 3 1

1531 1 20 3 2 1499 1 20 3 3 1499 2 20 3 1 1465 2 20 3 2 1468 2 20 3 3 1530 1 21 3 1 1440 1 21 3 2 1484 1 21 3 3 1499 2 21 3 1

1454 2 21 3 2 1442 1 22 3 1 1487 1 22 3 2 1436 1 22 3 3 1589 2 22 3 1 1591 2 22 3 2 1556 2 22 3 3 END

Table C-1 Example OpenBUGS Homogeneity Summary

mean sd MC_error val25pc median val975pc start sample mu[11] 15460 0122 000123 1522 1546 1570 5001 10000 mu[12] 15640 0137 000141 1537 1564 1591 5001 10000 mu[21] 17490 0113 000118 1727 1749 1772 5001 10000 mu[22] 17480 0129 000125 1722 1748 1774 5001 10000 mu[31] 14870 0073 000073 1473 1487 1501 5001 10000 mu[32] 14960 0069 000063 1482 1496 1509 5001 10000 muA[1] 15550 0092 000092 1536 1555 1573 5001 10000 muA[2] 17490 0085 000086 1732 1749 1765 5001 10000 muA[3] 14910 0050 000046 1481 1491 1501 5001 10000 sigbP[1] 3269 04985 000514 2376 3244 4324 5001 10000 sigbP[2] 2548 04385 000461 1733 2533 3472 5001 10000 sigbP[3] 1850 0292 000260 1329 1832 2477 5001 10000

sigtP[11] 2774 0298 000260 2268 2750 3442 5001 10000 sigtP[12] 4036 0429 000438 3290 4005 4986 5001 10000 sigtP[21] 2664 0301 000321 2155 2640 3313 5001 10000 sigtP[22] 3807 0395 000377 3116 3772 4662 5001 10000 sigtP[31] 2313 0250 000209 1880 2291 2873 5001 10000 sigtP[32] 1812 0195 000170 1473 1797 2243 5001 10000

24 Results The material appears to be homogeneous as can be seen in Figure C-1 Based on these results the between-tube variability is of the same order as within-tube variability

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

56

Component A Assay NEIF Component A Assay ND6 box plot gamma[11]

[111][112][113]

[114]

[115][116]

[117]

[118]

[119]

[1110][1111]

[1112]

[1113]

[1114]

[1115]

[1116][1117]

[1118]

[1119][1120]

[1121]

[1122]

114

box plot gamma[12]

[121][122]

[123]

[124][125]

[126]

[127][128]

[129]

[1210]

[1211][1212]

[1213]

[1214]

[1215]

[1216]

[1217]

[1218][1219]

[1220][1221]

[1222]

115

Component B Assay NEIF Component B Assay ND6

box plot gamma[21]

[211]

[212]

[213][214]

[215][216]

[217]

[218]

[219]

[2110][2111]

[2112]

[2113][2114]

[2115]

[2116]

[2117]

[2118]

[2119]

[2120]

[2121]

[2122]

129

box plot gamma[22]

[221]

[222][223]

[224][225][226][227]

[228]

[229]

[2210][2211]

[2212]

[2213]

[2214]

[2215]

[2216][2217]

[2218]

[2219]

[2220][2221][2222]

128

Component C Assay NEIF Component C Assay ND6

box plot gamma[31]

[311]

[312]

[313][314]

[315][316]

[317]

[318]

[319][3110]

[3111]

[3112]

[3113]

[3114]

[3115]

[3116]

[3117]

[3118]

[3119]

[3120]

[3121]

[3122]109

box plot gamma[32]

[321][322]

[323]

[324][325]

[326]

[327][328][329]

[3210]

[3211]

[3212]

[3213][3214]

[3215]

[3216]

[3217]

[3218]

[3219]

[3220][3221]

[3222]

11

Figure C-1 Homogeneity Assessment

Note These results were obtained using the 1-4 gravimetric dilution-adjusted λ but uncorrected for droplet volume

or the 1-10 volumetric dilution That is the vertical axis in each panel should be multiplied by 07349times10 = 7349 This constant difference in scale does not impact the homogeneity analysis Error bars span approximate 95 confidence intervals

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

57

3 Certification measurements 31 Data

Originally the experiment consisted of 5 plates containing the 3 components (A B C) and 12 assays Ten of the assays were for genomic DNA identification (for the certified value) and two of the assays were for mitochondrial detection which is used for informational values (and are not included in this data set) Each of the plates contained samples from a new vial from different boxes (noted in the ldquovial usedrdquo portion the number references the box) The assay positions were alternated across the five plates as well For the plate design and which tubes were used see

Figure 6 Basic Plate Layout Design for Certification Measurements It was later determined that the homogeneity and stability data sets were obtained using a different set of pipettes than the certification data set and since all data was to be combined for certification another plate was obtained using the original pipette For the complete certification data see

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component Stability data was measured at different temperatures over several weeks using assays NEIF and ND6 Only the 4degC data was also used for certification For the complete stability data see

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

32 Model It was decided that all three sets of data should be used to derive the certified values In this way the uncertainty of the certified values would better reflect any potential heterogeneity in the samples as well as any additional unknown factors that may have effected the results Because two different pipettes were used to obtain different sets of data there was a significant bias in some of the measurements Specifically the ldquocertificationrdquo data from the first five plates was obtained using pipette set ldquo1rdquo and the remaining measurements were obtained using pipette set ldquo2rdquo This bias was introduced into the statistical model as part of the mean Possible effects of different assays and type of data (heterogeneity certification and stability) were accounted for in the statistical model in terms of a random effect Specifically the response for each component (A B C) was represented as

119910119910119894119894119894119894119894119894~1198731198731205741205741198941198941198941198941198941198941205901205901198941198942 where ~N() denotes ldquodistributed as a normal distribution of some mean and variancerdquo i = 123 denotes type of measurement heterogeneity certification and stability j = 1hellip10 denotes assay 2PR4 POTP NEIF R4Q5 D5 ND6 D9 HBB1 ND14 22C3 k = 1 2 denotes pipette set 120574120574119894119894119894119894119894119894 = 120572120572119894119894119894119894 + 120573120573119894119894 120572120572119894119894119894119894~119873119873(micro1198941198941205901205901205721205722) 120573120573119894119894 = 120573120573 for data type 1 and 120573120573119894119894 = 0 for data types 2 and 3 For a given 2372a component the variance 1205901205901205721205722 measures the between assay variability and 1205901205901198941198942 measures the repeatability variability for each data type The data was analyzed using Bayesian MCMC programmed in OpenBUGSC2 with non-informative Gaussian priors for the means and Gamma priors for the variance components

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

58

33 OpenBUGS code inits and data Certification code ModelBegin for(i in 13)mu[i]~dnorm(010E-5) sig[i]~dgamma(10E-510E-5) sigalph~dgamma(10E-510E-5) for(i in 110)alpha[1i]~dnorm(mu[1] sigalph) for(i in 12)alpha[2i]~dnorm(mu[2] sigalph) alpha[3i]~dnorm(mu[3]sigalph) b~dnorm(010E-5) beta[1]lt-b beta[2]lt-00 for(i in 1Ntot)mean[i]lt-alpha[typ[i]asy[i]]+beta[pip[i]] for(i in 1Ntot)lamb[i]~dnorm(mean[i]sig[typ[i]]) ModelEnd Certification Inits list(sig=c(111) sigalph=1) Certification Data for Component A list(Ntot=275) lamb[] asy[] cmp[] typ[] pip[] 1667 1 1 1 1 1655 2 1 1 1 1647 3 1 1 1 1594 4 1 1 1 1655 5 1 1 1 1629 6 1 1 1 1655 7 1 1 1 1644 8 1 1 1 1645 9 1 1 1 1545 10 1 1 1 1698 1 1 1 1 1685 2 1 1 1 1665 3 1 1 1 1658 4 1 1 1 1683 5 1 1 1 1647 6 1 1 1 1652 7 1 1 1 1611 8 1 1 1 1707 9 1 1 1 1603 10 1 1 1 1668 1 1 1 1 1587 2 1 1 1 1678 3 1 1 1 1628 4 1 1 1 1738 5 1 1 1 1667 6 1 1 1 1695 7 1 1 1 1757 8 1 1 1 1771 9 1 1 1 1762 10 1 1 1 1739 1 1 1 1 1702 2 1 1 1 1614 3 1 1 1 1762 4 1 1 1 1649 5 1 1 1 1678 6 1 1 1 1653 7 1 1 1

175 8 1 1 1 1764 9 1 1 1 1759 10 1 1 1 1638 1 1 1 1 1676 2 1 1 1 1616 3 1 1 1 1559 4 1 1 1 1608 5 1 1 1 1611 6 1 1 1 1517 7 1 1 1 1557 8 1 1 1 1604 9 1 1 1 1675 10 1 1 1 1667 1 1 1 1 1638 2 1 1 1 1624 3 1 1 1 1587 4 1 1 1 1652 5 1 1 1 1589 6 1 1 1 1607 7 1 1 1 1582 8 1 1 1 1625 9 1 1 1 1616 10 1 1 1 1888 1 1 1 1 1868 2 1 1 1 1953 3 1 1 1 1779 4 1 1 1 1924 5 1 1 1 1861 6 1 1 1 1824 7 1 1 1 1809 8 1 1 1 1864 9 1 1 1 193 10 1 1 1 1939 1 1 1 1 1815 2 1 1 1 1902 3 1 1 1 1871 4 1 1 1 2124 5 1 1 1

1787 6 1 1 1 1844 7 1 1 1 1799 8 1 1 1 1857 9 1 1 1 1942 10 1 1 1 1597 1 1 1 1 1533 2 1 1 1 1462 3 1 1 1 1493 4 1 1 1 1538 5 1 1 1 1492 6 1 1 1 1528 7 1 1 1 1548 8 1 1 1 1513 9 1 1 1 1562 10 1 1 1 159 1 1 1 1 1404 2 1 1 1 1533 3 1 1 1 1519 4 1 1 1 1586 5 1 1 1 1516 6 1 1 1 1538 7 1 1 1 1581 8 1 1 1 1514 9 1 1 1 1431 1 1 1 2 1407 2 1 1 2 1487 3 1 1 2 145 4 1 1 2 1394 5 1 1 2 1396 6 1 1 2 1414 7 1 1 2 1401 8 1 1 2 1412 9 1 1 2 1435 10 1 1 2 1411 1 1 1 2 1395 2 1 1 2 1398 3 1 1 2 1439 4 1 1 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

59

1422 5 1 1 2 1393 6 1 1 2 139 7 1 1 2 1382 8 1 1 2 146 9 1 1 2 141 10 1 1 2 1632 1 1 2 2 1531 1 1 2 2 1524 1 1 2 2 1454 1 1 2 2 1465 1 1 2 2 1512 1 1 2 2 1418 1 1 2 2 1496 1 1 2 2 1588 1 1 2 2 1574 1 1 2 2 1557 1 1 2 2 15 1 1 2 2 1672 1 1 2 2 1551 1 1 2 2 1573 1 1 2 2 1516 1 1 2 2 1612 1 1 2 2 1598 1 1 2 2 1575 1 1 2 2 1658 1 1 2 2 1547 1 1 2 2 1529 1 1 2 2 1525 1 1 2 2 154 1 1 2 2 1458 1 1 2 2 154 1 1 2 2 1491 1 1 2 2 148 1 1 2 2 156 1 1 2 2 1581 1 1 2 2 1513 1 1 2 2 1508 1 1 2 2 1476 1 1 2 2 1634 1 1 2 2 1568 1 1 2 2 152 1 1 2 2 1516 1 1 2 2 1535 1 1 2 2 1514 1 1 2 2 1568 1 1 2 2 1611 1 1 2 2 1528 1 1 2 2 1511 1 1 2 2 1549 1 1 2 2 1533 1 1 2 2 1427 1 1 2 2 1504 1 1 2 2 1491 1 1 2 2 1455 1 1 2 2

1501 1 1 2 2 1637 1 1 2 2 157 1 1 2 2 1633 1 1 2 2 1515 1 1 2 2 1804 1 1 2 2 1538 1 1 2 2 1617 1 1 2 2 1628 1 1 2 2 1672 1 1 2 2 1596 1 1 2 2 1586 1 1 2 2 1578 1 1 2 2 1532 1 1 2 2 1616 2 1 2 2 1615 2 1 2 2 1582 2 1 2 2 1627 2 1 2 2 1607 2 1 2 2 1511 2 1 2 2 1479 2 1 2 2 1514 2 1 2 2 1474 2 1 2 2 1594 2 1 2 2 1571 2 1 2 2 1626 2 1 2 2 149 2 1 2 2 1617 2 1 2 2 1634 2 1 2 2 1613 2 1 2 2 1529 2 1 2 2 1571 2 1 2 2 1652 2 1 2 2 1647 2 1 2 2 1537 2 1 2 2 164 2 1 2 2 1589 2 1 2 2 1152 2 1 2 2 1605 2 1 2 2 1633 2 1 2 2 1483 2 1 2 2 1557 2 1 2 2 1615 2 1 2 2 1438 2 1 2 2 1713 2 1 2 2 1611 2 1 2 2 1586 2 1 2 2 1516 2 1 2 2 1614 2 1 2 2 1627 2 1 2 2 1567 2 1 2 2 1495 2 1 2 2 1535 2 1 2 2 1545 2 1 2 2 1607 2 1 2 2

1513 2 1 2 2 1582 2 1 2 2 1601 2 1 2 2 1555 2 1 2 2 1607 2 1 2 2 1586 2 1 2 2 1524 2 1 2 2 1656 2 1 2 2 154 2 1 2 2 1454 2 1 2 2 1669 2 1 2 2 1631 2 1 2 2 1608 2 1 2 2 1497 2 1 2 2 1639 2 1 2 2 1575 2 1 2 2 1567 2 1 2 2 1528 2 1 2 2 1511 2 1 2 2 1627 2 1 2 2 1612 2 1 2 2 1536 2 1 2 2 1568 1 1 3 2 152 1 1 3 2 1486 1 1 3 2 1441 1 1 3 2 1706 1 1 3 2 144 1 1 3 2 15 1 1 3 2 1537 1 1 3 2 1387 1 1 3 2 1463 1 1 3 2 157 1 1 3 2 1429 1 1 3 2 1667 1 1 3 2 1632 1 1 3 2 1589 1 1 3 2 1474 2 1 3 2 1543 2 1 3 2 1555 2 1 3 2 1599 2 1 3 2 1494 2 1 3 2 1624 2 1 3 2 1577 2 1 3 2 1593 2 1 3 2 1479 2 1 3 2 1517 2 1 3 2 1538 2 1 3 2 159 2 1 3 2 1479 2 1 3 2 1479 2 1 3 2 1522 2 1 3 2 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

60

Table C-2 Example OpenBUGS Certification Summary for Component A

mean sd MC_error val25pc median val975pc start sample b 2584 02948 0005376 2001 2588 3158 100001 100000 mu[1] 1417 02706 0005366 1364 1416 147 100001 100000 mu[2] 1559 008368 4776E-4 1543 1559 1576 100001 100000 mu[3] 1533 01504 0001419 1504 1533 1563 100001 100000

Certification Data for Component B list(Ntot=278) lamb[] asy[] cmp[] typ[] pip[] 1869 1 2 1 1 185 2 2 1 1 1709 3 2 1 1 1856 4 2 1 1 1848 5 2 1 1 1828 6 2 1 1 1861 7 2 1 1 1825 8 2 1 1 1871 9 2 1 1 1776 10 2 1 1 196 1 2 1 1 1781 2 2 1 1 1707 3 2 1 1 1787 4 2 1 1 1829 5 2 1 1 1826 6 2 1 1 1858 7 2 1 1 1869 8 2 1 1 1773 9 2 1 1 1902 10 2 1 1 2174 1 2 1 1 2202 2 2 1 1 2172 3 2 1 1 223 5 2 1 1 2094 6 2 1 1 1949 7 2 1 1 1971 8 2 1 1 2123 9 2 1 1 2107 10 2 1 1 2231 1 2 1 1 2292 2 2 1 1 2216 3 2 1 1 2078 4 2 1 1 2192 5 2 1 1 2034 6 2 1 1 211 7 2 1 1 2123 8 2 1 1 2066 9 2 1 1 2046 10 2 1 1 181 1 2 1 1 1865 2 2 1 1 1806 3 2 1 1 1878 4 2 1 1 1824 5 2 1 1 1768 6 2 1 1 1811 7 2 1 1 1763 8 2 1 1 1865 9 2 1 1 1815 10 2 1 1

1806 1 2 1 1 1801 2 2 1 1 1786 3 2 1 1 1768 4 2 1 1 179 5 2 1 1 167 6 2 1 1 1749 7 2 1 1 1741 8 2 1 1 1781 9 2 1 1 1866 10 2 1 1 1784 1 2 1 1 1735 2 2 1 1 1778 3 2 1 1 1835 4 2 1 1 1794 5 2 1 1 17 6 2 1 1 1799 7 2 1 1 1765 8 2 1 1 1754 9 2 1 1 1792 10 2 1 1 1821 1 2 1 1 1813 2 2 1 1 1813 3 2 1 1 1749 4 2 1 1 1848 5 2 1 1 1711 6 2 1 1 1779 7 2 1 1 1769 8 2 1 1 1725 9 2 1 1 1814 10 2 1 1 1718 1 2 1 1 1753 2 2 1 1 1773 3 2 1 1 1777 4 2 1 1 1803 5 2 1 1 1718 6 2 1 1 1802 7 2 1 1 1834 8 2 1 1 1731 9 2 1 1 1842 10 2 1 1 1768 1 2 1 1 1698 2 2 1 1 1792 3 2 1 1 174 4 2 1 1 1795 5 2 1 1 1693 6 2 1 1 1743 7 2 1 1 1745 8 2 1 1 1714 9 2 1 1 1791 1 2 1 2

1746 2 2 1 2 1803 3 2 1 2 1767 4 2 1 2 1745 5 2 1 2 1824 6 2 1 2 1752 7 2 1 2 1813 8 2 1 2 1648 9 2 1 2 1818 10 2 1 2 177 1 2 1 2 182 2 2 1 2 1803 3 2 1 2 1739 4 2 1 2 1773 5 2 1 2 1729 6 2 1 2 1744 7 2 1 2 1754 8 2 1 2 1671 9 2 1 2 1741 10 2 1 2 179 1 2 2 2 1712 1 2 2 2 1693 1 2 2 2 1754 1 2 2 2 1757 1 2 2 2 1697 1 2 2 2 1722 1 2 2 2 1884 1 2 2 2 1713 1 2 2 2 1759 1 2 2 2 1749 1 2 2 2 1748 1 2 2 2 1632 1 2 2 2 1752 1 2 2 2 1871 1 2 2 2 1688 1 2 2 2 1789 1 2 2 2 1826 1 2 2 2 1751 1 2 2 2 1738 1 2 2 2 1742 1 2 2 2 1761 1 2 2 2 1909 1 2 2 2 1719 1 2 2 2 1755 1 2 2 2 1745 1 2 2 2 1687 1 2 2 2 1709 1 2 2 2 1625 1 2 2 2 1802 1 2 2 2 1734 1 2 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

61

1809 1 2 2 2 1782 1 2 2 2 1712 1 2 2 2 1668 1 2 2 2 1535 1 2 2 2 1835 1 2 2 2 1725 1 2 2 2 1769 1 2 2 2 178 1 2 2 2 1759 1 2 2 2 1678 1 2 2 2 1836 1 2 2 2 1743 1 2 2 2 1885 1 2 2 2 1704 1 2 2 2 1749 1 2 2 2 1739 1 2 2 2 1735 1 2 2 2 1714 1 2 2 2 1703 1 2 2 2 1772 1 2 2 2 1727 1 2 2 2 1738 1 2 2 2 174 1 2 2 2 1718 1 2 2 2 1701 1 2 2 2 1673 1 2 2 2 1887 1 2 2 2 1743 1 2 2 2 1825 1 2 2 2 1849 1 2 2 2 188 1 2 2 2 1768 1 2 2 2 1713 1 2 2 2 166 1 2 2 2 1775 2 2 2 2 1824 2 2 2 2 167 2 2 2 2 1683 2 2 2 2 1693 2 2 2 2 169 2 2 2 2 1864 2 2 2 2 1709 2 2 2 2 1764 2 2 2 2

1741 2 2 2 2 1797 2 2 2 2 1626 2 2 2 2 1724 2 2 2 2 177 2 2 2 2 1744 2 2 2 2 1716 2 2 2 2 1805 2 2 2 2 1827 2 2 2 2 1752 2 2 2 2 1746 2 2 2 2 1666 2 2 2 2 187 2 2 2 2 1718 2 2 2 2 1483 2 2 2 2 1632 2 2 2 2 1759 2 2 2 2 1723 2 2 2 2 1684 2 2 2 2 1887 2 2 2 2 1756 2 2 2 2 1816 2 2 2 2 1792 2 2 2 2 1614 2 2 2 2 1566 2 2 2 2 1741 2 2 2 2 1752 2 2 2 2 1697 2 2 2 2 1762 2 2 2 2 1801 2 2 2 2 1766 2 2 2 2 1747 2 2 2 2 1744 2 2 2 2 1692 2 2 2 2 186 2 2 2 2 1756 2 2 2 2 1896 2 2 2 2 1814 2 2 2 2 1703 2 2 2 2 1722 2 2 2 2 1688 2 2 2 2 1821 2 2 2 2 1758 2 2 2 2 1787 2 2 2 2

1784 2 2 2 2 179 2 2 2 2 1681 2 2 2 2 1685 2 2 2 2 1798 2 2 2 2 171 2 2 2 2 1638 2 2 2 2 1931 2 2 2 2 1786 2 2 2 2 1768 2 2 2 2 1764 2 2 2 2 1819 2 2 2 2 1666 1 2 3 2 1626 1 2 3 2 188 1 2 3 2 1653 1 2 3 2 1733 1 2 3 2 1551 1 2 3 2 1884 1 2 3 2 1718 1 2 3 2 1549 1 2 3 2 1723 1 2 3 2 194 1 2 3 2 164 1 2 3 2 1565 1 2 3 2 1787 1 2 3 2 1655 2 2 3 2 167 2 2 3 2 1673 2 2 3 2 1705 2 2 3 2 1755 2 2 3 2 1625 2 2 3 2 1708 2 2 3 2 1705 2 2 3 2 1862 2 2 3 2 1686 2 2 3 2 1723 2 2 3 2 1909 2 2 3 2 1859 2 2 3 2 1897 2 2 3 2 1972 2 2 3 2 END

Table C-3 Example OpenBUGS Certification Summary for Component B

mean sd MC_error val25pc median val975pc start sample b 09337 03387 0007504 02695 09371 1592 100001 100000 mu[1] 1764 03101 0007483 1704 1763 1825 100001 100000 mu[2] 1748 008014 4719E-4 1733 1748 1764 100001 100000 mu[3] 1735 02338 0003093 1688 1736 1781 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

62

Certification data for Component C list(Ntot=280) lamb[] asy[] cmp[] typ[] pip[] 163 1 3 1 1 1596 2 3 1 1 1574 3 3 1 1 159 4 3 1 1 161 5 3 1 1 1589 6 3 1 1 1569 7 3 1 1 1565 8 3 1 1 1564 9 3 1 1 1548 10 3 1 1 1706 1 3 1 1 1576 2 3 1 1 1538 3 3 1 1 1555 4 3 1 1 1607 5 3 1 1 1548 6 3 1 1 1572 7 3 1 1 1681 8 3 1 1 1492 9 3 1 1 1655 10 3 1 1 1499 1 3 1 1 1576 2 3 1 1 1545 3 3 1 1 1584 4 3 1 1 1577 5 3 1 1 1599 6 3 1 1 1587 7 3 1 1 152 8 3 1 1 1479 9 3 1 1 1512 10 3 1 1 1497 1 3 1 1 1611 2 3 1 1 156 3 3 1 1 1546 4 3 1 1 1529 5 3 1 1 1478 6 3 1 1 1504 7 3 1 1 1551 8 3 1 1 1438 9 3 1 1 1494 10 3 1 1 1452 1 3 1 1 1347 2 3 1 1 1395 3 3 1 1 1442 4 3 1 1 1448 5 3 1 1 1463 6 3 1 1 1402 7 3 1 1 1474 8 3 1 1 1466 9 3 1 1 1439 10 3 1 1 1424 1 3 1 1 1397 2 3 1 1 135 3 3 1 1 1459 4 3 1 1 1602 5 3 1 1 1435 6 3 1 1 1433 7 3 1 1 1481 8 3 1 1 1482 9 3 1 1

1568 10 3 1 1 147 1 3 1 1 1411 2 3 1 1 1512 3 3 1 1 1452 4 3 1 1 1567 5 3 1 1 148 6 3 1 1 1442 7 3 1 1 143 8 3 1 1 1413 9 3 1 1 1486 10 3 1 1 1444 1 3 1 1 1427 2 3 1 1 1439 3 3 1 1 1468 4 3 1 1 1524 5 3 1 1 1424 6 3 1 1 1459 7 3 1 1 1467 8 3 1 1 1411 9 3 1 1 148 10 3 1 1 1485 1 3 1 1 1422 2 3 1 1 1438 3 3 1 1 1439 4 3 1 1 1515 5 3 1 1 1409 6 3 1 1 1469 7 3 1 1 1453 8 3 1 1 1465 9 3 1 1 1363 10 3 1 1 1458 1 3 1 1 1459 2 3 1 1 1465 3 3 1 1 1468 4 3 1 1 1457 5 3 1 1 1412 6 3 1 1 1474 7 3 1 1 1428 8 3 1 1 1425 9 3 1 1 1393 10 3 1 1 1357 1 3 1 2 1447 2 3 1 2 1448 3 3 1 2 1409 4 3 1 2 1389 5 3 1 2 1415 6 3 1 2 1431 7 3 1 2 1381 8 3 1 2 1403 9 3 1 2 1435 10 3 1 2 1364 1 3 1 2 1361 2 3 1 2 1404 3 3 1 2 1411 4 3 1 2 1429 5 3 1 2 1417 6 3 1 2 138 7 3 1 2 1336 8 3 1 2 1387 9 3 1 2

1372 10 3 1 2 1493 1 3 2 2 1516 1 3 2 2 1451 1 3 2 2 144 1 3 2 2 1523 1 3 2 2 1437 1 3 2 2 1485 1 3 2 2 1433 1 3 2 2 1463 1 3 2 2 1534 1 3 2 2 1533 1 3 2 2 145 1 3 2 2 1537 1 3 2 2 1515 1 3 2 2 1511 1 3 2 2 1449 1 3 2 2 1388 1 3 2 2 1475 1 3 2 2 1468 1 3 2 2 1549 1 3 2 2 153 1 3 2 2 1442 1 3 2 2 1385 1 3 2 2 1513 1 3 2 2 1528 1 3 2 2 1486 1 3 2 2 1472 1 3 2 2 1493 1 3 2 2 1477 1 3 2 2 1405 1 3 2 2 1499 1 3 2 2 1458 1 3 2 2 1514 1 3 2 2 1525 1 3 2 2 1527 1 3 2 2 149 1 3 2 2 1568 1 3 2 2 1452 1 3 2 2 1421 1 3 2 2 1455 1 3 2 2 1525 1 3 2 2 1531 1 3 2 2 144 1 3 2 2 1487 1 3 2 2 151 1 3 2 2 1577 1 3 2 2 1506 1 3 2 2 152 1 3 2 2 1483 1 3 2 2 1522 1 3 2 2 1463 1 3 2 2 1478 1 3 2 2 1505 1 3 2 2 1454 1 3 2 2 1555 1 3 2 2 1497 1 3 2 2 1515 1 3 2 2 1494 1 3 2 2 1594 1 3 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

63

1471 1 3 2 2 1394 1 3 2 2 1445 1 3 2 2 1446 1 3 2 2 1499 1 3 2 2 1484 1 3 2 2 1436 1 3 2 2 1565 2 3 2 2 149 2 3 2 2 1485 2 3 2 2 1529 2 3 2 2 1513 2 3 2 2 1492 2 3 2 2 1475 2 3 2 2 1497 2 3 2 2 1503 2 3 2 2 1448 2 3 2 2 1488 2 3 2 2 1496 2 3 2 2 1499 2 3 2 2 1469 2 3 2 2 1515 2 3 2 2 1543 2 3 2 2 1452 2 3 2 2 1458 2 3 2 2 1435 2 3 2 2 1499 2 3 2 2 1499 2 3 2 2 1589 2 3 2 2 1562 2 3 2 2 1547 2 3 2 2 1505 2 3 2 2 1508 2 3 2 2 152 2 3 2 2

1481 2 3 2 2 1547 2 3 2 2 1501 2 3 2 2 1491 2 3 2 2 1465 2 3 2 2 147 2 3 2 2 1483 2 3 2 2 1479 2 3 2 2 148 2 3 2 2 1543 2 3 2 2 1448 2 3 2 2 1531 2 3 2 2 1478 2 3 2 2 1465 2 3 2 2 1454 2 3 2 2 1591 2 3 2 2 1466 2 3 2 2 157 2 3 2 2 1471 2 3 2 2 1506 2 3 2 2 1491 2 3 2 2 1461 2 3 2 2 1485 2 3 2 2 1495 2 3 2 2 1516 2 3 2 2 1474 2 3 2 2 1483 2 3 2 2 1519 2 3 2 2 1475 2 3 2 2 1466 2 3 2 2 1565 2 3 2 2 1491 2 3 2 2 1435 2 3 2 2 1459 2 3 2 2

1421 2 3 2 2 1468 2 3 2 2 1556 2 3 2 2 1424 1 3 3 2 1427 1 3 3 2 1381 1 3 3 2 1394 1 3 3 2 1473 1 3 3 2 133 1 3 3 2 1446 1 3 3 2 1424 1 3 3 2 133 1 3 3 2 1452 1 3 3 2 1512 1 3 3 2 1314 1 3 3 2 1525 1 3 3 2 1475 1 3 3 2 1542 1 3 3 2 1461 2 3 3 2 1443 2 3 3 2 1547 2 3 3 2 1394 2 3 3 2 1364 2 3 3 2 1395 2 3 3 2 1522 2 3 3 2 1483 2 3 3 2 1452 2 3 3 2 1482 2 3 3 2 1439 2 3 3 2 1458 2 3 3 2 155 2 3 3 2 1502 2 3 3 2 152 2 3 3 2 END

Table C-4 Example OpenBUGS Certification Summary for Component C

mean sd MC_error val25pc median val975pc start sample b 09626 01701 0002428 06285 09633 1293 100001 100000 mu[1] 1399 01566 0002413 1369 1399 143 100001 100000 mu[2] 1491 005701 2386E-4 148 1491 1503 100001 100000 mu[3] 1449 0132 0001181 1423 1449 1475 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

64

34 Results Table C-5 lists the results in CopiesnL for the three components

Table C-5 Parameter Values for Components A B and C

Posterior Distribution Summary Statistics Component Parameter Mean SD 25 50 975

A

β 260 029 203 260 318 μ1 1415 027 1362 1415 1467 μ2 1559 010 1539 1559 158 μ3 1533 016 1501 1533 1565

B

β 095 034 029 095 162 μ1 1762 031 1701 1762 1822 μ2 1748 010 1728 1748 1769 μ3 1735 024 1687 1735 1783

C

β 097 017 064 097 130 μ1 1399 016 1368 1399 1429 μ2 1491 008 1476 1491 1507 μ3 1449 014 1421 1449 1477

Since in each case the three data set types represent independent measurements the three estimates can be combined to obtain a consensus value In this way a between data set type variability is included in the final standard uncertainty Table C-6 lists the results in Copies per Nanoliter obtained using the NIST Consensus Builder (NICOB)C3

Table C-6 Certified Values for Components A B and C

Component Mean u(Mean) Uk=2(Mean) A 151 07 14 B 175 03 06 C 145 05 10

4 Mito-to-nuclear DNA ratios (mtDNAnDNA)

41 Data Measurements were also obtained on the same samples using three mitochondrial detection assays mtND1 mtBatz and mtKav and a single nDNA assay (genomic) to compute the ratios mtDNAnDNA For the complete mtDNAnDNA data see

Table 16 mtDNAnDNA Ratios

42 Model For each component the data was analyzed using the following random effects model to account for between assay variability

119903119903119894119894~119873119873120572120572119894119894 1205901205901198861198861198861198861198861198861198861198861198861198862

where 119894119894 denotes assay (for A and B 1 = mtND1 2 = mtBatz 3 = mtKav for C 1 = mtBatz 2 = mtKav) and 120572120572119894119894~119873119873micro120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119886119886119886119886119886119886119886119886119886119886

2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

65

43 OpenBUGS code inits and data Two versions of the same model are required one for the results from the three mtDNA assays available for components A and B mtND1 mtBatz and mtKav and the other for the results from the two mtDNA assays available for component C mtBatz and mtKav mtDNAnDNA code for components A and B ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 13)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for components A and B list(sigt=c(111)) Data for Component A list(Ndata=102) set[] rep[] asy[] rat[] 1 1 1 1931 1 2 1 187 1 1 2 1845 1 2 2 1764 2 1 1 181 2 2 1 1906 2 1 2 1785 2 2 2 1837 3 1 1 1743 3 2 1 1761 3 1 2 1829 3 2 2 1847 4 1 1 1716 4 2 1 180 4 1 2 174 4 2 2 1813 5 1 1 1923 5 2 1 1835 5 1 2 1935 5 2 2 1851 6 1 1 1757 6 2 1 1724 6 3 1 1768 6 1 2 1743 6 2 2 1671 6 3 2 1766 6 1 3 1692 6 2 3 1662 6 3 3 1767 7 1 1 1924 7 2 1 1944 7 3 1 1893 7 1 2 1786

7 2 2 1752 7 3 2 1743 7 1 3 153 7 2 3 1542 7 3 3 1523 7 1 2 158 7 2 2 1557 7 3 2 1592 7 1 3 2103 7 2 3 210 7 3 3 2049 7 1 1 1749 7 2 1 1769 7 3 1 1777 7 1 2 1965 7 2 2 1889 7 3 2 1947 7 1 3 1813 7 2 3 1864 7 3 3 1864 8 1 1 178 8 2 1 1776 8 3 1 1706 8 1 2 1804 8 2 2 1751 8 3 2 1738 8 1 3 1769 8 2 3 1786 8 3 3 1761 9 1 2 1584 9 2 2 1544 10 1 1 1947 10 2 1 1928 10 3 1 1922 10 1 1 1829

10 2 1 1895 10 3 1 1851 10 1 1 1674 10 2 1 1617 11 1 2 1667 11 2 2 1813 11 3 2 1708 11 4 2 1598 11 1 2 1566 11 2 2 1687 11 3 2 1663 11 4 2 1589 11 1 2 1643 11 2 2 1668 11 3 2 1604 11 4 2 1573 12 1 2 1719 12 2 2 1716 12 3 2 167 12 4 2 1666 12 1 2 1581 12 2 2 1573 12 3 2 1601 12 4 2 1612 12 1 2 1546 12 2 2 1542 12 3 2 1538 12 4 2 1587 13 1 1 1617 13 2 1 1599 13 1 1 1622 13 2 1 1579 13 1 1 1575 13 2 1 1528 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

66

Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A

mean sd MC_error val25pc median val975pc start sample mu 1730 1443 004419 1703 1730 1759 100001 100000

data for Component B list(Ndata=87) set[] rep[] asy[] rat[] 1 1 1 2461 1 2 1 2401 1 1 2 2555 1 2 2 2608 2 1 1 2229 2 2 1 2200 2 1 2 2108 2 2 2 2192 3 1 1 2127 3 2 1 2198 3 1 2 2455 3 2 2 2506 4 1 1 2211 4 2 1 2199 4 1 2 2380 4 2 2 2249 5 1 1 2105 5 2 1 1943 5 1 2 2136 5 2 2 2119 6 1 1 1990 6 2 1 1921 6 3 1 1982 6 1 2 1990 6 2 2 1924 6 3 2 1958 6 1 3 1903 6 2 3 1831

6 3 3 1919 7 1 1 2304 7 2 1 2341 7 3 1 2332 7 1 2 2048 7 2 2 2078 7 3 2 2067 7 1 3 2399 7 2 3 2479 7 3 3 2493 7 1 1 2065 7 2 1 2166 7 3 1 2164 7 1 2 2314 7 2 2 2280 7 3 2 2258 7 1 3 1867 7 2 3 1987 7 3 3 1971 8 1 1 2046 8 2 1 2057 8 3 1 2024 8 1 2 1998 8 2 2 2043 8 3 2 1952 8 1 3 2077 8 2 3 2117 8 3 3 2041 9 1 2 1890 9 2 2 1886

10 1 2 1928 10 2 2 1916 10 3 2 1915 10 1 2 1947 10 2 2 1946 10 3 2 1961 10 4 2 1909 10 1 2 1953 10 2 2 1952 10 3 2 1941 10 4 2 1881 11 1 2 1865 11 2 2 1863 11 3 2 1828 11 4 2 1828 11 1 2 1944 11 2 2 1964 11 3 2 1947 11 4 2 1895 11 1 2 1908 11 2 2 1909 11 3 2 1940 11 4 2 1839 12 1 1 1688 12 2 1 1783 12 1 1 1823 12 2 1 1631 12 1 1 1920 12 2 1 1887 END

Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B

mean sd MC_error val25pc median val975pc start sample mu 2045 2246 0111 1999 2045 2088 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

67

mtDNAnDNA code for component C ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 12)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for component C list(sigt=c(11)) mtDNAnDNA data for component C list(Ndata=66) set[] rep[] asy[] rat[] 1 1 1 3201 1 2 1 3299 2 1 1 3347 2 2 1 3287 3 1 1 3217 3 2 1 3384 4 1 1 3200 4 2 1 3363 5 1 1 3294 5 2 1 3013 6 1 1 2805 6 2 1 2825 6 3 1 2829 6 1 2 2709 6 2 2 2738 6 3 2 2727 7 1 1 2797 7 2 1 2642 7 3 1 2792 7 1 2 2804 7 2 2 2789

7 3 2 2842 7 1 1 2715 7 2 1 2606 7 3 1 2687 7 1 2 2799 7 2 2 2628 7 3 2 2559 7 1 1 2639 7 2 1 2535 7 3 1 2657 7 1 2 2953 7 2 2 2757 7 3 2 2802 8 1 1 2927 8 2 1 2776 8 3 1 2762 8 1 2 3044 8 2 2 2830 8 3 2 2825 9 1 1 2501 9 2 1 2522 10 1 1 2832 10 2 1 2840

10 3 1 2788 10 4 1 2725 10 1 1 2835 10 2 1 2790 10 3 1 2736 10 4 1 2679 10 1 1 2803 10 2 1 2773 10 3 1 2757 10 4 1 2752 11 1 1 2599 11 2 1 2508 11 3 1 2551 11 4 1 2609 11 1 1 2617 11 2 1 2483 11 3 1 2626 11 4 1 2602 11 1 1 2665 11 2 1 2535 11 3 1 2570 11 4 1 2650 END

Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C

mean sd MC_error val25pc median val975pc start sample mu 2774 2367 01176 2729 2774 2821 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

68

Table C-10 Results for mtDNAnDNA Posterior Distribution Summary Statistics

Component Mean Std Dev 25 50 975 A 1730 14 1703 1730 1758 B 2045 23 1994 2047 2087 C 2774 23 2729 2774 2821

5 ReferencesC1 Lunn DJ Spiegelhalter D Thomas A Best N Statistics in Medicine 2009283049ndash3082C2 Polson NG Scott JG Bayesian Analysis 20127887ndash902C3 NIST Consensus Builder Software and userrsquos guide freely available at

httpsconsensusnistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

69

Appendix D Performance in Commercial qPCR Chemistries The performance of the SRM 2372a materials in commercial qPCR chemistries was assessed using manufacturerrsquos recommended protocols Each panel in Figures D-1 to D-4 displays the average cycle threshold (Ct) as a function of DNA concentration for the 5-point dilutions of SRM 2372a components In all cases the external calibrant was SRM 2372 component A Error bars represent approximate 95 confidence intervals Table D-1 lists the estimated [DNA] for the three SRM 2372a components for all commercial assays evaluated This Table also lists the Component BComponent A (BA) and Component CComponent A (CA) ratios calculated from the [DNA] estimates

Figure D-1 Results for the Innogenomics Commercial qPCR assays

20

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantLong(IQ_h)

ABCStd

[A] = (384 plusmn 26) ngμL[B] = (441 plusmn 24) ngμL[C] = (516 plusmn 24) ngμL

Rsup2 = 0999 se(Ct) = 01016

17

18

19

20

21

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantShort(IQ_d)

ABCStd

[A] = (435 plusmn 28) ngμL[B] = (543 plusmn 12) ngμL[C] = (510 plusmn 28) ngμL

Rsup2 = 0999 se(Ct) = 007

24

25

26

27

28

29

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYLong(IQHY_h)

ABCStd

[A] = (387 plusmn 24) ngμL[B] = (448 plusmn 21) ngμL[C] = (559 plusmn 31) ngμL

Rsup2 = 0998 se(Ct) = 01314

16

18

20

22

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYShort(IQHY_d)

ABCStd

[A] = (353 plusmn 23) ngμL[B] = (267 plusmn 15) ngμL[C] = (390 plusmn 18) ngμL

Rsup2 = 1000 se(Ct) = 00420

22

24

26

28

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYY(IQHY_y)

ABCStd

[A] = (572 plusmn 13) ngμL[B] = 0 ngμL[C] = (137 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

70

Figure D-2 Results for the Thermo Fisher Commercial qPCR assays

21

22

23

24

25

26

27

28

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HumanHuman(QFH_h)

ABCStd

[A] = (464 plusmn 13) ngμL[B] = (529 plusmn 19) ngμL[C] = (472 plusmn 22) ngμL

Rsup2 = 1000 se(Ct) = 004

22

23

24

25

26

27

28

29

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoHuman(QFD_h)

ABCStd

[A] = (701 plusmn 53) ngμL[B] = (561 plusmn 11) ngμL[C] = (472 plusmn 11) ngμL

Rsup2 = 0994 se(Ct) = 01922

24

26

28

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoMale(QFD_y)

ABCStd

[A] = (706 plusmn 49) ngμL[B] = 0 ngμL[C] = (122 plusmn 02) ngμL

Rsup2 = 0999 se(Ct) = 007

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPLarge(QFHP_h)

ABCStd

[A] = (340 plusmn 18) ngμL[B] = (432 plusmn 05) ngμL[C] = (342 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 00720

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPSmall(QFHP_d)

ABCStd

[A] = (434 plusmn 15) ngμL[B] = (514 plusmn 10) ngμL[C] = (466 plusmn 07) ngμL

Rsup2 = 1000 se(Ct) = 005

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioLarge(QFT_h)

ABCStd

[A] = (362 plusmn 23) ngμL[B] = (430 plusmn 06) ngμL[C] = (378 plusmn 07) ngμL

Rsup2 = 0999 se(Ct) = 00920

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioSmall(QFT_d)

ABCStd

[A] = (468 plusmn 08) ngμL[B] = (505 plusmn 08) ngμL[C] = (523 plusmn 14) ngμL

Rsup2 = 1000 se(Ct) = 00318

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioY(QFT_y)

ABCStd

[A] = (372 plusmn 11) ngμL[B] = 0 ngμL[C] = (110 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 003

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

71

Figure D-3 Results for the Qiagen Commercial qPCR assays

Figure D-4 Results for the Promega Commercial qPCR assays

19

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresAutosomal(QPH_h)

ABCStd

[A] = (554 plusmn 21) ngμL[B] = (479 plusmn 05) ngμL[C] = (381 plusmn 09) ngμL

Rsup2 = 1000 se(Ct) = 00418

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresY(QPH_y)

ABCStd

[A] = (358 plusmn 10) ngμL[B] = 0 ngμL[C] = (118 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 007

16

17

18

19

20

21

22

23

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProHuman(QPP_h)

ABCStd

[A] = (525 plusmn 49) ngμL[B] = (480 plusmn 42) ngμL[C] = (467 plusmn 45) ngμL

Rsup2 = 0999 se(Ct) = 01019

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProDegradation(QPP_d)

ABCStd

[A] = (567 plusmn 18) ngμL[B] = (558 plusmn 08) ngμL[C] = (484 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 00410

15

20

25

30

35

-50 -40 -30 -20 -10 00 10

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProY(QPP_y)

ABCStd

[A] = (538 plusmn 27) ngμL[B] = (00 plusmn 00) ngμL[C] = (162 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 003

16

18

20

22

24

-05 00 05 10 15 20

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorAutosomal(P_h)

ABCStd

[A] = (759 plusmn 64) ngμL[B] = (775 plusmn 24) ngμL[C] = (1056 plusmn 17) ngμL

Rsup2 = 0995 se(Ct) = 01818

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorY(P_y)

ABCStd

[A] = (596 plusmn 51) ngμL[B] = 0 ngμL[C] = (117 plusmn 01) ngμL

Rsup2 = 0994 se(Ct) = 020

18

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantAutosomal(PQ_h)

ABCStd

[A] = (436 plusmn 13) ngμL[B] = (606 plusmn 20) ngμL[C] = (504 plusmn 10) ngμL

Rsup2 = 1000 se(Ct) = 00218

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantDegradation(PQ_d)

ABCStd

[A] = (420 plusmn 18) ngμL[B] = (529 plusmn 16) ngμL[C] = (530 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 00516

18

20

22

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantY(PQ_y)

ABCStd

[A] = (410 plusmn 20) ngμL[B] = 0 ngμL[C] = (139 plusmn 05) ngμL

Rsup2 = 0999 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

72

Table D-1 Summary [DNA] Results for the Commercial qPCR Assays Estimated [DNA] ngμL Ratios

A B C BA CA Assaya xb u(x)c xb u(x)c xb u(x)c yd u(y)e yd u(y)e

IQd 435 28 543 12 510 28 125 008 117 010 IQh 384 26 441 24 516 24 115 010 134 011

IQHYd 353 23 267 15 390 18 076 006 111 009 IQHYh 387 24 448 21 559 31 116 009 144 012

Ph 759 64 775 24 1056 16 102 009 139 012 PQd 420 18 529 16 530 06 126 007 126 006 PQh 436 13 606 20 504 10 139 006 116 004

QFDh 701 53 561 11 472 11 080 006 067 005 QFHh 464 13 529 19 472 22 114 005 102 006

QFHPd 434 15 514 10 466 07 118 005 107 004 QFHPh 340 18 432 05 342 03 127 007 101 005

QFTd 468 08 505 08 523 14 108 003 112 004 QFTh 362 23 430 06 378 07 118 008 104 007 QPHh 554 21 479 05 380 09 086 003 069 003 QPPd 567 18 558 08 484 11 098 003 085 003 QPPh 525 49 480 42 467 45 092 012 089 012

IQHYy 572 13 NDf 137 06

NRg

024 001 Py 596 51 NDf 117 01 020 002

PQy 410 20 NDf 139 05 034 002 QFDy 706 49 NDf 122 02 017 001 QFTy 372 11 NDf 110 06 030 002 QPHy 358 10 NDf 118 03 033 001 QPPy 538 27 0002 0001 162 11 030 003

a Assay codes are listed in Table 8 b Mean of dilution-adjusted results from five dilutions of the SRM 2372a component c Standard uncertainty of the mean estimated from the stdanrd deviation of the five dilutions d Ratio of the estimated [DNA] of component B or C relative to that of component A e Standard uncertainty of the ratio estimated from the standard uncertainties of the individual components f Not detected Component B was prepared using tissue from only female donors it is not expected to

respond to male-specific assays g Not relevant

  • Abstract
  • Keywords
  • Technical Information Contact for this SRM
  • Table of Contents
  • Table of Tables
  • Table of Figures
  • Purpose and Description
  • Warning SRM 2372a is a Human Source Material
  • Storage and Use
  • History and Background
    • Figure 1 Sales History of SRM 2372
      • Experimental Methods
        • Polymerase Chain Reaction Assays
          • Table 1 NIST-Developed Human Nuclear DNA Assays
          • Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays
          • Table 2 NIST-Optimized Human Mitochondrial DNA Assays
          • Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays
            • Droplet Digital Polymerase Chain Reaction Measurements
              • Nuclear DNA (nDNA)
                • Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays
                  • Mitochondrial DNA (mtDNA)
                    • Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays
                    • Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays
                        • Sample Preparation
                        • Absorbance Measurements
                        • SRM Unit Production
                        • Component Homogeneity
                          • Figure 4 Plate Layout Design for Homogeneity Measurements
                            • Component Stability
                              • Figure 5 Plate Layout Design for Stability Measurements
                                • Component Certification Measurements
                                  • Table 6 Tubes used for Certification Measurements
                                  • Figure 6 Basic Plate Layout Design for Certification Measurements
                                    • Mitochondrial DNA Measurements
                                      • Table 7 Run Parameters for Mitochondrial DNA
                                        • Performance in Commercial qPCR Chemistries
                                          • Table 8 Commercial qPCR Chemistries Tested with SRM 2372a
                                          • Figure 7 Plate Layout Design for Performance Assessments
                                              • Measurement Results
                                                • Absorbance Spectrophotometry
                                                  • Table 9 Absorbance at Selected Wavelengths
                                                  • Figure 8 Absorbance Spectra of SRM 2372a Components A B and C
                                                    • Droplet Volume
                                                    • Certification Measurements
                                                      • Homogeneity Results
                                                        • Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component
                                                        • Table 11 Within- and Between-Tube Relative Standard Deviations
                                                          • Stability Results
                                                            • Figure 9 ddPCR Stability Measurements as Function of Time
                                                            • Table 12 Stability Data as λacute Copy Number per Nanoliter of Component
                                                              • Ten-Assay Certification Results
                                                                • Figure 10 Comparison of Assay Performance
                                                                • Table 13 Certification Data as λacute Copy Number per Nanoliter of Component
                                                                  • Estimation of Copy Number Concentration
                                                                    • Table 14 Measured Copy Number Concentrations
                                                                      • Conversion of Copy Number to Mass Concentration
                                                                        • Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution
                                                                            • Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)
                                                                              • Figure 11 mtDNAnDNA for Components A B and C
                                                                              • Table 16 mtDNAnDNA Ratios
                                                                              • Table 17 mtDNAnDNA Summary Results
                                                                                • Performance in Commercial qPCR Chemistries
                                                                                  • Figure 12 qPCR Autosomal Quantitation Performance Summary
                                                                                  • Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary
                                                                                      • Qualification of ddPCR as Fit-For-Purpose
                                                                                        • Figure 14 Absorbance Spectra of SRM 2372 Components A and B
                                                                                        • Digital Polymerase Chain Reaction Assays
                                                                                        • Heat Treatments (Melting of the DNA into Single Strands)
                                                                                        • Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a
                                                                                          • Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a
                                                                                            • Quantitative Evaluation of ssDNA Content
                                                                                              • ddPCR Measurement of ssDNA Proportion
                                                                                              • Proof of Principle
                                                                                                • Table 19 Comparison of Snap-Cooling to Slow-Cooling
                                                                                                  • Time at Denaturation Temperature
                                                                                                    • Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures
                                                                                                      • Temperature
                                                                                                        • Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a
                                                                                                          • DNA Concentration in the ddPCR Solution
                                                                                                            • Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution
                                                                                                              • Equivalence of Assays
                                                                                                                • Table 23 Comparison of Assays
                                                                                                                    • ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials
                                                                                                                      • Table 24 Evaluation of Limiting Ratio
                                                                                                                          • Conclusions and Recommendations
                                                                                                                            • Recommended Certification Values
                                                                                                                              • Table 25 Recommended Values and 95 Uncertainties for Certification
                                                                                                                                • Use of SRM 2372a for Value-Assigning In-House Reference Materials
                                                                                                                                  • Certificate of Analysis
                                                                                                                                  • References
                                                                                                                                  • Appendix A DNA Extraction Method
                                                                                                                                    • Procedure
                                                                                                                                    • Reagents
                                                                                                                                    • Reference
                                                                                                                                      • Appendix B Spectrophotometric Measurement Protocol
                                                                                                                                        • Spectrophotometer Calibration
                                                                                                                                          • Establish Baseline
                                                                                                                                          • Enable Verification of Wavelength Calibration
                                                                                                                                          • Enable Verification of Absorbance Calibration
                                                                                                                                            • For ldquonativerdquo dsDNA Samples
                                                                                                                                              • Verify Optical Balance of Cuvettes
                                                                                                                                                • For NaOH denatured ssDNA Samples
                                                                                                                                                  • Prepare a 04 molL NaOH Stock Solution
                                                                                                                                                  • Prepare the Sample Solutions
                                                                                                                                                  • Measure Sample Solutions
                                                                                                                                                    • After All Sample Measurements are Complete
                                                                                                                                                      • Enable Verification of Spectrophotometer Stability
                                                                                                                                                      • Clean Up
                                                                                                                                                        • Evaluation
                                                                                                                                                        • Reference
                                                                                                                                                          • Appendix C SRM 2372a Statisticianrsquos report
                                                                                                                                                            • 1 Introduction
                                                                                                                                                            • 11 OpenBUGS Analysis
                                                                                                                                                            • 12 OpenBUGS Exemplar Output
                                                                                                                                                            • 2 Homogeneity assessment
                                                                                                                                                            • 21 Data
                                                                                                                                                            • 22 Model
                                                                                                                                                            • 23 OpenBUGS code and data
                                                                                                                                                              • Table C-1 Example OpenBUGS Homogeneity Summary
                                                                                                                                                                • 24 Results
                                                                                                                                                                  • Figure C-1 Homogeneity Assessment
                                                                                                                                                                    • 3 Certification measurements
                                                                                                                                                                    • 31 Data
                                                                                                                                                                    • 32 Model
                                                                                                                                                                    • 33 OpenBUGS code inits and data
                                                                                                                                                                      • Table C-2 Example OpenBUGS Certification Summary for Component A
                                                                                                                                                                      • Table C-3 Example OpenBUGS Certification Summary for Component B
                                                                                                                                                                      • Table C-4 Example OpenBUGS Certification Summary for Component C
                                                                                                                                                                        • 34 Results
                                                                                                                                                                          • Table C-5 Parameter Values for Components A B and C
                                                                                                                                                                          • Table C-6 Certified Values for Components A B and C
                                                                                                                                                                            • 4 Mito-to-nuclear DNA ratios (mtDNAnDNA)
                                                                                                                                                                            • 41 Data
                                                                                                                                                                            • 42 Model
                                                                                                                                                                            • 43 OpenBUGS code inits and data
                                                                                                                                                                              • Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A
                                                                                                                                                                              • Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B
                                                                                                                                                                              • Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C
                                                                                                                                                                              • Table C-10 Results for mtDNAnDNA
                                                                                                                                                                                • 5 References
                                                                                                                                                                                  • Appendix D Performance in Commercial qPCR Chemistries
                                                                                                                                                                                    • Table D-1 Summary [DNA] Results for the Commercial qPCR Assays
Page 4: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

i

Abstract Standard Reference Material (SRM) 2372a is intended for use in the value assignment of human genomic deoxyribonucleic acid (DNA) forensic quantitation materials A unit of SRM 2372a consists of three well-characterized human genomic DNA materials solubilized in 10 mmolL 2-amino-2-(hydroxymethyl)-13 propanediol hydrochloride (Tris HCl) and 01 mmolL ethylenediaminetetraacetic acid disodium salt (disodium EDTA) pH 80 buffer (TE-4) The three component genomic DNA materials labeled A B and C are respectively derived from a single male donor a single female donor and 13 mixture of a male and a female donor A unit of the SRM consists of one 05 mL tube of each component each tube containing approximately 55 microL of DNA solution Each of these tubes is labeled and is sealed with a color-coded screw cap This publication documents the production analytical methods and statistical evaluations involved in production of this SRM

Keywords Human mitochondrial DNA (mtDNA)

Human nuclear DNA (nDNA) Standard Reference Material (SRM)

Technical Information Contact for this SRM Please address technical questions about this SRM to srmsnistgov where they will be assigned to the appropriate Technical Project Leader responsible for support of this material For sales and customer service inquiries please contact srminfonistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

ii

Table of Contents Purpose and Description 1 Warning SRM 2372a is a Human Source Material 1 Storage and Use 1 History and Background 2 Experimental Methods 4

Polymerase Chain Reaction Assays 4 Droplet Digital Polymerase Chain Reaction Measurements 7

Nuclear DNA (nDNA) 7 Mitochondrial DNA (mtDNA) 8

Sample Preparation 9 Absorbance Measurements 10 SRM Unit Production 10 Component Homogeneity 10 Component Stability 11 Component Certification Measurements 12 Mitochondrial DNA Measurements 13 Performance in Commercial qPCR Chemistries 13

Measurement Results 15 Absorbance Spectrophotometry 15 Droplet Volume 16 Certification Measurements 16

Homogeneity Results 16 Stability Results 18 Ten-Assay Certification Results 20 Estimation of Copy Number Concentration 22 Conversion of Copy Number to Mass Concentration 22

Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA) 23 Performance in Commercial qPCR Chemistries 25

Qualification of ddPCR as Fit-For-Purpose 27 Digital Polymerase Chain Reaction Assays 28 Heat Treatments (Melting of the DNA into Single Strands) 28 Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a 29 Quantitative Evaluation of ssDNA Content 30

ddPCR Measurement of ssDNA Proportion 30 Proof of Principle 31 Time at Denaturation Temperature 33 Temperature 35 DNA Concentration in the ddPCR Solution 36 Equivalence of Assays 38

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials 40 Conclusions and Recommendations 43

Recommended Certification Values 43 Use of SRM 2372a for Value-Assigning In-House Reference Materials 44

Certificate of Analysis 44 References 44

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

iii

Appendix A DNA Extraction Method 46 Procedure 46 Reagents 46 Reference 47

Appendix B Spectrophotometric Measurement Protocol 48 Spectrophotometer Calibration 48

Establish Baseline 48 Enable Verification of Wavelength Calibration 48 Enable Verification of Absorbance Calibration 48

For ldquonativerdquo dsDNA Samples 48 Verify Optical Balance of Cuvettes 48

For NaOH denatured ssDNA Samples 49 Prepare a 04 molL NaOH Stock Solution 49 Prepare the Sample Solutions 49 Measure Sample Solutions 50

After All Sample Measurements are Complete 50 Enable Verification of Spectrophotometer Stability 50 Clean Up 50

Evaluation 50 Appendix C SRM 2372a Statisticianrsquos report 51

1 Introduction 51 11 OpenBUGS Analysis 51 12 OpenBUGS Exemplar Output 51 2 Homogeneity assessment 52 21 Data 52 22 Model 52 23 OpenBUGS code and data 52 24 Results 55 3 Certification measurements 57 31 Data 57 32 Model 57 33 OpenBUGS code inits and data 58 34 Results 64 4 Mito-to-nuclear DNA ratios (mtDNAnDNA) 64 41 Data 64 42 Model 64 43 OpenBUGS code inits and data 65 5 References 68

Appendix D Performance in Commercial qPCR Chemistries 69

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

iv

Table of Tables Table 1 NIST-Developed Human Nuclear DNA Assays 5 Table 2 NIST-Optimized Human Mitochondrial DNA Assays 6 Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays 8 Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays 8 Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays 8 Table 6 Tubes used for Certification Measurements 12 Table 7 Run Parameters for Mitochondrial DNA 13 Table 8 Commercial qPCR Chemistries Tested with SRM 2372a 14 Table 9 Absorbance at Selected Wavelengths 15 Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component 17 Table 11 Within- and Between-Tube Relative Standard Deviations 18 Table 12 Stability Data as λacute Copy Number per Nanoliter of Component 19 Table 13 Certification Data as λacute Copy Number per Nanoliter of Component 21 Table 14 Measured Copy Number Concentrations 22 Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution 23 Table 16 mtDNAnDNA Ratios 24 Table 17 mtDNAnDNA Summary Results 25 Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a 29 Table 19 Comparison of Snap-Cooling to Slow-Cooling 32 Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures 34 Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 36 Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution 37 Table 23 Comparison of Assays 39 Table 24 Evaluation of Limiting Ratio 41 Table 25 Recommended Values and 95 Uncertainties for Certification 43 Table C-1 Example OpenBUGS Homogeneity Summary 55 Table C-2 Example OpenBUGS Certification Summary for Component A 60 Table C-3 Example OpenBUGS Certification Summary for Component B 61 Table C-4 Example OpenBUGS Certification Summary for Component C 63 Table C-5 Parameter Values for Components A B and C 64 Table C-6 Certified Values for Components A B and C 64 Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A 66 Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B 66 Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C 67 Table C-10 Results for mtDNAnDNA 68 Table D-1 Summary [DNA] Results for the Commercial qPCR Assays 72

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

i

Table of Figures Figure 1 Sales History of SRM 2372 3 Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays 6 Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays 7 Figure 4 Plate Layout Design for Homogeneity Measurements 11 Figure 5 Plate Layout Design for Stability Measurements 11 Figure 6 Basic Plate Layout Design for Certification Measurements 12 Figure 7 Plate Layout Design for Performance Assessments 14 Figure 8 Absorbance Spectra of SRM 2372a Components A B and C 16 Figure 9 ddPCR Stability Measurements as Function of Time 18 Figure 10 Comparison of Assay Performance 20 Figure 11 mtDNAnDNA for Components A B and C 24 Figure 12 qPCR Autosomal Quantitation Performance Summary 26 Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary 26 Figure 14 Absorbance Spectra of SRM 2372 Components A and B 27 Figure 15 Comparison of Snap-Cooling to Slow-Cooling 32 Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures 33 Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 35 Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution 37 Figure 19 Comparison of Assays 38 Figure 20 Evaluation of Limiting Ratio 40 Figure 21 Degrees of Equivalence 42 Figure C-1 Homogeneity Assessment 56 Figure D-1 Results for the Innogenomics Commercial qPCR assays 69 Figure D-2 Results for the Thermo Fisher Commercial qPCR assays 70 Figure D-3 Results for the Qiagen Commercial qPCR assays 70 Figure D-4 Results for the Promega Commercial qPCR assays 71

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

1

Purpose and Description Standard Reference Material (SRM) 2372a is intended for use in the value assignment of human genomic deoxyribonucleic acid (DNA) forensic quantitation materials A unit of SRM 2372a consists of three well-characterized human genomic DNA materials solubilized in 10 mmolL 2-amino-2-(hydroxymethyl)-13 propanediol hydrochloride (Tris HCl) and 01 mmolL ethylenediaminetetraacetic acid disodium salt (disodium EDTA) pH 80 buffer (TE-4) The three component genomic DNA materials labeled A B and C are respectively derived from a single male donor a single female donor and 13 mixture of a male and a female donor A unit of the SRM consists of one 05 mL tube of each component each tube containing approximately 55 microL of DNA solution Each of these tubes is labeled and is sealed with a color-coded screw cap NIST is guided by and adheres to the ethical principles set forth in the Belmont Report1 SRM 2372a was developed after an appropriate human subject research determination by the NIST Human Subjects Protections Office

Warning SRM 2372a is a Human Source Material The three SRM 2372a components are human genomic DNA extracts prepared at NIST from buffy coat white blood cells from single-source anonymous donors All buffy coats were purchased from Interstate Blood Bank (Memphis TN) The supplier reported that each donor unit used in the preparation of this product was tested by FDA-licensed tests and found to be negative for human immunodeficiency virus (HIV-1 RNA and Anti-HIV 12) hepatitis B (HBV DNA and HBsAg) hepatitis C (HCV RNA and Anti-HCV) West Nile virus (WNV RNA) and syphilis Additionally each donor was tested according to FDA guidelines for Trypanosoma cruzi (Chagas) and found to be negativenon-reactive However no known test method can offer complete assurance that infectious agents are absent from this material Accordingly this human blood-based product should be handled at the Biosafety Level 12 SRM 2372a components and derived solutions should be disposed of in accordance with local state and federal regulations

Storage and Use Until required for use SRM 2372a should be stored in the dark between 2 degC and 8 degC The SRM 2372a component tubes should be mixed briefly and centrifuged (without opening the tube cap) prior to removing sample aliquots for analysis For the certified values to be applicable materials should be withdrawn immediately after opening the tubes and processed without delay Certified values do not apply to any material remaining in recapped tubes The certification only applies to the initial use and the same results are not guaranteed if the remaining material is used later

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

2

History and Background SRM 2372a is the second member of the SRM 2372 Human DNA Quantitation Standard series The supply of the initial material SRM 2372 was exhausted in May 2017 SRM 2372a is derived from commercially obtained buffy coat cells from different donors than those used in SRM 2372 A series of interlaboratory studies coordinated by what was then the Biotechnology Division of the National Institute of Standards and Technology (NIST) during the 1990s demonstrated that the DNA quantitation technologies then in use by the forensic human identification community adequately met the needs of the DNA typing tools then available34 However our 2004 study demonstrated that the increasingly sensitive and informative polymerase chain reaction (PCR)-based multiplex typing tools required more reliable determinations of DNA input quantity for best performance5 Quantitative PCR (qPCR) methods for DNA quantitation require calibration to an external standard of defined DNA concentration ([DNA]) expressed in units of nanograms of DNA per microliter of extract solution The true [DNA] of a commercial secondary standard may vary from its nominal concentration depending on supplier and lot number To enable providers of secondary standards to more reliably evaluate their materials and forensic testing laboratories to validate their use of commercial quantitation methods and supplies in 2007 we delivered the human DNA calibration standard SRM 2372 Human DNA Quantitation Standard6 The quantity of DNA in each of the three components of SRM 2372 was originally assigned based on the absorbance of double-stranded DNA (dsDNA) at 260 nm7 However by 2012 the absorbance properties of the materials had increased beyond the certified uncertainty limits due to slow but progressive changes in DNA tertiary structure Sales of the SRM were restricted until the DNA quantities were reassigned based on measurements of single-stranded DNA (ssDNA) in early 20138-10 Figure 1 displays the sales history of this relatively high-selling SRM While SRM 2372 was fit for the practical purpose of harmonizing DNA quantitation measurements to an accessible and reproducible reference material the conversion of both dsDNA and ssDNA absorbance to mass [DNA] is through conventional proportionality constants that are not metrologically traceable to the modern reference system for measurements the International System of Units (SI) Responding to requests from the forensic and clinical communities for calibration standards that provide traceability to the SI we and others have established droplet digital polymerase chain reaction (ddPCR) as a potentially primary measurement process that can provide traceable results11-18 The copy numbers of human nuclear DNA (nDNA) in the reaction mixture per partition symbolized as λ were determined by ddPCR using ten PCR assays optimized for use with our ddPCR platform These assays target loci on eight chromosomes with three assays on widely separated locations on one chromosome (see Table 1)

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

3

Figure 1 Sales History of SRM 2372

Each small black dot represents the sale of one unit of SRM 2372 The red lines summarize the average rate of sales as originally certified (2007 to 2012) and after re-certification (2013 to 2017) The magenta labels state the average yearly sales rate during these two periods

The nDNA copy number concentration human haploid genome equivalents (HHGE) per nanoliter component solution and symbolized here as λacute is calculated as

copy number [DNA] = λ = λ(VF) HHGEnL

where V is the average droplet volume and F is the total dilution factor the volume fraction of the component solution in the ddPCR reaction mixture The manufacturer of the ddPCR platform used a Bio-Rad QX200 Droplet Digital PCR System (Hercules CA) system does not provide metrologically traceable droplet volume information This lack of information prevented metrologically traceable conversion of the number per partition measurements to the desired number per reaction volume units In consequence NIST developed a measurement method that provides metrologically traceable droplet volumes19 We show elsewhere that the mass [DNA] nanograms of human DNA per microliter component can be traceably estimated as

mass [DNA] = 3031timesλacute ngmicroL

where the constant is the product of the number of nucleotide basepairs (bp) in the human reference genome the average bp molecular weight Avogadrorsquos number of bp per mole and

2372

2372(Recertif ied)

050

010

00

1713 unitsyear

1416 unitsyear

2007 2009 2011 2013 2015 2017

Uni

ts S

old

Calendar Year

Human DNA Quantitation StandardSRM 2372

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

4

unit scaling factors18 The three components of SRM 2372a are certified for both copy number and mass [DNA] Mitochondrial DNA (mtDNA) analysis within the forensic community is useful when the nDNA may be degraded or is in low-to-undetectable quantity The challenges of mtDNA testing include contamination of the standard due to routine laboratory use degradation of plasmids or synthetic oligomers or improper commutability of the standard relative to casework samples Commonly cell line materials are used as standards for qPCR with human genomic DNA which may lead to an incorrect mtDNA-to-nDNA ratio (mtDNAnDNA) for unknown samples Non-certified values of mtDNAnDNA for the three components are provided to bridge the current gap between well characterized mitochondrial DNA quantification assays and the availability of commercial standards With the understanding that the forensic DNA community primarily uses qPCR as a method for DNA quantification the performance of the three SRM 2372a materials was evaluated in eleven commercial qPCR chemistries currently used employed within the forensic DNA community

Experimental Methods Production and value-assignment of any certified reference material (CRM) is a complex process In addition to acquiring processing and packaging the materials included in the CRM the materials must be sufficiently homogenous across the units stable over time within its packaging and well-characterized to deliver results that are fit-for-purpose within the intended measurement community20 The process has been particularly complicated for SRM 2372a because we needed to determine the metrological properties of the required ddPCR measurement systems before we could trust the results they provide The following sections describe the methods and processes used to produce and characterize the SRM 2372a materials including sample preparation SRM unit production PCR assays ddPCR measurement processes spectrophotometric absorbance measurements homogeneity testing stability testing ddPCR nDNA and mtDNAnDNA measurements and performance in commercial qPCR assessments Many of these processes were mutually dependent therefore the sections are presented in a ldquodefinitionalrdquo order that sequentially introduces terminology Polymerase Chain Reaction Assays

Ten PCR assays were developed for characterization of SRM 2372a Each assay was confirmed to target one locus per HHGE using the National Center for Biotechnology Informationrsquos BLASTblastn system21 Primers were purchased from Eurofins Operon (Huntsville AL) TaqMan probes labeled with 6-carboxyfluorescein (FAM) were purchased from Thermo Fisher (Waltham MA) FAM-labeled Blackhole Quencher and Blackhole Quencher+ probes were purchased from LGC Biosearch Technologies (Novato CA) Table 1 details the ten nDNA assays used to certify copy number and mass [DNA] of the SRM 2372a components FAM-labeled TaqMan probes were purchased from Thermo Fisher

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

5

and used in monoplex reactions VIC (a proprietary dye)-labeled probes were purchased from the same source and were used when nDNA and mtDNA were analyzed in the same reaction mixture Figure 2 displays exemplar droplet patterns for the assays All assays were developed at NIST and have been optimized for ddPCR18 Table 2 details three mitochondrial DNA (mtDNA) assays used to estimate mtDNAnDNA in the SRM 2372a components Figure 3 displays exemplar droplet patterns for the assays These assays were optimized for ddPCR from real-time quantitative PCR (qPCR) assays from the literature All primers probes were diluted with TE-4 buffer purchased from Affymetrix-USB Products (Affymetrix Inc Cleveland OH) to a nominal 5 micromol working solution for all single-plex assays (individual nDNA and mtDNA assays) For duplex assays (combined nDNA and mtDNA) primer and probe working solutions were diluted to a nominal 10 micromol

Table 1 NIST-Developed Human Nuclear DNA Assays

Assay Target

Chromosome Band Accession Primers and Probe a

Amplicon Length bp

NEIF Gene EIF5B

Chr 2 p111-q111 NC_00000212

F gccaaacttcagccttctcttc R ctctggcaacatttcacactaca PB+ tcatgcagttgtcagaagctg

67

2PR4 Gene RPS27A

Chr 2 p16 NC_00000212

F cgggtttgggttcaggtctt R tgctacaatgaaaacattcagaagtct PB tttgtctaccacttgcaaagctggccttt

97

POTP STR TPOX

Chr 2 p253 NC_00000212

F ccaccttcctctgcttcacttt R acatgggtttttgcctttgg PT caccaactgaaatatg

60

NR4Q Gene DCK

Chr 4 q133-q211 NC_00000412

F tggtgggaatgttcttcagatga R tcgactgagacaggcatatgtt PB+ tgtatgagaaacctgaacgatggt

83

D5 STR D5S2500

Chr 5 q112 NC_00000510

F ttcatacaggcaagcaatgcat R cttaaagggtaaatgtttgcagtaatagat PT ataatatcagggtaaacaggg

75

ND6 STR D6S474

Chr 6 q21-22 NC_00000612

F gcatggctgagtctaaagttcaaag R gcagcctcagggttctcaa PB+ cccagaaccaaggaagatggt

82

D9 STR D9S2157

Chr 9 q342 NC_00000912

F ggctttgctgggtactgctt R ggaccacagcacatcagtcact PT cagggcacatgaat

60

HBB1 Gene HBB

Chr 11 p155 NC_00001110

F gctgagggtttgaagtccaactc R ggtctaagtgatgacagccgtacct PT agccagtgccagaagagccaagga

76

ND14 STR D14S1434

Chr 14 q3213C NC_0000149

F tccaccactgggttctatagttc R ggctgggaagtcccacaatc PB+ tcagactgaatcacaccatcag

109

22C3 Gene PMM1

Chr 22 q132 NC_00002210

F cccctaagaggtctgttgtgttg R aggtctggtggcttctccaat PB caaatcacctgaggtcaaggccagaaca

78

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

6

a F Forward primer R Reverse primer PB Blackhole quencher probe (FAM labeled) PB+ Blackhole Plus quencher probe (FAM labeled) PT TaqMan MGB probe (FAM labeled)

Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays

Table 2 NIST-Optimized Human Mitochondrial DNA Assays

Assay Location Primers and Probe a Amplicon Length bp Ref

mtND1 34853504 F ccctaaaacccgccacatct

69 22 35333553 R gagcgatggtgagagctaaggt 35063522 PT ccatcaccctctacatc

mtBatz 84468473 F aatattaaacacaaactaccacctacct

79 23 85038524 R tggttctcagggtttgttataa 84758491 PT cctcaccaaagcccata

mtKav 1318813308 F ggcatcaaccaaccacaccta

105 24 1336913392 R attgttaaggttgtggatgatgga 1331013325 PT cattcctgcacatctg

a F Forward primer R Reverse primer PT TaqMan MGB probe (FAM labeled)

2PR4

POTP

NEI

F

R4Q

5

D5

ND6

D9

HBB1

ND1

4

22C3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

7

Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays

Droplet Digital Polymerase Chain Reaction Measurements

Nuclear DNA (nDNA) All nDNA ddPCR measurements used in this study followed the same measurement protocol

bull Droplets were generated on the Auto Droplet Generator using the Bio-Rad ldquoddPCR ldquoSupermix for Probes (No dUTP)rdquo (Bio-Rad cat 186-3024 Control 64079080) and ldquoDroplet Generation Oil for Probesrdquo (Bio-Rad cat 186-4110 Control 64046051) Note desoxyuridine 5-triphosphatase (dUTP) is an enzyme sometimes included in PCR reaction mixtures to minimize PCR carryover contamination

bull Table 3 details the composition of the ddPCR mastermix setup for a given assay and sample where ldquomastermix refers to reaction mixture of manufacturerrsquos proprietary PCR reagent ldquoSupermixrdquo forward and reverse PCR primers for a given assay probe for that assay the sample DNA and sometimes additional water to produce the desired value dilution factor F This setup was used for all assays and samples

bull Non-Template Controls (NTCs) for each assay were included in every analysis bull Droplets were thermal cycled on a ProFlex thermal cycler (Applied Biosystems) for

95 degC for 10 min followed by 60 cycles of 94 degC for 05 min and 61 degC for 1 min then 98 degC for 10 min before a 4 degC hold until the plate was removed

bull Droplets were read on the QX200 Droplet Reader and analysis was performed using the QuantaSoft Analysis Software version 1740917

mtND1 mtBatz mtKav

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

8

bull The numbers of positive and negative droplets at the end of 60 cycles were determined and were exported into a spreadsheet for further analysis Assay-specific intensity thresholds were determined by visual inspection for each assay for each measurement session

Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays

Concentration nDNA Mastermix Microliter

per Reaction 2X ldquoSupermix for Probes (no dUTPs)rdquo 125

5 μmolL Forward Primer 1875 5 μmolL Reverse Primer 1875 5 μmolL Probe (FAM) 125

PCR Grade Water 50 14 Diluted Sample 25 Total Volume 250

Mitochondrial DNA (mtDNA) All the mtDNA measurements used in this study followed the nDNA protocol described above but with different mastermix setups The duplex mtDNA measurements used the setup described in Table 4 The monoplex mtDNA measurements used the setup described in Table 5

Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 125

10 μmolL Forward Primer 09375 10 μmolL Reverse Primer 09375 10 μmolL Probe (VIC) 0625

PCR Grade Water 95 nDNA Mastermix from Table 3 05 Total Volume 250

Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 120

10 μmolL Forward Primer 09 10 μmolL Reverse Primer 09 10 μmolL Probe (FAM) 06

PCR Grade Water 72 Diluted Sample (See Table 7) 24

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

9

Total Volume 240 Sample Preparation

The buffy coat white blood cells used to prepare SRM 2372a materials were purchased from Interstate Blood Bank These materials were from four single-source anonymous individuals supplied to us labeled as Female1 Female2 Male1 and Male2 In our laboratories each donor buffy coat bag was aliquoted (5 mL per aliquot) into sterile 50 mL conical tubes and stored at 4 degC prior to extraction The modified salt-out manual extraction protocol described in Appendix A was performed for each aliquot with rehydration of the DNA in TE-4 buffer From 150 mL to 250 mL of solubilized DNA for each component was accumulated through the extraction of the 5 mL aliquots The DNA was stored in perfluoroalkoxy polymer (PFA) containers that had been bleach sterilized thoroughly rinsed with deionized water rinsed with ethanol and air-dried The minimum volume for production of 2000 units was 120 mL Preliminary estimates of the mass [DNA] of the re-solubilized components were obtained using a Nanodrop microvolume spectrophotometer Dilutions of the component materials were made until the absorbance at 260 nm was approximately 1 corresponding to a nominal mass [DNA] of 50 ngμL of dsDNA7 After addition of the TE-4 buffer the containers were placed on an orbital shaker and gently rotated for several hours prior to refrigerated storage at 4 ordmC The four single-source components were initially screened with the nDNA ddPCR measurement process using four nDNA assays 2PR4 NR4Q ND6 and 22C3 Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate The gravimetric 1rarr4 dilution was prepared by weighing an empty PFA vial and recording the weight then pipetting 30 microL TE-4 into the vial and recording the weight then pipetting 10 microL DNA and recording the final weight The dilution factor was calculated in a spreadsheet These screening measurements verified that the individual component materials were fit-for-purpose The four single-source components were evaluated with the nDNA ddPCR measurement process using the ten nDNA assays Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate Quantitative information was derived from the average of the ten nDNA assays to prepare the component C mixture Thirty milliliters of sample Male2 (estimated at 452 ngμL mass [DNA]) were combined with 80 mL of sample Female2 (estimated at 507 ngμL mass [DNA]) and 10 mL of TE-4 buffer to create a 1 male nDNA to 3 female nDNA producing a 14 malehuman mixture After addition of the TE-4 buffer the component C container was gently rotated on the orbital shaker for several hours prior to refrigerated storage at 4 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

10

Absorbance Measurements

In February 2017 absorbance spectra of the recently prepared fully diluted nominally dsDNA and ssDNA solutions of the three components were acquired with a BioCary 100 spectrophotometer using the measurement protocol described in Appendix B In December 2017 additional spectra were acquired for the remaining bulk solution of components A and B and for SRM 2372 components A and B These solutions had been stored at 4 degC in well-sealed PFA containers There was no remaining bulk component C solution SRM Unit Production

In the morning of the day a component was transferred into the sterile component tubes (Sarstedt Biosphere SC micro tube-05 mL Newton NC) used to package the SRM the container holding the diluted component was removed from the refrigerator and placed on a slow orbital shaker for 2 h The tubes were opened and placed under the laminar flow hood in rows of 5 within 80-hole tube racks for a total of 30 tubes per rack Tubes were placed in every second row of the racks to facilitate filling with an Eppendorf Repeater Xstream pipette (Eppendorf North America Inc Hauppauge NY) fitted with a 1-mL positive displacement tip set to dispense 55 microL per tube The component container was removed from the shaker a magnetic stir-bar was added the container was placed on a magnetic stir plate in the laminar flow hood and the material was stirred gently Filling proceeded until 2200 vials were filled with pipet tip replacement after filling every 200 tubes The filling process consisted of two individuals manually filling each vial within a biosafety cabinet (one physically pipetting each sample the other verifying that each tube was filled) and four individuals tightly closing the lids to each tube on a sterilized and covered bench When all tubes in a rack were filled the rack was moved out of the hood the tubes checked for proper filling their lids closed and the tubes transferred to 100-unit pre-labeled storage boxes As the 100-unit boxes were filled they were transferred to the labeling room where SRM labels were applied by hand and colored inserts were added to the lids red for component A white for component B and blue for component C After all tubes for a given component were labeled the 22 boxes were placed in a locked refrigerator Filling of each of the components took 60 min to 90 min All units were transferred to the refrigerator within 4 hours of beginning the filling process All units of the three materials were allowed to equilibrate for a minimum of two weeks prior to beginning homogeneity analysis Component Homogeneity

The ddPCR measurement protocol described above was used to assess the nDNA homogeneity of the three components One tube from the approximate center of each of the 22 storage boxes was assayed using the NEIF and ND6 assays (see Table 1) A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Three technical replicates were obtained for every tube A total

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

11

of six plates were used two plates per component one with the NEIF assay and one with the ND6 assay Figure 4 details the layout of the samples

Figure 4 Plate Layout Design for Homogeneity Measurements

Component Stability

Stability of the materials was evaluated at 4 degC room temperature (RT asymp 22 degC) and 37 degC Tubes from the homogeneity study were divided among the three temperatures for the stability study Additionally two never-opened tubes (one from box 8 and one from box 12 of each component) were held at each temperature Two tubes at each temperature for each component were opened sampled and resealed 4 times over the course of 8 weeks Measurements were made at week 1 week 3 week 5 and week 8 The measurement on week 8 was from an unopened tube (from box 8) for each of the components at all temperatures The data from the homogeneity study was used as timepoint 0 in the stability study anchored to 4 degC by the storage recommendations A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used for both the NEIF and ND6 assays All temperatures components and both assays fit onto one plate for timepoints 1 to 5 Twelve NTCs placed in rows G and H were included for each assay Figure 5 details the sample layout

Figure 5 Plate Layout Design for Stability Measurements

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

ND6 ND6 ND6

NEIFNEIFNEIFComponent A

Component A

Component B

Component B

Component C

Component C

1 2 3 4 5 6 7 8 9 10 11 12A A-7 A-7 B-7 B-7 C-7 C-7 A-7 A-7 B-7 B-7 C-7 C-7B A-10 A-10 B-10 B-10 C-10 C-10 A-10 A-10 B-10 B-10 C-10 C-10C A-9 A-9 B-9 B-9 C-9 C-9 A-9 A-9 B-9 B-9 C-9 C-9D A-12 A-12 B-12 B-12 C-12 C-12 A-12 A-12 B-12 B-12 C-12 C-12E A-8 A-8 B-8 B-8 C-8 C-8 A-8 A-8 B-8 B-8 C-8 C-8F A-11 A-11 B-11 B-11 C-11 C-11 A-11 A-11 B-11 B-11 C-11 C-11G NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTCH NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

NEIF

4C

RT

37C

ND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

12

Component Certification Measurements

After establishing homogeneity and stability measurements intended to be used to certify nDNA values and inform mtDNA values were started using randomly selected SRM tubes from the boxes listed in Table 7 The ten nDNA assays described in Table 1 and two of the three mtDNA assays described in Table 2 were run simultaneously on the same sample plate A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used in each of the six measurement campaigns

Table 6 Tubes used for Certification Measurements

Component Plate A B C

dd20170711 Box 1 Box 1 Box 1 dd20170712a Box 10 tube a Box 10 tube a Box 10 tube a dd20170712b Box 10 tube b Box 10 tube b Box 10 tube b dd20170713 Box 20 Box 20 Box 20 dd20170714 Box 5 Box 5 Box 5 dd20170801 Box 5 Box 5 Box 5

Every sample was analyzed in duplicate with every assay in all six plates Component A was always located in rows B and C component B in rows D and E component C in rows F and G and NTCs in rows A and H Figure 6 displays the basic layout of the ten nDNA and two mtDNA assays In addition to the plate column assignment scheme shown the three assay groupings were also ordered as Group 2 Group 3 Group 1 and Group 3 Group 1 Group 2 to address potential positional bias

Group 1 Group 2 Group 3 2PR4 POTP NEIF NR4Q D5 ND6 D9 HBB1 ND14 22C3 mtND1 mtBatz 1 2 3 4 5 6 7 8 9 10 11 12 A NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Comp A B Comp A C Comp B D Comp B E Comp C F Comp C G

H NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Figure 6 Basic Plate Layout Design for Certification Measurements

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

13

Mitochondrial DNA Measurements

A ldquoduplexed reactionrdquo was used to determine mtDNAnDNA where 1 μL of one of the nDNA mastermix was added to 49 μL of the mtDNA mastermix resulting in a 1rarr50 volumetric dilution of the nDNA The nDNA was targeted with a FAM-labeled probe while the mtDNA was targeted with a VIC labeled probe The primer and probe concentrations were increased from 5 micromol to 10 micromol in the duplexed reactions Additional testing of the mtDNA assays were performed in monoplex with FAM-labeled probes to determine the additional sources of bias between sampling assay setup and dilution factors Table 7 outlines the different methods examined to determine mtDNAnDNA for the three components

Table 7 Run Parameters for Mitochondrial DNA Droplet

Generation nDNA mtDNA

Plate Component Assay Dilution Reps Assay Dilution Reps dd20170824 A manualDG HBB1 1rarr4 3 mtND1 1rarr200 9 dd20170825 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170829 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170831 AB autoDG HBB1 1rarr3 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr6 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr12 2 mtND1 1rarr10rarr100 3 dd20170907 ABC manualDG HBB1 1rarr3 2 mtBatz 1rarr10rarr100 3 dd20170913 AB manualDG NEIF 1rarr4 3 mtND1 mtBatz mtKav 1rarr10rarr100 3 dd20170913 C manualDG NEIF 1rarr4 3 mtBatz mtKav 1rarr10rarr100 3 Performance in Commercial qPCR Chemistries

Table 8 lists the 11 qPCR chemistries that were commercially available in late 2017 All three of the SRM 2372a components were evaluated with these chemistries using component A of SRM 2372 as the reference standard For each component of SRM 2372a four tubes were pulled from box 22 and combined into one tube A 1rarr5 dilution of all samples was prepared followed by three serial 1rarr2 dilutions A five-point serial dilution of each SRM 2372a component was amplified in triplicate for each chemistry This consisted of the neat material the 1rarr5 dilution then the three serial 1rarr2 dilutions A six-point serial dilution of SRM 2372 Component A was also created for the standard curve for the commercial qPCR chemistries This consisted of the neat material followed by a 1rarr5 dilution then followed by four 1rarr2 dilutions serially The six-point serial dilution series of SRM 2372 Component A was also amplified in triplicate These commercial chemistries were amplified in individual plates on an Applied Biosystems 7500 HID Real-Time PCR Instrument (Foster City CA) following the manufacturersrsquo recommended protocols for reaction mix setup and cycling conditions Figure 7 details the plate layout for the qPCR plates Data was analyzed with the HID Real-Time PCR Analysis Software v12 and further analyzed in a spreadsheet

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

14

Table 8 Commercial qPCR Chemistries Tested with SRM 2372a

Manufacturer Commercial Kit Target Codea

Innogenomics

InnoQuant Long IQh Short IQd

InnoQuantHY Long IQHYh Short IQHYd Y IQHYy

Thermo Fisher Quantifiler Duo Human QFDh Male QFDy

Quantifiler HP Large QFHPh Small QFHPd

Quantifiler Human Human QFHh

Quantifiler Trio Large QFTh Small QFTd Y QFTy

Qiagen

Quantiplex Autosomal QPh

Quantiplex Hyres Autosomal QPHh Y QPHy

Quantiplex Pro Human QPPh Degradation QPPd Y QPPy

Promega

Plexor Autosomal Ph Y Py

PowerQuant Autosomal PQh Degradation PQd Y PQy

a Assays are coded with a short acronym derived from the kit name followed by the following codes for the type of target ldquohrdquo autosomal non-degradation described as ldquoAutosomalrdquo ldquoHumanrdquo or ldquoLongrdquo ldquodrdquo autosomal degradation described as ldquoDegradationrdquo or ldquoShortrdquo ldquoyrdquo Y-chromosome described as ldquoMalerdquo or ldquoYrdquo

1 2 3 4 5 6 7 8 9 10 11 12

A

B 2372-A_57 2372-A_57 2372-A_57 2372a-neat_A

2372a-neat_A

2372a-neat_A

2372a-neat_B

2372a-neat_B

2372a-neat_B

2372a-neat_C

2372a-neat_C

2372a-neat_C

C 2372-A_11

4 2372-A_11

4 2372-A_11

4 2372a-15_A

2372a-15_A

2372a-15_A

2372a-15_B

2372a-15_B

2372a-15_B

2372a-15_C

2372a-15_C

2372a-15_C

D 2372-A_5

7 2372-A_5

7 2372-A_5

7 2372a-110_A

2372a-110_A

2372a-110_A

2372a-110_B

2372a-110_B

2372a-110_B

2372a-110_C

2372a-110_C

2372a-110_C

E 2372-A_2

85 2372-A_2

85 2372-A_2

85 2372a-120_A

2372a-120_A

2372a-120_A

2372a-120_B

2372a-120_B

2372a-120_B

2372a-120_C

2372a-120_C

2372a-120_C

F 2372-A_1

425 2372-A_1

425 2372-A_1

425 2372a-140_A

2372a-140_A

2372a-140_A

2372a-140_B

2372a-140_B

2372a-140_B

2372a-140_C

2372a-140_C

2372a-140_C

G 2372-A_0

7125 2372-A_0

7125 2372-A_0

7125

H NTC NTC NTC

Figure 7 Plate Layout Design for Performance Assessments

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

15

Measurement Results After preparing the SRM 2372a bulk solutions their approximate 50 ngmicroL mass [DNA] were confirmed spectrophotometrically The bulk solutions were then packaged into individual tubes The materials contained in the tubes were characterized for homogeneity stability nDNA copy number [DNA] and mtDNAnDNA ratio The following sections detail the results of these measurements Absorbance Spectrophotometry

By widely accepted convention an aqueous DNA solution having an absorbance of 10 at 260 nm corresponds to a mass [DNA] of 50 ngmicroL for dsDNA and (37 to 40) ngmicroL for ssDNA7 Absorbance at (230 270 280 and 330) nm are traditionally used in the assessment of DNA quality725 Due to minor baseline shifts during acquisition of some spectra it was necessary to bias-adjusted all spectra to have absorbance of 00 at 330 nm Table 9 lists the resulting bias-adjusted values for selected wavelengths and the conventional mass [DNA] conversion of the absorbance at 260 nm The close agreement between the mass concentration values for the nominally dsDNA and the converted-to-ssDNA spectra confirm that the bulk solutions were correctly prepared to be asymp 50 ngμL dsDNA The slight increase in absorbance over the 10 months between acquiring the spectra suggests that like SRM 2372 the tertiary structure of the dsDNA in the SRM 2372a materials may be changing slowly with time Figure 8 displays the spectra

Table 9 Absorbance at Selected Wavelengths

Wavelength nm Form Date Component 230 260 270 280 ngμL dsDNA 2142017 A 0394 0931 0758 0488 466

B 0480 1112 0905 0583 556 C 0451 1006 0815 0524 503

ssDNA 2142017 A 0390 0614 0540 0316 454 B 0457 0723 0634 0372 535 C 0411 0641 0563 0330 474

dsDNA 12142017 A 0402 0956 0775 0503 478 B 0484 1128 0914 0589 564

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

16

Figure 8 Absorbance Spectra of SRM 2372a Components A B and C

dsDNA is denatured to ssDNA by diluting with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to DNA mass concentration

Droplet Volume

Droplet volumes for the two lots of ldquoSupermixrdquo used in the studies described in this report were determined using NISTrsquos Special Test Method 11050S-D as described in Dagata et al19 The lot used for most of the ddPCR analyses produced (07349 plusmn 00085) nL droplets where the ldquoplusmnrdquo value is the combined standard uncertainty The lot used for analyses performed after October 11 2017 produced (07376 plusmn 00085) nL droplets NIST measurements have demonstrated that droplet volumes are not influenced by the presence or absence of DNA in the sample solution Measurements repeated over time on an earlier lot of the Supermix product used in these studies suggest that droplet volumes remain constant over the shelf-life of the lot However droplet volumes for different lots of this product as well as for different Supermix products have been observed to differ by several percent We intend to document these measurements in a future report Other researchers have demonstrated that non-constant droplet volume distributions may contribute to ddPCR measurement imprecision and bias2627 However these effects become significant only with distributions wider than observed using Special Test Method 11050S-D19 and at copydroplet values larger than the λ le 05 copydroplet typically used in our measurements Certification Measurements

Homogeneity Results Table 10 lists the λacute nDNA copy number values adjusted for both droplet volume and sample dilution from the homogeneity measurements The data was analyzed using the three-level Gaussian hierarchical ANOVA model described in Appendix C Table 11 lists the values of the within- and between-tube variance components expressed as percent relative standard deviations At the 95 confidence level all components are homogeneous

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372a-A2141721417-NaOH121417

220 260 300 340

Wavelength nm

2372a-B2141721417-NaOH121517

220 260 300 340

Wavelength nm

2372a-C21417

21417-NaOH

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

17

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 NEIF A-1-01 1632 1529 1511 B-1-01 1790 1909 1885 C-1-01 1493 1385 1510 ND6 1616 1640 1582 NAe 1870 1860 1565 1562 1466

NEIF A-1-02 1531 1525 1549 B-1-02 1712 1719 1704 C-1-02 1516 1513 1577 ND6 1615 1589 1601 1775 1718 1756 1490 1547 1570

NEIF A-1-03 1524 1540 1533 B-1-03 1693 1755 1749 C-1-03 1451 1528 1506 ND6 1582 1152 1555 1824 1483 1896 1485 1505 1471

NEIF A-1-04 1454 1458 1427 B-1-04 1754 1745 1739 C-1-04 1440 1486 1520 ND6 1627 1605 1607 1670 1632 1814 1529 1508 1506

NEIF A-1-05 1465 1540 1504 B-1-05 1757 1687 1735 C-1-05 1523 1472 1483 ND6 1607 1633 1586 1683 1759 1703 1513 1520 1491

NEIF A-1-06 1512 1491 1491 B-1-06 1697 1709 1714 C-1-06 1437 1493 1522 ND6 1511 1483 1524 1693 1723 1722 1492 1481 1461

NEIF A-1-07 1418 1480 1455 B-1-07 1722 1625 1703 C-1-07 1485 1477 1463 ND6 1479 1557 1656 1690 1684 1688 1475 1547 1485

NEIF A-1-08 1496 1560 1501 B-1-08 1884 1802 1772 C-1-08 1433 1405 1478 ND6 1514 1615 1540 1864 1887 1821 1497 1501 1495

NEIF A-1-09 1588 1581 1637 B-1-09 1713 1734 1727 C-1-09 1463 1499 1505 ND6 1474 1438 1454 1709 1756 1758 1503 1491 1516

NEIF A-1-10 1574 1513 1570 B-1-10 1759 1809 1738 C-1-10 1534 1458 1454 ND6 1594 1713 1669 1764 1816 1787 1448 1465 1474

NEIF A-1-11 1557 1508 1633 B-1-11 1749 1782 1740 C-1-11 1533 1514 1555 ND6 1571 1611 1631 1741 1792 1784 1488 1470 1483

NEIF A-1-12 1500 1476 1515 B-1-12 1748 1712 1718 C-1-12 1450 1525 1497 ND6 1626 1586 1608 1797 1614 1790 1496 1483 1519

NEIF A-1-13 1672 1634 1804 B-1-13 1632 1668 1701 C-1-13 1537 1527 1515 ND6 1490 1516 1497 1626 1566 1681 1499 1479 1475

NEIF A-1-14 1551 1568 1538 B-1-14 1752 1535 1673 C-1-14 1515 1490 1494 ND6 1617 1614 1639 1724 1741 1685 1469 1480 1466

NEIF A-1-15b 1793 1722 1733 B-1-15 1871 1835 1887 C-1-15 1511 1568 1594 ND6 1818 1753 1762 1770 1752 1798 1515 1543 1565

NEIF A-1-16 1573 1520 1617 B-1-16 1688 1725 1743 C-1-16 1449 1452 1471 ND6 1634 1627 1575 1744 1697 1710 1543 1448 1491

NEIF A-1-17 1516 1516 1628 B-1-17 1789 1769 1825 C-1-17 1388 1421 1394 ND6 1613 1567 1567 1716 1762 1638 1452 NAe 1435

NEIF A-1-18 1612 1535 1672 B-1-18 1826 1780 1849 C-1-18 1475 1455 1445 ND6 1529 1495 1528 1805 1801 1931 1458 1531 1459

NEIF A-1-19 1598 1514 1596 B-1-19 1751 1759 1880 C-1-19 1468 1525 1446 ND6 1571 1535 1511 1827 1766 1786 1435 1478 1421

NEIF A-1-20 1575 1568 1586 B-1-20 1738 1678 1768 C-1-20 1549 1531 1499 ND6 1652 1545 1627 1752 1747 1768 1499 1465 1468

NEIF A-1-21 1658 1611 1578 B-1-21 1742 1836 1713 C-1-21 1530 1440 1484 ND6 1647 1607 1612 1746 1744 1764 1499 1454 NAe

NEIF A-1-22 1547 1528 1532 B-1-22 1761 1743 1660 C-1-22 1442 1487 1436 ND6 1537 1513 1536 1666 1692 1819 1589 1591 1556

NEIF A-1-15c 1456 1448 1523 ND6 1525 1402 1436

NEIF A-2-15d 1679 1599 1674 ND6 1581 1550 1584

a Tube coded as Component-Tube-Box b These results were identified as a probable error in the preparation or volumetric delivery of the sample

solution into the microfluidic device and were not used in the homogeneity evaluations c Repeat assessment of the solution in sample A-1-15 d Results for a second tube from box 15 analyzed at the same time as the repeat assessment of A-1-15 e Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

18

Table 11 Within- and Between-Tube Relative Standard Deviations

Within-Tube Component NEIF ND6 Between-Tube

A 28 40 33 B 27 38 25 C 23 18 19

Stability Results Table 12 lists the λacute volume- and dilution-adjusted nDNA copy number values from the stability measurements Figure 9 summarizes these results The results for the tubes held at 4 degC room temperature (RT asymp 22 degC) and 37 degC are very similar for all three components The absence of between-temperature differences identifies the between-week differences as attributable to variation in the measurement processes rather than sample instability These results indicate that the SRM 2372a genomic copy number is thermally stable from 4 degC to 37 degC over an extended period

Figure 9 ddPCR Stability Measurements as Function of Time

Each dot-and-bar symbol represents the mean λacute and the approximate 95 expanded uncertainty for all results for a given component at a given storage temperature The thick black horizontal lines represent an approximate 95 confidence interval on the λacute values estimated from the homogeneity data

13

14

15

16

17

18

19

0 1 2 3 4 5 6 7 8

λacute H

aplo

id n

DNA

Copi

es μL

Weeks

A

0 1 2 3 4 5 6 7 8

Weeks

B

4 degC 22 degC 37 degC

0 1 2 3 4 5 6 7 8

Weeks

C

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

19

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Week Temp Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3

1 4 degC NEIF A-10 1568 1441 B-10 1666 1653 C-10 1424 1394 ND6 A-10 1474 1494 B-10 1655 1705 C-10 1461 1394

NEIF A-07 1500 1463 B-07 1718 1640 C-07 1446 1452 ND6 A-07 1479 1479 B-07 1708 1686 C-07 1522 1482

RT (22 degC) NEIF A-12 1674 1608 B-12 1805 1777 C-12 1506 1495 ND6 A-12 1522 1507 B-12 1759 1788 C-12 1478 1461 NEIF A-09 1635 1539 B-09 1791 1793 C-09 1411 NAb ND6 A-09 1543 1519 B-09 1661 1682 C-09 1482 1496

37 degC NEIF A-11 1614 1637 B-11 1748 1744 C-11 1515 1464 ND6 A-11 1494 1502 B-11 1740 1697 C-11 1495 1535 NEIF A-08 1592 1614 B-08 1691 1725 C-08 1528 1500 ND6 A-08 1581 1595 B-08 1727 1699 C-08 1458 1483 3 4 degC NEIF A-10 1520 1706 B-10 NAb 1733 C-10 1427 1473 ND6 A-10 1543 1624 B-10 1670 1755 C-10 1443 1364 NEIF A-07 1537 1570 B-07 1549 1565 C-07 1424 1512 ND6 A-07 1517 1479 B-07 1705 1723 C-07 1483 1439

RT (22 degC) NEIF A-12 1521 1607 B-12 1817 1842 C-12 1566 1471 ND6 A-12 1644 1597 B-12 1845 1786 C-12 1504 1511 NEIF A-09 1432 1626 B-09 1825 1856 C-09 1392 1387 ND6 A-09 1539 1528 B-09 1734 1737 C-09 1535 1465

37 degC NEIF A-11 1651 1685 B-11 1759 1779 C-11 1437 1440 ND6 A-11 1680 1594 B-11 1806 1806 C-11 1471 1509 NEIF A-08 1580 1613 B-08 1740 1791 C-08 1582 1590 ND6 A-08 1592 1524 B-08 1769 1694 C-08 1468 1520 5 4 degC NEIF A-10 1486 1440 B-10 1626 1551 C-10 1381 1330 ND6 A-10 1555 1577 B-10 1673 1625 C-10 1547 1395 NEIF A-07 1387 1429 B-07 1723 1787 C-07 1330 1314 ND6 A-07 1538 1522 B-07 1862 1909 C-07 1452 1458

RT (22 degC) NEIF A-12 1601 1631 B-12 1718 1752 C-12 1546 1541 ND6 A-12 1699 1677 B-12 1772 1721 C-12 1490 1471 NEIF A-09 1381 1484 B-09 1675 1694 C-09 1371 1376 ND6 A-09 1625 1615 B-09 1904 1815 C-09 1468 1472

37 degC NEIF A-11 1456 1517 B-11 1720 1704 C-11 1505 1493 ND6 A-11 1650 1658 B-11 1148 1152 C-11 1527 1600 NEIF A-08 1590 1666 B-08 1745 1723 C-08 1469 1540 ND6 A-08 1235 1244 B-08 1809 1749 C-08 1508 1529 8 4 degC NEIF A-08a 1667 1632 1589 B-08a 1880 1884 1940 C-08a 1525 1475 1542 ND6 A-08a 1599 1593 1590 B-08a 1859 1897 1972 C-08a 1550 1502 1520

RT (22 degC) NEIF A-08b 1693 1666 1634 B-08b 1769 1884 1771 C-08b 1595 1557 1545 ND6 A-08b 1644 1620 1647 B-08b 1862 1738 1811 C-08b 1611 1585 1535

37 degC NEIF A-08c 1620 1576 1608 B-08c 1934 1952 1956 C-08c 1595 1600 1569 ND6 A-08c 1576 1553 1532 B-08c 1889 1844 1889 C-08c 1569 1514 1581

a Tube coded as Component-Tube-Box b Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

20

Ten-Assay Certification Results The original intention was to certify the copy number [DNA] of the three components with results from five independently prepared sets of samples analyzed over four days Each sample of the three components was from a new tube from different boxes and was evaluated with 10 nDNA assays The assay positions were alternated across these five ldquo10-assay certificationrdquo plates The λ results for these five plates are systematically somewhat larger than those for the homogeneity and stability data On review we realized that a different workspace and its suite of pipettes had been used to prepare the 1rarr10 dilution of the 1rarr4 diluted components into the reaction mastermix The ldquohighrdquo workstation was used for stability and homogeneity measurements while the ldquolowrdquo workstation was used for certification measurements Gravimetric testing of the pipettes established that the volumes delivered by these alternate pipettes was sufficiently biased to account for the observed λ differences A sixth ldquocertificationrdquo plate was prepared using the original (and now re-verified) pipettes The λ results or this plate agreed well with the homogeneity and stability data Table 13 lists the λacute volume- and dilution-adjusted nDNA copy number values for all six of the ten-assay datasets Figure 10 displays the extent of agreement among the ten assays While there are small-but-consistent differences between the assays there is only a 4 difference between the assays providing smallest and largest λ

Figure 10 Comparison of Assay Performance

The blue dot-and-bars represent the mean and its combined standard uncertainty for the A B and C components for each assay relative to the over-all grand mean for each component The black dot-and-bars represent combination of three component results for each assay The solid horizontal line represents the median difference of the assay means from the grand mean the dotted horizontal lines bound an approximate 95 confidence region about the median

2PR4POTPNEIFR4Q5 D5ND6 D9 HBB1ND14 22C3-10

-8

-6

-4

-2

0

2

4

6

8

10

Rela

tive

Diffe

renc

e fr

om M

ean

Assays

Component MeansAssay MeansMedianU95

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

21

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

7112017 2PR4 1667 1698 1869 1960 1630 1706 POTP 1655 1685 1850 1781 1596 1576 NEIF 1647 1665 1709 1707 1574 1538 R4Q5 1594 1658 1856 1787 1590 1555 D5 1655 1683 1848 1829 1610 1607 ND6 1629 1647 1828 1826 1589 1548 D9 1655 1652 1861 1858 1569 1572 HBB1 1644 1611 1825 1869 1565 1681 ND14 1645 1707 1871 1773 1564 1492 22C3 1545 1603 1776 1902 1548 1655

7122017 2PR4 1668 1739 2174 2231 1499 1497 POTP 1587 1702 2202 2292 1576 1611 NEIF 1678 1614 2172 2216 1545 1560 R4Q5 1628 1762 NA 2078 1584 1546 D5 1738 1649 2230 2192 1577 1529 ND6 1667 1678 2094 2034 1599 1478 D9 1695 1653 1949 2110 1587 1504 HBB1 1757 1750 1971 2123 1520 1551 ND14 1771 1764 2123 2066 1479 1438 22C3 1762 1759 2107 2046 1512 1494

7122017 2PR4 1638 1667 1810 1806 1452 1424 POTP 1676 1638 1865 1801 1347 1397 NEIF 1616 1624 1806 1786 1395 1350 R4Q5 1559 1587 1878 1768 1442 1459 D5 1608 1652 1824 1790 1448 1602 ND6 1611 1589 1768 1670 1463 1435 D9 1517 1607 1811 1749 1402 1433 HBB1 1557 1582 1763 1741 1474 1481 ND14 1604 1625 1865 1781 1466 1482 22C3 1675 1616 1815 1866 1439 1568

7132017 2PR4 1888 1939 1784 1821 1470 1444 POTP 1868 1815 1735 1813 1411 1427 NEIF 1953 1902 1778 1813 1512 1439 R4Q5 1779 1871 1835 1749 1452 1468 D5 1924 2124 1794 1848 1567 1524 ND6 1861 1787 1700 1711 1480 1424 D9 1824 1844 1799 1779 1442 1459 HBB1 1809 1799 1765 1769 1430 1467 ND14 1864 1857 1754 1725 1413 1411 22C3 1930 1942 1792 1814 1486 1480

7142017 2PR4 1597 1590 1718 1768 1485 1458 POTP 1533 1404 1753 1698 1422 1459 NEIF 1462 1533 1773 1792 1438 1465 R4Q5 1493 1519 1777 1740 1439 1468 D5 1538 1586 1803 1795 1515 1457 ND6 1492 1516 1718 1693 1409 1412 D9 1528 1538 1802 1743 1469 1474 HBB1 1548 1581 1834 1745 1453 1428 ND14 1513 1514 1731 1714 1465 1425 22C3 1562 NA 1842 NA 1363 1393

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

22

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

101217 2PR4 1431 1411 1791 1770 1357 1364 POTP 1407 1395 1746 1820 1447 1361 NEIF 1487 1398 1803 1803 1448 1404 R4Q5 1450 1439 1767 1739 1409 1411 D5 1394 1422 1745 1773 1389 1429 ND6 1396 1393 1824 1729 1415 1417 D9 1414 1390 1752 1744 1431 1380 HBB1 1401 1382 1813 1754 1381 1336 ND14 1412 1460 1648 1671 1403 1387 22C3 1435 1410 1818 1741 1435 1372

Estimation of Copy Number Concentration It was decided that all three sets of results (homogeneity stability and certification) would be used to derive the certified values with the λ of the first five ldquoten-assayrdquo plates adjusted by a data-driven estimate of the volumetric bias This bias was incorporated into the statistical model providing better estimates of the λ means while capturing all of the variance components (homogeneity stability between-assay and between data sets) as random effects using Bayesian MCMC programmed in OpenBUGS28 Table 14 lists the estimated values and their uncertainties See Appendix C for details

Table 14 Measured Copy Number Concentrations

Genomic Copies per Nanoliter of Solution Component Mean a u(Mean) b U95(Mean) c CV95

d A 151 07 14 93 B 175 03 06 34 C 145 05 10 69

a Mean the experimentally determined most probable estimate of the true value b u(Mean) the standard uncertainty of the Mean c Mean plusmnU95(Mean) the interval which is believed with about 95 confidence to contain the true value d CV95 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the counting unit one29 2) the validity of the Poisson endpoint transformation for digital PCR endpoint assays when applied to samples providing le 05 copies per droplet 3) the determination that the copies are dispersed as dsDNA and 4) calibrated mean droplet volume measurements made at NIST during sample dilution and mastermix preparation Conversion of Copy Number to Mass Concentration As described in Reference 18 metrologically valid conversion from copies per nanoliter to nanograms per microliter requires quantitative estimates for the number of nucleotide bases per reference human genome and the average nucleotide mass Using the values estimated in that report and asserting that each target copy corresponds to one HHGE the required transformation is

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

23

[DNA] ngmicroL

= [DNA] HHGE

nLtimes

3301 ngHHGE

The combined standard uncertainty is

119906119906 [DNA] ng

microL = 1199061199062 [DNA] HHGE

nL + 0531100

times[DNA] HHGE

nL 2

Both the measured and the conversion factor uncertainties are associated with ldquolargerdquo degrees of freedom Therefor the approximate 95 confidence uncertainty is estimated

11988011988095 [DNA] ng

microL = 2 times 119906119906

[DNA] ngmicroL

Table 15 lists the transformation of the Table 14 copy number values These values agree well with the conventional 260 nm absorbance values averaging (4 plusmn 5) larger

Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution

ngmicroL Component Value a u(Measurement)b u(Conversion)c U95(Value)d CV95

e A 498 23 03 47 93 B 578 10 03 21 36 C 479 17 03 33 70

a Estimated true value b Standard uncertainty component from the ddPCR measurements c Standard uncertainty estimated from the relative uncertainty of the conversion factor d Value plusmnU95(Value) is believed with about 95 confidence to contain the true value e 100U95(Value)Value the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the measured estimates of copies per nanoliter and 2) determinations made by us from internationally recognized literature resources of the average number of nucleotide bases per HHGE and the mean molecular weights of the bases18 Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)

While not intended to provide certified values mtDNAnDNA was measured as part of 13 experiments over 5 months Measurements were made using dilution-adjusted λ for the mtND1 mtBatz andor mtKav mitochondrial assays referenced against the dilution-adjusted λ of either the NEIF or HBB1 nDNA assays By chance both DNAs in the component C mixture have a binding site mutation that silences the mtND1 probe Figure 11 graphically summarizes the results by assay and component Table 16 lists the data Table 17 summarizes results from a random effects model that accounts for between-assay variability See Appendix C for details

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

24

Due to the limited number of mtDNA assays used in these measurements these results are metrologically traceable only to the assays used

Figure 11 mtDNAnDNA for Components A B and C

Each dot-and-bar symbol represents a mean mtDNAnDNA and its approximate 95 expanded uncertainty The dotted horizontal lines represent approximate 95 confidence intervals on the ratios

Table 16 mtDNAnDNA Ratios

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

20170711a 1 mtND1 1931 1870 2461 2401 1 mtBatz 1845 1764 2555 2608 3201 3299

20170711b 1 mtND1 1810 1906 2229 2200 1 mtBatz 1785 1837 2108 2192 3347 3287

20170712 1 mtND1 1743 1761 2127 2198 1 mtBatz 1829 1847 2455 2506 3217 3384

20170713 1 mtND1 1716 1800 2211 2199 1 mtBatz 1740 1813 2380 2249 3200 3363

20170714 1 mtND1 1923 1835 2105 1943 1 mtBatz 1935 1851 2136 2119 3294 3013

20170814 1 mtND1 1780 1776 1706 2046 2057 2024 1 mtBatz 1804 1751 1738 1998 2043 1952 2927 2776 2762 1 mtKav 1769 1786 1761 2077 2117 2041 3044 2830 2825

20170824 1 mtND1 1947 1928 1922 2 mtND1 1829 1895 1851 3 mtND1 1674 1617

20170825 1 mtBatz 1667 1813 1708 1598 1928 1916 1915 2832 2840 2788 2725 2 mtBatz 1566 1687 1663 1589 1947 1946 1961 1909 2835 2790 2736 2679 3 mtBatz 1643 1668 1604 1573 1953 1952 1941 1881 2803 2773 2757 2752

20170829 1 mtBatz 1719 1716 1670 1666 1865 1863 1828 1828 2599 2508 2551 2609 2 mtBatz 1581 1573 1601 1612 1944 1964 1947 1895 2617 2483 2626 2602 3 mtBatz 1546 1542 1538 1587 1908 1909 1940 1839 2665 2535 2570 2650

20170831 1 mtND1 1617 1599 1688 1783

ND1

ND1

ND1Ka

v

Kav

Kav

Batz

Batz

Batz

A

B

C

160

210

260

mtD

NAn

DNA

Ratio

Assay

Bind

ing s

ite m

utat

ion

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

25

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

2 mtND1 1622 1579 1823 1631 3 mtND1 1575 1528 1920 1887

20170907 1 mtBatz 1584 1544 1890 1886 2501 2522 20170913 1 mtND1 1757 1724 1768 1990 1921 1982

1 mtBatz 1743 1671 1766 1990 1924 1958 2805 2825 2829 1 mtKav 1692 1662 1767 1903 1831 1919 2709 2738 2727

20171016 1 mtND1 1924 1944 1893 2304 2341 2332 1 mtBatz 1786 1752 1743 2048 2078 2067 2797 2642 2792 1 mtKav 1530 1542 1523 2399 2479 2493 2804 2789 2842 2 mtBatz 1580 1557 1592 2715 2606 2687 2 mtKav 2103 2100 2049 2799 2628 2559 3 mtND1 1749 1769 1777 2065 2166 2164 3 mtBatz 1965 1889 1947 2314 2280 2258 2639 2535 2657 3 mtKav 1813 1864 1864 1867 1987 1971 2953 2757 2802

Table 17 mtDNAnDNA Summary Results

mtDNAnDNA Component Value a u(Value)b U95(Value)c CV95

d A 1737 16 38 17 B 2058 22 44 21 C 2791 23 47 17

a Estimated true value b Standard uncertainty c Value plusmnU95(Value) is believed with about 95 confidence to contain the true value d 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) Performance in Commercial qPCR Chemistries

Traditional qPCR assays only quantify relative to a reference standard Thus qPCR assays do not provide absolute [DNA] for the individual materials but they can provide [DNA] of one component relative to another Figure 12 summarizes the BA and CA ratios for the commercial autosomal (Human) targets from the eleven assays evaluated (see Table 8) The Y-chromosome (Male) assays cannot be summarized in this manner as the (all female) component B does not include a Y chromosome Figure 13 summaries the mass [DNA] results for the (all male) component A and (male-female mixture) component C Since this material was prepared as a nominal 13 malefemale mixture the expected malehuman (total) ratio is 14 sjc As displayed in these figures there is intrinsic variability between the SRM 2372a components and between the commercial qPCR chemistries examined This reflects the difference in the targets and amplification chemistries for each of the commercial kits Many of the commercial chemistries on the market to date contain multi-copy targets for both autosomal and Y-chromosomal quantification to increase the sensitivity of the assays

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

26

Figure 12 qPCR Autosomal Quantitation Performance Summary

Open circles denote the BA CA ratios the 17 commercial qPCR assays evaluated each labeled with the code listed in Table 8 The solid black circle represents ratios as estimated by ddPCR The red lines bound the approximate 95 confidence intervals on the ddPCR-estimated ratios

Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary

Open squares denote the ngμL values for SRM 2372a Components A (male) and C (mixture of male and female) for the seven commercial Y-chromosome qPCR assays evaluated each labeled with the code listed in Table 8 The dotted red lines represent malehuman ratios of 13 14 and 15

Appendix D provides a graphical analysis of all the data generated in the performance assessments for each of the targets amplified in the eleven commercial qPCR chemistries and the [DNA] values for all of the assays

dd

IQ_h

IQ_d

IQHY_h

IQHY_d

QFD_h

QFH_h QFHP_h

QFHP_dQFT_h

QFT_d

QPH_h

QPP_hQPP_d

P_h

PQ_h

PQ_d

06

08

10

12

14

07 09 11 13

CA

BA

IQHY_y

QFD_y

QFT_y

QPH_y

QPP_y

P_y

PQ_y

1413

15

10

12

14

16

18

30 40 50 60 70

[DNA

] C n

gμL

[DNA]A nguL

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

27

Qualification of ddPCR as Fit-For-Purpose We have documented the requirements for establishing metrological traceability of ddPCR human genomic assay results expressed as copies per nanoliter of sample in a series of three peer-reviewed papers151617 Through the experimental evidence they are based upon these papers demonstrate that ddPCR can accurately enumerate the number of independently migrating genomic copies in an aqueous extract Our most recent paper establishes the metrological traceability of these ddPCR results when transformed into the nanogram per microliter units used by the forensic and clinical communities18 However our previous studies have not addressed a fundamental requirement for correctly interpreting digital PCR measurements In broad simplification end-point assay systems count the number of independently partitioning entities regardless of the number of copies the entities contain With assays that target just one site per HHGE entities that migrate as ssDNA will yield counts that are twice the value had they migrated as dsDNA30 Should some originally dsDNA be converted over time to ssDNA ddPCR measurements of the number of independently migrating entities would increase although the number of HHGE would be unchanged Measurements of mass [DNA] provided by absorbance spectrophotometry and ddPCR reflect different properties absorbance at 260 nm is sensitive to the spatial configuration of the nucleotides as well as their number while ddPCR counts independently migrating entities The SRM 2372a materials have absorbance spectra shown in Figure 8 consistent with being predominantly dsDNA The SRM 2372 materials now have the absorbance spectra shown in Figure 14 which are consistent with them being entirely ssDNA

Figure 14 Absorbance Spectra of SRM 2372 Components A and B dsDNA is denatured to ssDNA by diluting the source material with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to mass [DNA]

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372-A200720122012-NaOH121417

0

4

6

0

4

6

220 260 300 340

Wavelength nm

2372-B200720122012-NaOH121517

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

28

To ensure the validity of ddPCR-based value assignments it is thus necessary to evaluate the nature of the partitioning entities Should a solution of what is thought to be dsDNA contain an appreciable fraction of ssDNA then the ddPCR results would require adjustment31 To avoid further delay in making SRM 2372a available to the forensic community the following sections detail a series of experiments that have enabled us to conclude with high confidence that the SRM 2372a components have a dsDNA conformation and that for this material ldquoentitiesrdquo can be considered dsDNA ldquocopiesrdquo The remaining stock of the SRM 2372 components also have a dsDNA conformation however the strands in the old materials are much less robustly connected than they are in the new materials Digital Polymerase Chain Reaction Assays

Three nDNA assays were used HBB1 NEIF and ND6 The HBB1 and NEIF assays reproducibly yield very similar results for DNA extracts from many different donors The HBB1 target is at the tip of the short (p) arm of chromosome 11 The NEIF target is in the centromere region of chromosome 2 The ND6 assay was selected for use because it typically yields the lowest results of the ten nDNA assays used to certify the SRM 2372a components (see Figure 10 below) The ND6 target is about halfway down the long (q) arm of chromosome 6 Heat Treatments (Melting of the DNA into Single Strands)

In all experiments the non-heated control and the heat-treated diluted solutions were stored in the same type of PCR reaction tube The starting point for the heat treatment was to determine if there was a difference between snap-cooling on ice and slower cooling at 25 degC Two tubes of the diluted material were placed in a GeneAmp 9700 thermal cycler (Applied Biosystems) and cycled up to 98 degC and held for 15 min After 15 min one tube was removed and placed on ice while the remaining tube was cooled at 1 degCs in the thermal cycler to 25 degC then placed on ice until prepared for ddPCR Additional heat treatments included cycling samples on a GeneAmp 9700 thermal cycler

a) from 98 degC for 5 min then cooling to 25 degC for 2 min During the 25 degC hold a randomly selected tube was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 45 min at 98 degC

b) from 94 degC for 1 min then cooling to 25 degC for 1 min During the 1 min hold at 25 degC a random sample was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 9 min at 94 degC

c) from 90 degC for 1 min then cooling to 25 degC for 1 min for a total of 20 cycles Samples were pulled at 25 degC after 1 min 3 min 5 min 10 min and 20 min and placed on ice

After the above experiments heat treatments were performed using a ProFlex thermal cycler This thermal cycler has six temperature zones that can be le 5 degC apart This enables heating

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

29

tubes at the same time but at different temperatures After treatment tubes were cooled at 5 degCs to 4 degC in the thermal cycler While designed for use with PCR plates a MicroAmp 96-well tray for the VeriFlex block enabled its use with single PCR reaction tubes Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a

As a first step in evaluating whether the DNA in the SRM materials disperse as dsDNA or ssDNA a direct comparison of the five materials was made using the HBB1 and NEIF assays Table 18 reports summary statistics for the ddPCR measurements reported as λ copiesdroplet and the estimated mass [DNA] values from the ddPCR results18 from the conventional dsDNA relationship7 and from the conventional ssDNA relationship after denaturing with 04 molL NaOH89 While the mass [DNA] values differ somewhat between the three methods there is no evidence for the SRM materials segregating exclusively or even mostly as ssDNA The extent of the hypochromic interactions among the nucleotides does not accurately reflect the extent to which the two strands of dsDNA have separated

Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a

λ copies per droplet λ copies per droplet ngmicroL Sample Assay Nrep

a Mean CVb ncmbc Mean CVb ddPCRd dsDNAe ssDNAf

2372-A HBB1 4 03645 19 8 03644 20 566g 524 57 2372-A NEIF 4 03644 23 2372-B HBB1 4 03513 23 8 03487 35 580h 536 61 2372-B NEIF 4 03461 47 2372a-A HBB1 4 02603 27 8 02649 30 474 465 454 2372a-A NEIF 4 02694 24 2372a-B HBB1 4 02908 34 8 02891 37 517 556 534 2372a-B NEIF 4 02873 45 2372a-C HBB1 4 02757 28 8 02711 28 485 502 474 2372a-C NEIF 4 02665 16

a Number of technical replicates for the given assay b CV = 100(standard deviation)mean c Combined number of technical replicates ignoring potential differences between the assays d Estimated from the ddPCR λ measurements listed in this Table18 e Estimated from the conventional relationship between the mass [DNA] of dsDNA and absorbance at

260 nm7 using absorbance spectra acquired shortly after solutions were produced f Estimated from the relationship between the mass [DNA] of ssDNA and absorbance at 260 nm after 11

dilution with 04 molL NaOH89 g Values adjusted to measured (132 plusmn19) volume loss of archived units due to evaporation h Values adjusted to measured (71 plusmn25) volume loss of archived due to evaporation

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

30

Quantitative Evaluation of ssDNA Content

Absorbance spectrophotometry is routinely used to quantitatively assess DNA tertiary structure3233 On storage ldquoat physiological pH the intrastrand repulsion between the negatively charged phosphate groups forces the double helix into more rigid rodlike conformationrdquo34 suggesting a mechanism for the observed slow changes in the SRM 2372 materials As demonstrated in Table 18 absorbance at 260 nm does not necessarily correlate with ddPCR behavior Yet ldquofurthermore the repulsion between phosphate groups on opposite strands tends to separate the complementary strandsrdquo34 Therefore a method that is independent of hypochromism is needed to adequately evaluate the proportion (if any) of ldquotruerdquo (completely disjoint) ssDNA in the SRM materials Circular dichroism spectroscopy is sensitive to the conformational structures of DNA in solution35 and different DNA configurations migrate differentially in agarose gel electrophoresis7 Both techniques have been used to qualitatively discriminate ssDNA from dsDNA in ldquopurerdquo single-sequence materials but neither technique appears to be able to quantify small proportions of ssDNA in a complex multi-chromosome genomic material Wilson and Ellison have proposed a real-time chamber digital PCR (cdPCR) technique that can estimate excess proportions of ssDNA copies from the crossing-threshold distribution31 However their method is insufficiently sensitive for low ssDNA proportions It is well known that heating DNA beyond the melting temperature separates mammalian dsDNA into ssDNA36 albeit we have shown using cdPCR that boiling dsDNA for some minutes also renders some target sites non-amplifiable15 In the following section we document a ddPCR-based approach for estimating the proportion if any of ssDNA in a sample that involves less destructive heating While the method is not yet optimized the results from these preliminary studies demonstrate that the SRM 2372 and 2372a components disperse into the ddPCR droplets as dsDNA ddPCR Measurement of ssDNA Proportion Given two otherwise identical aliquots of a given DNA material one subjected to some treatment and the other untreated the ratio between the average number of copies per droplet of treated material and that of its untreated sibling λtλu quantitatively estimates the change in the number of independently dispersing entities Should the treatment reduce the number of accessible amplifiable targets (AAT) λtλu will be less than 1 Reduction in AAT may result from damage to the target itself andor conformational changes to the fragment containing the target that render the target inaccessible or non-amplifiable16 Assuming no denaturation of dsDNA to ssDNA either because of an ineffective treatment or the material having started as ldquopurerdquo ssDNA if the treatment does not itself damage targets then λtλu will equal 1 Should treatment convert at least some dsDNA to ssDNA without damaging targets the number of AAT will increase and λtλu will be greater than 1 Assuming complete conversion of ldquopurerdquo dsDNA to ldquopurerdquo ssDNA without other damage then λtλu will equal 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

31

Values greater than 1 but less than 2 indicate some conversion of dsDNA to ssDNA but also some combination of (a) incomplete denaturation of dsDNA to ssDNA (b) partial renaturation of ssDNA to dsDNA after treatment but before ddPCR analysis (c) damage to targets during the process or ndash of greatest interest ndash (d) the original sample material was an impure mixture of dsDNA and ssDNA The following discussions graphically summarize sets of λtλu followed by tabular listings of the relevant experimental conditions ddPCR measurements and the numerical λtλu values The information in these tables includes

Plate Index Experiment identifier indicates date performed in YYYYMMDD format Material Sample material SRM 2372-A 2372-B 2372a-A 2372a-B or 2372a-C Assay ddPCR assay HBB1 NEIF or ND6 degC Denaturation temperature in degC Min Length of time the treated material was held at denaturation temperature in

minutes Cool Method used to cool treated material from denaturation temperature Snap

Slow R1 or R5 where ldquoSnaprdquo indicates cooling on ice ldquoSlowrdquo cooling at room temperature to 25 degC ldquoR1rdquo cooling in a thermal cycler to 25 degC at 1 degCs and ldquoR5rdquo cooling in a thermal cycler to 4 degC at 5 degCs

F Dilution factor expressed as (volume Material) per (final volume of diluted solution) 12 14 or 18

n Number of technical replicates Mean Mean of technical replicates CV Coefficient of Variation expressed as a percentage 100 Standard Deviation

Mean

λtλu the copies per droplet result of the treated solution λt divided by the copies per droplet result of its untreated sibling λu

u(λtλu) Standard uncertainty of λtλu 119906119906(120582120582t 120582120582ufrasl ) = 120582120582t 120582120582ufrasl CV(120582120582t)100

2

+ CV(120582120582u)100

2

Proof of Principle Figure 15 summarizes two replicate sets of measurements performed to assess whether the SRM 2372 and 2372a materials respond similarly whether the denatured materials renature if allowed to cool slowly and whether results are (reasonably) repeatable Table 19 details the experimental conditions The λtλu values for the four materials are similar all much greater than 1 but not approaching 2 Most of the ratios are on or close to the ldquoslow cooling is the same as snap coolingrdquo line suggesting that renaturation is not a significant concern While the replicate results for the SRM 2372-A and 2372a-A materials do not overlap as closely as those for SRM 2372-B and 2372a-B similar treatment conditions produce similar results

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

32

Figure 15 Comparison of Snap-Cooling to Slow-Cooling

Symbols mark the means of replicate measurements made using the same dilutions on successive days with bars representing plusmn one standard combined uncertainty Results are labeled by material ldquoArdquo and ldquoBrdquo for SRM 2372 components and ldquoaArdquo ldquoaBrdquo and ldquoaCrdquo for SRM 2372a components The diagonal black line represents equality between the snap- and slow-cooling treatments

Table 19 Comparison of Snap-Cooling to Slow-Cooling

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171218 2372-A HBB1 14 4 03830 29

dd20171218 2372-A HBB1 98 15 Slow 14 2 05644 18 1474 0051 dd20171218 2372-A HBB1 98 15 Snap 14 4 05852 15 1528 0051

dd20171218 2372a-A HBB1 14 4 03029 18

dd20171218 2372a-A HBB1 98 15 Slow 14 4 04430 21 1463 0040 dd20171218 2372a-A HBB1 98 15 Snap 14 4 04454 33 1471 0055

dd20171218 2372-B HBB1 14 4 03594 31

dd20171218 2372-B HBB1 98 15 Slow 14 4 05806 18 1615 0058 dd20171218 2372-B HBB1 98 15 Snap 14 4 05825 13 1620 0055

dd20171218 2372a-B HBB1 14 4 03383 15

dd20171218 2372a-B HBB1 98 15 Slow 14 4 05043 38 1491 0061 dd20171218 2372a-B HBB1 98 15 Snap 14 4 05254 20 1553 0039

dd20171219 2372-A HBB1 14 4 03357 23

dd20171219 2372-A HBB1 98 15 Slow 14 4 05340 16 1591 0046 dd20171219 2372-A HBB1 98 15 Snap 14 4 05340 12 1591 0042

A

aA

B

aB

A

aA

B

aB

14

15

16

17

14 15 16 17

λ tλ

u Sn

ap C

oole

d

λtλu Slow Cooled

Rep 1Rep 2Snap = Slow

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

33

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171219 2372a-A HBB1 14 4 02735 24

dd20171219 2372a-A HBB1 98 15 Slow 14 4 04151 03 1518 0037 dd20171219 2372a-A HBB1 98 15 Snap 14 4 03918 21 1433 0045 dd20171219 2372-B HBB1 14 4 03346 22

dd20171219 2372-B HBB1 98 15 Slow 14 4 05476 17 1637 0045 dd20171219 2372-B HBB1 98 15 Snap 14 4 05480 29 1638 0059 dd20171219 2372a-B HBB1 14 4 03158 19

dd20171219 2372a-B HBB1 98 15 Slow 14 4 04798 16 1519 0038 dd20171219 2372a-B HBB1 98 15 Snap 14 4 04856 17 1538 0039

Time at Denaturation Temperature Figure 16 summarizes λtλu for the SRM 2372-B material as a function of denaturation temperature and the cumulative time the solution is held at that temperature Table 20 details the experimental conditions

Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures

The symbols represent the ratio of the mean result of three technical replicates of the untreated and each of the treated solutions the bars represent combined standard uncertainties The lines denote regression fits to the function λtλu = atimese-bt where t is the cumulative time Values and the standard uncertainties of the coefficients are listed in the legend

12

14

16

18

20

0 5 10 15 20

λ t λ

u

Cummulative Treatment Time min

degC a u (a ) b u (b )90 1804 0039 00198 0000494 1788 0029 00099 0002998 1855 0019 00067 00019

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

34

For all three of the denaturation temperatures studied the relationship between λtλu and cumulative time t can be modeled as

λtλu = ae-bt

where a is the extrapolated value of λtλu at t = 0 and b is the rate of change in AAT per minute As expected the rate of loss due to the treatment increases with increasing treatment temperature and cumulative time at that temperature37 Given the much longer treatment time for the 98 degC series the agreement among the three t = 0 parameter (a) values is remarkable This suggests that the first 1 min of treatment efficiently separates dsDNA into ssDNA The pooled relative standard uncertainty of these regression parameter is 17 somewhat smaller than the 25 pooled relative standard uncertainties of the individual λtλu However the plusmnstandard uncertainty interval of the 1-min λtλu results for the 94 degC and 90 degC both overlap the extrapolated t = 0 value suggesting that reliable results can be obtained using a single treatment of short duration

Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171228 2372-B HBB1 14 3 03650 17 dd20171228 2372-B HBB1 98 5 R1 14 3 06157 12 1687 0035 dd20171228 2372-B HBB1 98 10 R1 14 3 05493 08 1505 0028 dd20171228 2372-B HBB1 98 15 R1 14 3 05019 13 1375 0029 dd20171228 2372-B HBB1 98 20 R1 14 3 04583 21 1256 0034 dd20171228 2372-B HBB1 98 25 R1 14 3 04154 32 1138 0041 dd20171228 2372-B HBB1 98 30 R1 14 3 03784 14 1037 0023 dd20171228 2372-B HBB1 98 35 R1 14 3 03408 10 0934 0019 dd20171228 2372-B HBB1 98 40 R1 14 3 02980 01 0816 0014 dd20171228 2372-B HBB1 98 40 R1 14 3 02815 08 0771 0015

dd20171229 2372-B HBB1 14 3 03708 27

dd20171229 2372-B HBB1 94 1 R1 14 3 06598 26 1779 0056 dd20171229 2372-B HBB1 94 2 R1 14 3 06553 32 1767 0063 dd20171229 2372-B HBB1 94 3 R1 14 3 06301 20 1699 0045 dd20171229 2372-B HBB1 94 4 R1 14 3 06252 14 1686 0037 dd20171229 2372-B HBB1 94 5 R1 14 3 06577 14 1773 0039 dd20171229 2372-B HBB1 94 6 R1 14 3 06156 07 1660 0030 dd20171229 2372-B HBB1 94 7 R1 14 3 06118 10 1650 0033 dd20171229 2372-B HBB1 94 8 R1 14 3 06256 12 1687 0035 dd20171229 2372-B HBB1 94 9 R1 14 3 05999 10 1618 0032

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

35

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180104 2372-B HBB1 14 4 03713 25 dd20180104 2372-B HBB1 90 1 R1 14 3 06916 22 1863 0052 dd20180104 2372-B HBB1 90 3 R1 14 3 06465 06 1741 0032 dd20180104 2372-B HBB1 90 5 R1 14 3 06431 20 1732 0045 dd20180104 2372-B HBB1 90 10 R1 14 3 06014 25 1620 0049 dd20180104 2372-B HBB1 90 15 R1 14 3 06024 10 1623 0032 dd20180104 2372-B HBB1 90 20 R1 14 3 06003 14 1617 0035

Temperature Figure 17 summarizes the effect of denaturation temperature using a single treatment of 05 min on the 2372-B and 2372a-B materials Table 21 details the experimental conditions

Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

The symbols represent the ratio of the mean result of four technical replicates of the untreated and three replicates of the treated solutions the bars represent combined standard uncertainties In all cases the duration of the treatment was 05 min The lines denote regression fits to the linear function λtλu = a +bT where Tis the denaturation temperature

The SRM 2372 and 2372a materials respond very differently to changes in denaturation temperature The older spectrophotometrically ssDNA material suffers somewhat greater loss in λtλu as temperature increases The new spectrophotometrically dsDNA material suffers much greater loss in λtλu as temperature decreases

14

15

16

17

18

19

20

82 86 90 94 98

λ tλ

u

Temperature degC

2372-B

2372a-B

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

36

Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180109 2372-B HBB1 14 4 03577 25 dd20180109 2372-B HBB1 100 05 R5 14 3 06314 16 1765 0051

dd20180104a 2372-B HBB1 14 4 03577 05

dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032 dd20180104a 2372-B HBB1 94 05 R5 14 3 06459 25 1806 0046 dd20180104a 2372-B HBB1 90 05 R5 14 2 06591 08 1842 0017 dd20180104a 2372-B HBB1 88 05 R5 14 3 06589 05 1842 0013 dd20180104a 2372-B HBB1 84 05 R5 14 3 06729 10 1881 0022 dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

We hypothesize that in the older SRM 2372 material the between-strand bond is relatively fragile and can be completely severed at relatively low temperature while avoiding much temperature-dependent AAT loss In contrast the between-strand bond in the SRM 2372a material requires higher temperature to completely disassociate the strands This suggests that there is not a sample-independent ldquooptimumrdquo combination of denaturation temperature and treatment duration However 98 degC for 05 min appears to provide comparable results for these two materials DNA Concentration in the ddPCR Solution Figure 18 summarizes the effect of [DNA] in the ddPCR solution with SRM 2372-B at a denaturation temperature of 98 degC and treatment duration of 05 min Table 22 details the experimental conditions

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

37

Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution

The solid black symbols represent the ratio of the mean result of four technical replicates of the untreated and treated solutions with the λtλu values shown on the left-hand vertical axis the bars represent combined standard uncertainties The solid blue circles represent λu and the open red circles λt with values shown on the right-hand vertical axis at this graphical scale the standard deviations of the technical replicates are covered by the symbol The diagonal blue and red lines represent the log-linear relationships between just the 18 and 14 λ values

The primary effect of [DNA] appears to be the loss of linearity in λt with copiesdroplet values greater than about 1 copiesdroplet There may be some advantage to use somewhat greater dilutions than the routine 14 however the small increase in λtλu may be offset by the increased variability with small λu This suggests the use of dilutions that give λu values of (01 to 04) copiesdroplet

Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180117 2372-B HBB1 12 4 07127 11 dd20180117 2372-B HBB1 98 05 R5 12 4 12331 14 1730 0031

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 3 06151 15 1843 0040

dd20180117 2372-B HBB1 18 4 01549 18 dd20180117 2372-B HBB1 98 05 R5 18 4 02897 20 1870 0051

12 14 18

170

175

180

185

190

02

04

06

08

101214

λ T

arge

t D

ropl

et

λ tλ

u

Dilution Factor

λ₊λᵤλᵤλ₊

λtλu

λu

λt

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

38

Equivalence of Assays Figure 19 summarizes the agreement between the HBB1 NEIF and ND6 assays at a denaturation temperature of 98 degC and treatment duration of 05 min Table 23 details the experimental conditions

Figure 19 Comparison of Assays

The symbols compared the λtλu results from the HBB1 and either the NEIF or ND6 assays Each λtλu value is the ratio of the mean result of four technical replicates of an untreated and three or four replicates of the treated solutions The bars represent combined standard uncertainties The diagonal line represents equality between the assays

There is little or no difference between the λtλu results from the HBB1 NEIF and ND6 assays suggesting that the chromosomal location of the assay target has little influence Since the NEIFHBB1 comparison includes results for both the SRM 2372 and 2372a materials this equivalence is not material-specific

14

16

18

20

14 16 18 20

λ tλ

u NE

IF o

r ND6

λtλu HBB1

NEIFHBB1

ND6HBB1

2nd Assay = HBB1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

39

Table 23 Comparison of Assays

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu)

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045 dd20180105 2372a-B ND6 94 05 R5 14 3 05526 36 1711 0074 dd20180105 2372a-B ND6 90 05 R5 14 3 05326 18 1649 0051 dd20180105 2372a-B ND6 88 05 R5 14 3 05077 22 1572 0052 dd20180105 2372a-B ND6 86 05 R5 14 3 05038 23 1560 0053 dd20180105 2372a-B ND6 84 05 R5 14 3 04906 10 1519 0041

dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

40

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials

Figure 20 summarizes the currently available λtλu results obtained by denaturing at 98 degC for 05 min rapidly cooling in the thermal cycler with dilutions that provide asymp 03 copiesdroplet The figure also displays the consensus result and its 95 confidence region estimated using the Hierarchical Bayes method provided in the NIST Consensus Builder (NICOB) system3839 Table 24 details the experimental conditions

Figure 20 Evaluation of Limiting Ratio

The symbols represent the ratio of the mean result of three or four technical replicates of the untreated and treated solutions the bars represent combined standard uncertainties The thick horizontal line represents a consensus value covering all displayed results the dotted horizontal lines bracket the 95 confidence interval about the consensus value

The Hierarchical Bayes consensus result for these values is 183 with a 95 confidence interval of plusmn003 All but three of the 16 λtλu values are within one standard uncertainty of this consensus value with most of the λtλu means lying within the 95 confidence interval This suggests that there are no statistically significant differences between the SRM 2372 and 2372a materials with regard to ssDNA proportion

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

17

18

19

20

λ tλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

41

Table 24 Evaluation of Limiting Ratio

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180104a 2372-B HBB1 14 4 03577 05 dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032

dd20180108 2372-B HBB1 14 3 03463 09 dd20180108 2372-B HBB1 98 05 R5 14 3 06562 14 1895 0031

dd20180109 2372-B HBB1 14 3 03577 25 dd20180109 2372-B HBB1 98 05 R5 14 3 06495 21 1816 0059

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 4 06151 15 1843 0040

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

42

The agreement among the measured λtλu is more directly visualized using estimated unilateral ldquoDegrees of Equivalencerdquo d Each d is the absolute difference between the measured and consensus values

119889119889 = |120582120582119905119905 120582120582119906119906frasl minus consensus|

and is associated with a 95 confidence interval U95(d) that combines the uncertainty estimates of the measured and consensus values The U95(d) calculation details are specific to the consensus estimation method39 Values of d plusmn U95(d) that include 0 indicate that the λtλu result does not differ from the majority at about a 95 level of confidence Figure 21 displays the d for the λtλu results for the Hierarchical Bayes consensus value39 By this figure of merit all the λtλu can be considered as members of a single population

Figure 21 Degrees of Equivalence

The symbols represent the Degree of Equivalence d for the λtλu results relative to the Hierarchical Bayes consensus value The bars represent 95 level of confidence uncertainty intervals d plusmn U95(d) The horizontal line marks the d = 0 ldquono differencerdquo value Results with d plusmn U95(d) that include this line are not significantly different from the majority

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

-024

-018

-012

-006

000

006

012

018

024

Degr

ees o

f Equ

ival

ence

λtλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

43

Conclusions and Recommendations With high confidence all tubes of the SRM 2372a solution have the same nDNA copy number content within measurement repeatability Additionally within measurement repeatability all tubes of the SRM 2372a have the same mtDNAnDNA With high confidence the nDNA copy number content of the SRM 2372a solution is thermally stable from 4 degC to 37 degC over an extended period The solution should not be shipped or stored below 4 degC With high confidence the ddPCR measurements used to establish the entity and mass concentrations and the mtDNAnDNA ratios are adequately precise and unbiased With high confidence the DNA in the SRM 2372a solutions is now dsDNA and can be expected to remain dsDNA for at least ten years when stored at 4 degC Recommended Certification Values

Since the component materials are believed to be homogeneous are of very similar composition and were prepared and processed similarly the observed differences in the measured entity concentration uncertainties are likely artifacts of the measurement process To limit over-interpretation of these differences we recommend that the copy per nanoliter values of the three components be assigned same 95 relative uncertainty interval (CV95) based on the largest of the measured CV95 values (93 ) rounded up to 10 Since the ngμL transformation does not appreciably inflate this relative uncertainty we recommend that the mass concentration values also be assigned CV95 of 10 Similarly we recommend that the mtDNAnDNA ratios be rounded to integer values all three components be assigned the same CV95 and that this uncertainty be based on the largest (21 ) of the measured CV95 rounded up to 25 Table 25 lists the values that we recommend be used in the Certificate of Analysis

Table 25 Recommended Values and 95 Uncertainties for Certification

Component Units Value U95(Value) a A Copies per nanoliter 151 15 B Copies per nanoliter 175 18 C Copies per nanoliter 145 15 A ngμL 498 50

B ngμL 578 58 C ngμL 479 48 A mtDNAnDNA 174 4

B mtDNAnDNA 206 5 C mtDNAnDNA 279 7

a The interval Value plusmn U95(Value) is believed with about 95 confidence to contain the true value of the measurand

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

44

Use of SRM 2372a for Value-Assigning In-House Reference Materials

The within-component and between-assay variability observed for the commercial qPCR assays is not unexpected since different qPCR methods exploit qualitatively different molecular targets Given the intrinsic variability of the human genome the prospect of developing a human DNA source material that responds equally to all current and future qPCR assays is improbable All three components should be used to elucidate potential bias when using these materials to value assign in-house DNA reference and control materials for qPCR assays

Certificate of Analysis In accordance with ISO Guide 31 2000 a NIST SRM certificate is a document containing the name description and intended purpose of the material the logo of the US Department of Commerce the name of NIST as a certifying body instructions for proper use and storage of the material certified property value(s) with associated uncertainty(ies) method(s) used to obtain property values the period of validity if appropriate and any other technical information deemed necessary for its proper use A Certificate is issued for an SRM certified for one or more specific physical or engineering performance properties and may contain NIST reference information or both values in addition to certified values A Certificate of Analysis is issued for an SRM certified for one or more specific chemical properties Note ISO Guide 31 is updated periodically check with ISO for the latest version [httpswwwnistgovsrmsrm-definitions] For the most current version of the Certificate of Analysis for NIST SRM 2372a Human DNA Quantification Standard please visit httpswww-snistgovsrmorsview_detailcfmsrm=2372a

References

1 National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research Department of Health Education and Welfare (1978) Belmont Report Ethical Principles and Guidelines for the Protection of Human Subjects of Research DHEW Publication No (OS) 78-0012 httpvideocastnihgovpdfohrp_belmont_reportpdf

2 CDCNIH Biosafety in Microbiological and Biomedical Laboratories 5th ed HHS publication No (CDC) 21-1112 Chosewood LC Wilson DE Eds US Government Printing Office Washington DC (2009) httpwwwcdcgovbiosafetypublicationsbmbl5indexhtm

3 Duewer DL Kline MC Redman JW Newall PJ Reeder DJ NIST Mixed Stain Studies 1 and 2 Interlaboratory Comparison of DNA Quantification Practice and Short Tandem Repeat Multiplex Performance with Multiple-Source Samples J Forensic Sci 200146(5)1199-1210 PMID 11569565

4 Kline MC Duewer DL Redman JW Butler JM NIST Mixed Stain Study 3 DNA quantitation accuracy and its influence on short tandem repeat multiplex signal intensity Anal Chem 200375(10)2463-2469 httpsdoiorg101021ac026410i

5 Kline MC Duewer DL Redman JW Butler JM Results from the NIST 2004 DNA Quantitation Study J Forensic Sci 200550(3)571-578 PMID 15932088

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

45

6 Kline MC Duewer DL Travis JC Smith MV Redman JW Vallone PM Decker AE Butler JM Production and Certification of Standard Reference Material 2372 Human DNA Quantitation Standard Anal Bioanal Chem 2009394(4)1183-1192 httpsdoiorg101007s00216-009-2782-0

7 Sambrook J and Russell DW (2001) Molecular Cloning a Laboratory Manual Cold Spring Harbor Laboratory Press Cold Spring Harbor New York

8 EU-JRC Protocol NK603 ndash Community Reference Laboratory Event-specific method for the quantification of maize line NK603 using real-time PCR Section 33 Spectrophotometric Measurement of DNA Concentration httpgmo-crljrceceuropaeusummariesNK603-WEB-ProtocolValidationpdf

9 ISO 215712005(E) Foodstuffs ndash Methods of analysis for the detection of genetically modified organisms and derived products ndash Nucleic acid extraction Annex B Methods for quantitation of extracted DNA pp 34-36 International Organization for Standardization Geneva (2005)

10 Certificate of Analysis Standard Reference Material 2372 Human DNA Quantitation Standard (2013) Gaithersburg MD USA Standard Reference Materials Program National Institute of Standards and Technology httpswww-snistgovsrmorsview_certcfmsrm=2372

11 Vogelstein B Kinzler KW Digital PCR Proc Natl Acad Sci USA 1999969236-9241 12 Hindson BJ Ness KD Masquelier DA Belgrader P Heredia NJ Makarewicz AJ Bright IJ Lucero

MY Hiddessen AL Legler TC Kitano TK Hodel MR Petersen JF Wyatt PW Steenblock ER Shah PH Bousse LJ Troup CB Mellen JC Wittmann DK Erndt NG Cauley TH Koehler RT So AP Dube S Rose KA Montesclaros L Wang S Stumbo DP Hodges SP Romine S Milanovich FP White HE Regan JF Karlin-Neumann GA Hindson CM Saxonov S Colston BW High-Throughput Droplet Digital PCR System for Absolute Quantitation of DNA Copy Number Anal Chem 2011838604-8610 httpsdoiorg101021ac202028g

13 Pinheiro LB Coleman VA Hindson CM Herrmann J Hindson BJ Bhat S Emslie KR Evaluation of a Droplet Digital Polymerase Chain Reaction Format for DNA Copy Number Quantification AnalChem 2012841003-1011 httpsdoiorg101021ac202578x

14 Corbisier P Pinheiro L Mazoua S Kortekaas AM Chung PY Gerganova T Roebben G Emons H Emslie K DNA copy number concentration measured by digital and droplet digital quantitative PCR using certified reference materials Anal Bioanal Chem 2015407(7)1831-1840 httpsdoiorg101007s00216-015-8458-z

15 Duewer DL Kline MC Romsos EL Real-time cdPCR opens a window into events occurring in the first few PCR amplification cycles Anal Bioanal Chem 2015407(30)9061-9069 httpsdoiorg101007s00216-015-9073-8

16 Kline MC Romsos EL Duewer DL Evaluating Digital PCR for the Quantification of Human Genomic DNA Accessible Amplifiable Targets Anal Chem 201688(4)2132-2139 httpsdoiorg101021acsanalchem5b03692

17 Kline MC Duewer DL Evaluating Droplet Digital Polymerase Chain Reaction for the Quantification of Human Genomic DNA Lifting the Traceability Fog Anal Chem 201789(8)4648-4654 httpsdoiorg101021acsanalchem7b00240

18 Duewer DL Kline MC Romsos EL Toman B Evaluating Droplet Digital PCR for the quantification of human genomic DNA Conversion from Copies per Nanoliter to Nanogram Genomic DNA per Microliter Anal Bioanal Chem 2018in press httpsdoiorg101007s00216-018-0982-1

19 Dagata JA Farkas N Kramer JA Method for measuring the volume of nominally 100 μm diameter spherical water in oil emulsion droplets NIST Special Publication 260-184 2016 httpsdoiorg106028NISTSP260-184

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

20 ISO 170342016 General requirements for the competence of reference material producers International Organization for Standardization Geneva (2016)

21 NCBI Standard Nucleotide BLAST httpsblastncbinlmnihgovBlastcgiPAGE_TYPE=BlastSearch

22 Timken MD Swango KL Orrego C Buoncristiani MR A Duplex Real-Time qPCR Assay for the Quantification of Human Nuclear and Mitochondrial DNA in Forensic Samples Implications for Quantifying DNA in Degraded Samples J Forensic Sci 200550(5)1044-60 PMID 16225209

23 Walker JA Hedges DJ Perodeau BP Landry KE Stoilova N Laborde ME Shewale J Sinha SK Batzer MA Multiplex polymerase chain reaction for simultaneous quantitation of human nuclear mitochondrial and male Y-chromosome DNA application in human identification Anal Biochem 200533789ndash97 httpsdoiorg101016jab200409036

24 Kavlick MF Lawrence HS Merritt RT Fisher C Isenberg A Robertson JM Budowle B Quantification of Human Mitochondrial DNA Using Synthesized DNA Standards J Forensic Sci 201156(6) 1457ndash1463 httpsdoiorg101111j1556-4029201101871x

25 Stulnig TM Amberger A Exposing contaminating phenol in nucleic acid preparations BioTechniques 199416(3) 403-404

26 Majumdar N Wessel T Marks J Digital PCR Modeling for Maximal Sensitivity Dynamic Range and Measurement Precision PLoS ONE 201510(3)e0118833 httpsdoi101371journalpone0118833

27 Tellinghuisen J Partition Volume Variability in Digital Polymerase Chain Reaction Methods Polydispersity Causes Bias but Can Improve Precision Anal Chem 2016881283-12187 httpsdoiorg101021acsanalchem6b03139

28 Lunn DJ Spiegelhalter D Thomas A Best N (2009) Statistics in Medicine 283049ndash3082 29 De Biegravevre P Dybkaer R Fajgelj A Hibbert DB Metrological traceability of measurement results in

chemistry Concepts and implementation Pure Appl Chem 201183(10)1873-1935 httpsdoiorg101007s00769-011-0805-y

30 Huggett JF Foy CA Benes V Emslie K Garson JA Haynes R Hellemans J Kubista M Mueller RD Nolan T Pfaffl MW Shipley GL Vandesompele J Wittwer CT Bustin SA The Digital MIQE Guidelines Minimum Information for Publication of Quantitative Digital PCR Experiments Clin Chem 201359(6)892-902 httpsdoiorg101373clinchem2013206375

31 Wilson PJ Ellison SLR Extending digital PCR analysis by modelling quantification cycle data BMC Bioinfo 201617421 httpsdoiorg101186s12859-016-1275-3

32 Wang X Son A Effects of pretreatment on the denaturation and fragmentation of genomic DNA for DNA hybridization Environ Sci Processes Impacts 2013152204 httpsdoiorg101039c3em00457k

33 Nwokeoji AO Kilby PM Portwood DE Dickman MJ Accurate Quantification of Nucleic Acids Using Hypochromicity Measurements in Conjunction with UV Spectrophotometry Anal Chem 201789(24)13567-13574 https101021acsanalchem7b04000

34 Farkas DH Kaul KL Wiedbrauk DL Kiechle FL Specimen collection and storage for diagnostic molecular pathology investigation Arch Pathol Lab Med 1996120(6)591-6

35 Kypr J Kejnovskaacute I Renčiuk D Vorličkovaacute M Circular dichroism and conformational polymorphism of DNA Nucleic Acids Res 200937(6)1713-1725 httpsdoiorg101093nargkp026

36 Strachan T Read AP (1999) Human Molecular Genetics 2nd Ed John Wiley amp Sons NY

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

37 Bhat S McLaughlin JL Emslie KR Effect of sustained elevated temperature prior to amplification on template copy number estimation using digital polymerase chain reaction Analyst 2011136724ndash32 httpsdoiorg101039c0an00484g

38 NIST Consensus Builder (NICOB) httpsconsensusnistgov 39 Koepke A Lafarge T Toman B Possolo A NIST Consensus Builder Userrsquos Manual

httpsconsensusnistgovNISTConsensusBuilder-UserManualpdf

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

Appendix A DNA Extraction Method The following method has been adapted from Miller et al (1988)A1 Procedure

1 Transfer approximately 5 mL of buffy coat into a 50 mL sterile polypropylene centrifuge tube and add 30 mL Red Blood Cell Lysis Buffer

2 Mix by gentle inversion and let stand at room temperature for 15 min

3 Centrifuge at approximately 209 rads (2000 rpm) for 20 min

4 Discard supernatant

5 Add 9 mL of Nucleic Lysis Buffer and re-suspend the pelleted cells by agitating with a plastic transfer pipette

6 Add 300 microL of 20 sodium dodecyl sulfate (SDS) and 18 mL of Protease K Solution (freshly prepared 2 mg to 5 mg Protease K per 1 mL Protease K Buffer)

7 Mix by inversion followed by incubation overnight or longer at 37 ordmC to 40 ordmC in a shaking water bath The length of incubation is determined by the time required to solubilize the cells present in the tube

8 Add 4 mL saturated ammonium acetate solution (in deionized water stored at 2 ordmC to 8 ordmC) Shake vigorously for 15 s

9 Centrifuge at approximately 471 rads (4500 rpm) for 10 min

10 Pipette supernatant into a 50 mL sterile polypropylene centrifuge tube and add two volumes of room temperature absolute ethanol Mix by gentle inversion until the DNA aggregates and floats to the top of the tube

11 Spool the aggregated DNA from the current tube into a fresh tube of 80 ethanol using a sterile inoculation loop At this step multiple tubes of precipitated DNA may be pooled to make a large lot of the purified material Mix by gentle inversion

12 Transfer the aggregated DNA to a fresh dry 50 mL polypropylene centrifuge tube Remove as much ethanol as possible by squeezing the DNA against the wall of the tube with a sterile inoculation loop Air-dry the material in a laminar flow hood

13 Resolubilize DNA in TE-4 buffer and store material at 4 ordmC Reagents

Red Blood Cell Lysis Buffer 144 mmolL ammonium chloride 1 mmolL sodium bicarbonate in deionized (DI) water Store at 2 ordmC to 8 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

Nucleic Lysis Buffer 10 mmolL Tris-HCl 400 mmolL sodium chloride 20 mmolL disodium ethylenediaminetetraacetic acid in DI Water adjust pH to 82 Store at 2 ordmC to 8 ordmC

Protease K Buffer

20 mmolL disodium ethylenediaminetetraacetic acid 10 (mass per volume) Sodium Dodecyl Sulfate (SDS) in DI water Store at 2 ordmC to 8 ordmC

Protease K

Sigma-Aldrich cat P6556 Tritirachium album lyophilized powder Store at ndash20 ordmC

Ammonium Acetate

192 molL ammonium acetate (saturated) in DI water Store at 2 ordmC to 8 ordmC

Absolute Ethanol (200 proof)

Store at room temperature Reference

A1 Miller SA Dykes DD Polesky HF A simple salting out procedure for extracting DNA from human nucleated cells Nucleic Acids Res 198816(3)1215 httpsdoiorg101093nar1631215

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

48

Appendix B Spectrophotometric Measurement Protocol The following is for use with a scanning recording spectrophotometer The procedures include the spectrophotometer calibration measurement of the ldquonativerdquo dsDNA and measurement of the NaOH denatured ssDNA The procedure for evaluating solutions after treatment with NaOH is adapted from references89 Spectrophotometer Calibration

Establish Baseline Set the spectrophotometer to acquire data from 220 nm to 345 nm in steps of 01 nm Zero the instrument with nothing (air) in the sample and reference cells Acquire an air-air baseline of the instrument with nothing (air) in the sample and reference cells If the trace has significant structure or excessive noise take corrective action If the trace is acceptable proceed with data collection Enable Verification of Wavelength Calibration SRM 2034 is a solution of holmium oxide in 10 perchloric acid contained in a sealed quartz cuvette delivering an intrinsic UVVis wavelength standard for absorption spectrophotometryB1 Evaluation of the location of peak maxima in the spectra of this solution enables verification of the spectrophotometers wavelength calibration Place SRM 2034 in the sample cell Create and name a file to hold the verification spectra Perform three replicate SRM 2034-to-air scans saving each scan to the file Remove SRM 2034 from the spectrophotometer and store appropriately Enable Verification of Absorbance Calibration SRM 2031 consists of three metal-on-fused silica filters held in cuvette-sized filter holders plus an empty filter holder The three filters are certified for transmittance and transmittance density (-log10(transmittance) effectively absorbance) at several wavelengths with nominal wavelength-independent transmittances of 10 30 and 90 Evaluation of the absorbance of each of these three filters at the certified 250 nm 280 nm and 340 nm wavelengths enables verification of the spectrophotometers absorbance calibration Place the empty SRM 2031 filter holder in the reference cell Remove the protective covers from the 10 transmittance filter and place in the sample cell Perform three replicate scans saving each scan to file Remove the filter from the spectrophotometer replace the protective covers and store appropriately Repeat for the 30 and 90 filters Close the file containing the verification spectra For ldquonativerdquo dsDNA Samples

Verify Optical Balance of Cuvettes The measurements of the DNA-containing samples are made using a 100 microL quartz cuvette in the sample cell compared to an optically matched cuvette in the reference cell The optical equivalence of these two cuvettes can be verified by filling both cuvettes with appropriate volumes of a sample solution matrix (TE-4 buffer) that does not contain DNA Fill two dry

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

49

clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes followed by a methanol rinse in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquonativerdquo dsDNA remove and clean both the reference cuvette and the sample cuvette For NaOH denatured ssDNA Samples

Prepare a 04 molL NaOH Stock Solution To a 15 mL disposable centrifuge tube add 016 g NaOH pellets add deionized (DI) water to the 10 mL line and mix until all the NaOH pellets have dissolved Centrifuge at 524 rads (5000 rpm) for 5 min For every 9 samples to be analyzed aliquot 1 mL of this 04 molL NaOH stock solution into a 15 mL screw top tube Use these tubes to prepare the NaOH blanks and DNA samples In a 15 mL screw cap vial mix 200 microL TE-4 buffer with 200 microL 04 molL sodium hydroxide solution to achieve a final NaOH concentration of 02 molL Vortex 5 s then centrifuge at 1257 rads (12 000 rpm) for 1 min Fill two dry clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Prepare the Sample Solutions For each unit of material to be measured (Tube or vial)

a) Vortex the unit for 5 s b) Centrifuge at 1257 rads (12 000 rpm) for 1 min c) Remove a 100 microL aliquot and place in a new labeled vial d) Add 100 microL of the freshly prepared 04 molL NaOH solution (with the same pipet) e) Vortex 5 s f) Centrifuge at 1257 rads (12 000 rpm) for 1 min

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

50

Measure Sample Solutions Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes rinse with methanol in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the NaOH denatured DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquodenaturedrdquo ssDNA close the file holding the spectra for this series Remove and clean both the reference cuvette and the sample cuvette After All Sample Measurements are Complete

Enable Verification of Spectrophotometer Stability The stability of the spectrophotometer over the course of the data collection can be evaluated by comparing spectra for SRM 2034 and 2031 taken at the start of data collection with spectra taken at the end of data collection Repeat the wavelength and absorbance verification measurements described above collecting the spectra in a suitably named file Clean Up Transfer all relevant data files from the spectrophotometers storage to a transfer device Clean the spectrophotometer work area as necessary and restore the SRMs and cuvettes to their appropriate storage locations Evaluation

Export the files to a spreadsheet Average the replicate spectra Subtract the average absorbance at A320 from the absorbance at A260 resulting in the corrected absorbance at A260 The conventional spectrophotometric estimate of the mass concentration of the dsDNA in the sample is achieved by multiplying the corrected absorbance by 50 The conventional estimate for the NaOH-treated ssDNA is achieved by multiplying the corrected absorbance by 37times2 where the ldquotimes2rdquo adjusts for the 1rarr2 volumetric dilution of the sample Note reliable spectrophotometric measurements require corrected A260 to be greater than 005 Reference

B1 Travis JC Acosta JC Andor G Bastie J Blattner P Chunnilall CJ Crosson SC Duewer DL Early EA Hengstberger F Kim C-S Liedquist L Manoocheri F Mercader F Metzdorf J Mito A Monardh LAG Nevask S Nilsson M Noeumll M Rodriguez AC Ruiacutez A Schirmacher A Smith MV Valencia G van Tonder N Zwinkels J Intrinsic Wavelength Standard Absorption Bands in Holmium Oxide Solution for UVvisible Molecular Absorption Spectrophotometry J Phys Chem Ref Data 200534(1)41-56

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

51

Appendix C SRM 2372a Statisticianrsquos report Extracted from Blaza Toman

November 30 2017 1 Introduction Candidate material for SRM 2372a is composed of three components labeled A B and C All components are human genomic DNA and prepared at NIST from buffy coat white blood cells from single-source anonymous donors Component A and B are single source (malefemale respectively) and Component C is a 1 male to 3 female mixture of two single-source donors Droplet digital PCR (ddPCR) assays were used to obtain measurements of the concentration of copies per microliter of sample Ten NIST-developed human nuclear DNA and three literature-derived human mitochondrial DNA PCR assays were used in this study

11 OpenBUGS Analysis All parameters in the following sections were estimated using Markov Chain Monte Carlo (MCMC) implemented in the OpenBUGS freewareC1 OpenBUGS is a software application for the Bayesian analysis of complex statistical models

12 OpenBUGS Exemplar Output All following sections list the OpenBUGs model (ldquocoderdquo) initialization values (ldquoinitsrdquo) if they are needed and the data used to evaluate the various data sets discussed in the body of this report homogeneity stability nDNA certification and mtDNAnDNA value assignment The provided information in each of the following sections is sufficient to produce estimates but the estimates may not exactly reproduce the exemplar results provided The MCMC paradigm does not produce ldquoanalyticalrdquo values but approaches them asymptotically as the number o model updates increases All exemplar results presented here are based upon the following bull 10-fold ldquothinningrdquo to minimize autocorrelation there were 10 MCMC estimates for every

model update used in parameter estimation bull a ldquoburn inrdquo of 5 000 updates before beginning parameter estimation and bull 10 000 model updates

The following statistics are provided in the tables listing the exemplar results

mean arithmetic mean of the parameterrsquos model update values sd standard deviation of the parameterrsquos model update values MC_error Monte Carlo standard deviation of the mean meanradic(number of model updates) val25pc 25th percentile of the model update values median 50th percentile of the model update values val975pc 975th percentile of the model update values startnumber of ldquoburn inrdquo updates sample number of model update values

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

52

2 Homogeneity assessment 21 Data

Samples of the material of all three components were drawn from 22 different boxes Two assays (NEIF ND6) were used to obtain 3 replicates of the measurement in copies per nanoliter for each sample and component For the complete homogeneity data see

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

22 Model The data were analyzed using a three-level Gaussian hierarchical ANOVA model

119910119910119894119894119894119894119894119894~119873119873120572120572119894119894119894119894119894119894 120590120590119908119908119894119894119905119905ℎ119894119894119894119894 119887119887119887119887119887119887 1198941198941198941198942

where ~N() denotes ldquodistributed as a normal distribution of some mean μ and variancerdquo 119894119894 denotes component (1 for A 2 for B 3 for C) 119895119895 denotes assay (1 for NEIF 2 for ND6) 119896119896 denotes box (1hellip22) and 120572120572119894119894119894119894119894119894~119873119873micro119894119894119894119894120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119887119887119887119887119887119887119894119894

2

All variance components were given a Half Cauchy prior as described in Reference C2 Note The Half Cauchy prior is specified by ldquo~dnorm(0 00016)I(0001)rdquo The ldquoIrdquo notation is for censoring It cuts off any random draws from the Normal distribution that are lower than 0001 An informative prior is needed when there are less than four data per distribution and the half Cauchy is a good general-purpose prior

23 OpenBUGS code and data Homogeneity code ModelBegin for(i in 13)xiNsb[i] ~ dnorm(0 00016)I(0001) chSqNsb[i] ~ dgamma(0505) sigb[i] lt- xiNsb[i]sqrt(chSqNsb[i]) for(i in 13)for(j in 12)mu[ij]~dnorm(010E-5) for(i in 13) muA[i]lt-mean(mu[i])sigbP[i]lt-100(muA[i]sqrt(sigb[i])) for(i in 13)for(j in 12)xiNs[ij] ~ dnorm(0 00016)I(0001) chSqNs[ij]~dgamma(0505) sigt[ij]lt-xiNs[ij]sqrt(chSqNs[ij]) sigtP[ij]lt-100(mu[ij]sqrt(sigt[ij])) for(i in 13)for(j in 12)for(k in 122)alpha[ijk]~dnorm(mu[ij]sigb[i]) for(idx in 1393)lamb[idx]~dnorm(alpha[cmp[idx]asy[idx]box[idx]]sigt[cmp[idx]asy[idx]]) ModelEnd Inits None required

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

53

Homogeneity Data lamb[] asy[] box[] cmp[] rep[] 1632 1 1 1 1 1529 1 1 1 2 1511 1 1 1 3 1616 2 1 1 1 1640 2 1 1 2 1582 2 1 1 3 1531 1 2 1 1 1525 1 2 1 2 1549 1 2 1 3 1615 2 2 1 1 1589 2 2 1 2 1601 2 2 1 3 1524 1 3 1 1 1540 1 3 1 2 1533 1 3 1 3 1582 2 3 1 1 1152 2 3 1 2 1555 2 3 1 3 1454 1 4 1 1 1458 1 4 1 2 1427 1 4 1 3 1627 2 4 1 1 1605 2 4 1 2 1607 2 4 1 3 1465 1 5 1 1 1540 1 5 1 2 1504 1 5 1 3 1607 2 5 1 1 1633 2 5 1 2 1586 2 5 1 3 1512 1 6 1 1 1491 1 6 1 2 1491 1 6 1 3 1511 2 6 1 1 1483 2 6 1 2 1524 2 6 1 3 1418 1 7 1 1 1480 1 7 1 2 1455 1 7 1 3 1479 2 7 1 1 1557 2 7 1 2 1656 2 7 1 3 1496 1 8 1 1 1560 1 8 1 2 1501 1 8 1 3 1514 2 8 1 1 1615 2 8 1 2 1540 2 8 1 3 1588 1 9 1 1 1581 1 9 1 2 1637 1 9 1 3 1474 2 9 1 1 1438 2 9 1 2 1454 2 9 1 3 1574 1 10 1 1 1513 1 10 1 2 1570 1 10 1 3 1594 2 10 1 1 1713 2 10 1 2 1669 2 10 1 3

1557 1 11 1 1 1508 1 11 1 2 1633 1 11 1 3 1571 2 11 1 1 1611 2 11 1 2 1631 2 11 1 3 1500 1 12 1 1 1476 1 12 1 2 1515 1 12 1 3 1626 2 12 1 1 1586 2 12 1 2 1608 2 12 1 3 1672 1 13 1 1 1634 1 13 1 2 1804 1 13 1 3 1490 2 13 1 1 1516 2 13 1 2 1497 2 13 1 3 1551 1 14 1 1 1568 1 14 1 2 1538 1 14 1 3 1617 2 14 1 1 1614 2 14 1 2 1639 2 14 1 3 1456 1 15 1 1 1448 1 15 1 2 1523 1 15 1 3 1525 2 15 1 1 1402 2 15 1 2 1436 2 15 1 3 1573 1 16 1 1 1520 1 16 1 2 1617 1 16 1 3 1634 2 16 1 1 1627 2 16 1 2 1575 2 16 1 3 1516 1 17 1 1 1516 1 17 1 2 1628 1 17 1 3 1613 2 17 1 1 1567 2 17 1 2 1567 2 17 1 3 1612 1 18 1 1 1535 1 18 1 2 1672 1 18 1 3 1529 2 18 1 1 1495 2 18 1 2 1528 2 18 1 3 1598 1 19 1 1 1514 1 19 1 2 1596 1 19 1 3 1571 2 19 1 1 1535 2 19 1 2 1511 2 19 1 3 1575 1 20 1 1 1568 1 20 1 2 1586 1 20 1 3 1652 2 20 1 1 1545 2 20 1 2 1627 2 20 1 3 1658 1 21 1 1

1611 1 21 1 2 1578 1 21 1 3 1647 2 21 1 1 1607 2 21 1 2 1612 2 21 1 3 1547 1 22 1 1 1528 1 22 1 2 1532 1 22 1 3 1537 2 22 1 1 1513 2 22 1 2 1536 2 22 1 3 1790 1 1 2 1 1909 1 1 2 2 1885 1 1 2 3 1870 2 1 2 2 1860 2 1 2 3 1712 1 2 2 1 1719 1 2 2 2 1704 1 2 2 3 1775 2 2 2 1 1718 2 2 2 2 1756 2 2 2 3 1693 1 3 2 1 1755 1 3 2 2 1749 1 3 2 3 1824 2 3 2 1 1483 2 3 2 2 1896 2 3 2 3 1754 1 4 2 1 1745 1 4 2 2 1739 1 4 2 3 1670 2 4 2 1 1632 2 4 2 2 1814 2 4 2 3 1757 1 5 2 1 1687 1 5 2 2 1735 1 5 2 3 1683 2 5 2 1 1759 2 5 2 2 1703 2 5 2 3 1697 1 6 2 1 1709 1 6 2 2 1714 1 6 2 3 1693 2 6 2 1 1723 2 6 2 2 1722 2 6 2 3 1722 1 7 2 1 1625 1 7 2 2 1703 1 7 2 3 1690 2 7 2 1 1684 2 7 2 2 1688 2 7 2 3 1884 1 8 2 1 1802 1 8 2 2 1772 1 8 2 3 1864 2 8 2 1 1887 2 8 2 2 1821 2 8 2 3 1713 1 9 2 1 1734 1 9 2 2 1727 1 9 2 3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

54

1709 2 9 2 1 1756 2 9 2 2 1758 2 9 2 3 1759 1 10 2 1 1809 1 10 2 2 1738 1 10 2 3 1764 2 10 2 1 1816 2 10 2 2 1787 2 10 2 3 1749 1 11 2 1 1782 1 11 2 2 1740 1 11 2 3 1741 2 11 2 1 1792 2 11 2 2 1784 2 11 2 3 1748 1 12 2 1 1712 1 12 2 2 1718 1 12 2 3 1797 2 12 2 1 1614 2 12 2 2 1790 2 12 2 3 1632 1 13 2 1 1668 1 13 2 2 1701 1 13 2 3 1626 2 13 2 1 1566 2 13 2 2 1681 2 13 2 3 1752 1 14 2 1 1535 1 14 2 2 1673 1 14 2 3 1724 2 14 2 1 1741 2 14 2 2 1685 2 14 2 3 1871 1 15 2 1 1835 1 15 2 2 1887 1 15 2 3 1770 2 15 2 1 1752 2 15 2 2 1798 2 15 2 3 1688 1 16 2 1 1725 1 16 2 2 1743 1 16 2 3 1744 2 16 2 1 1697 2 16 2 2 1710 2 16 2 3 1789 1 17 2 1 1769 1 17 2 2 1825 1 17 2 3 1716 2 17 2 1 1762 2 17 2 2 1638 2 17 2 3 1826 1 18 2 1 1780 1 18 2 2 1849 1 18 2 3 1805 2 18 2 1 1801 2 18 2 2 1931 2 18 2 3 1751 1 19 2 1 1759 1 19 2 2 1880 1 19 2 3 1827 2 19 2 1 1766 2 19 2 2

1786 2 19 2 3 1738 1 20 2 1 1678 1 20 2 2 1768 1 20 2 3 1752 2 20 2 1 1747 2 20 2 2 1768 2 20 2 3 1742 1 21 2 1 1836 1 21 2 2 1713 1 21 2 3 1746 2 21 2 1 1744 2 21 2 2 1764 2 21 2 3 1761 1 22 2 1 1743 1 22 2 2 1660 1 22 2 3 1666 2 22 2 1 1692 2 22 2 2 1819 2 22 2 3 1493 1 1 3 1 1385 1 1 3 2 1510 1 1 3 3 1565 2 1 3 1 1562 2 1 3 2 1466 2 1 3 3 1516 1 2 3 1 1513 1 2 3 2 1577 1 2 3 3 1490 2 2 3 1 1547 2 2 3 2 1570 2 2 3 3 1451 1 3 3 1 1528 1 3 3 2 1506 1 3 3 3 1485 2 3 3 1 1505 2 3 3 2 1471 2 3 3 3 1440 1 4 3 1 1486 1 4 3 2 1520 1 4 3 3 1529 2 4 3 1 1508 2 4 3 2 1506 2 4 3 3 1523 1 5 3 1 1472 1 5 3 2 1483 1 5 3 3 1513 2 5 3 1 1520 2 5 3 2 1491 2 5 3 3 1437 1 6 3 1 1493 1 6 3 2 1522 1 6 3 3 1492 2 6 3 1 1481 2 6 3 2 1461 2 6 3 3 1485 1 7 3 1 1477 1 7 3 2 1463 1 7 3 3 1475 2 7 3 1 1547 2 7 3 2 1485 2 7 3 3 1433 1 8 3 1

1405 1 8 3 2 1478 1 8 3 3 1497 2 8 3 1 1501 2 8 3 2 1495 2 8 3 3 1463 1 9 3 1 1499 1 9 3 2 1505 1 9 3 3 1503 2 9 3 1 1491 2 9 3 2 1516 2 9 3 3 1534 1 10 3 1 1458 1 10 3 2 1454 1 10 3 3 1448 2 10 3 1 1465 2 10 3 2 1474 2 10 3 3 1533 1 11 3 1 1514 1 11 3 2 1555 1 11 3 3 1488 2 11 3 1 1470 2 11 3 2 1483 2 11 3 3 1450 1 12 3 1 1525 1 12 3 2 1497 1 12 3 3 1496 2 12 3 1 1483 2 12 3 2 1519 2 12 3 3 1537 1 13 3 1 1527 1 13 3 2 1515 1 13 3 3 1499 2 13 3 1 1479 2 13 3 2 1475 2 13 3 3 1515 1 14 3 1 1490 1 14 3 2 1494 1 14 3 3 1469 2 14 3 1 1480 2 14 3 2 1466 2 14 3 3 1511 1 15 3 1 1568 1 15 3 2 1594 1 15 3 3 1515 2 15 3 1 1543 2 15 3 2 1565 2 15 3 3 1449 1 16 3 1 1452 1 16 3 2 1471 1 16 3 3 1543 2 16 3 1 1448 2 16 3 2 1491 2 16 3 3 1388 1 17 3 1 1421 1 17 3 2 1394 1 17 3 3 1452 2 17 3 1 1435 2 17 3 3 1475 1 18 3 1 1455 1 18 3 2 1445 1 18 3 3 1458 2 18 3 1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

55

1531 2 18 3 2 1459 2 18 3 3 1468 1 19 3 1 1525 1 19 3 2 1446 1 19 3 3 1435 2 19 3 1 1478 2 19 3 2 1421 2 19 3 3 1549 1 20 3 1

1531 1 20 3 2 1499 1 20 3 3 1499 2 20 3 1 1465 2 20 3 2 1468 2 20 3 3 1530 1 21 3 1 1440 1 21 3 2 1484 1 21 3 3 1499 2 21 3 1

1454 2 21 3 2 1442 1 22 3 1 1487 1 22 3 2 1436 1 22 3 3 1589 2 22 3 1 1591 2 22 3 2 1556 2 22 3 3 END

Table C-1 Example OpenBUGS Homogeneity Summary

mean sd MC_error val25pc median val975pc start sample mu[11] 15460 0122 000123 1522 1546 1570 5001 10000 mu[12] 15640 0137 000141 1537 1564 1591 5001 10000 mu[21] 17490 0113 000118 1727 1749 1772 5001 10000 mu[22] 17480 0129 000125 1722 1748 1774 5001 10000 mu[31] 14870 0073 000073 1473 1487 1501 5001 10000 mu[32] 14960 0069 000063 1482 1496 1509 5001 10000 muA[1] 15550 0092 000092 1536 1555 1573 5001 10000 muA[2] 17490 0085 000086 1732 1749 1765 5001 10000 muA[3] 14910 0050 000046 1481 1491 1501 5001 10000 sigbP[1] 3269 04985 000514 2376 3244 4324 5001 10000 sigbP[2] 2548 04385 000461 1733 2533 3472 5001 10000 sigbP[3] 1850 0292 000260 1329 1832 2477 5001 10000

sigtP[11] 2774 0298 000260 2268 2750 3442 5001 10000 sigtP[12] 4036 0429 000438 3290 4005 4986 5001 10000 sigtP[21] 2664 0301 000321 2155 2640 3313 5001 10000 sigtP[22] 3807 0395 000377 3116 3772 4662 5001 10000 sigtP[31] 2313 0250 000209 1880 2291 2873 5001 10000 sigtP[32] 1812 0195 000170 1473 1797 2243 5001 10000

24 Results The material appears to be homogeneous as can be seen in Figure C-1 Based on these results the between-tube variability is of the same order as within-tube variability

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

56

Component A Assay NEIF Component A Assay ND6 box plot gamma[11]

[111][112][113]

[114]

[115][116]

[117]

[118]

[119]

[1110][1111]

[1112]

[1113]

[1114]

[1115]

[1116][1117]

[1118]

[1119][1120]

[1121]

[1122]

114

box plot gamma[12]

[121][122]

[123]

[124][125]

[126]

[127][128]

[129]

[1210]

[1211][1212]

[1213]

[1214]

[1215]

[1216]

[1217]

[1218][1219]

[1220][1221]

[1222]

115

Component B Assay NEIF Component B Assay ND6

box plot gamma[21]

[211]

[212]

[213][214]

[215][216]

[217]

[218]

[219]

[2110][2111]

[2112]

[2113][2114]

[2115]

[2116]

[2117]

[2118]

[2119]

[2120]

[2121]

[2122]

129

box plot gamma[22]

[221]

[222][223]

[224][225][226][227]

[228]

[229]

[2210][2211]

[2212]

[2213]

[2214]

[2215]

[2216][2217]

[2218]

[2219]

[2220][2221][2222]

128

Component C Assay NEIF Component C Assay ND6

box plot gamma[31]

[311]

[312]

[313][314]

[315][316]

[317]

[318]

[319][3110]

[3111]

[3112]

[3113]

[3114]

[3115]

[3116]

[3117]

[3118]

[3119]

[3120]

[3121]

[3122]109

box plot gamma[32]

[321][322]

[323]

[324][325]

[326]

[327][328][329]

[3210]

[3211]

[3212]

[3213][3214]

[3215]

[3216]

[3217]

[3218]

[3219]

[3220][3221]

[3222]

11

Figure C-1 Homogeneity Assessment

Note These results were obtained using the 1-4 gravimetric dilution-adjusted λ but uncorrected for droplet volume

or the 1-10 volumetric dilution That is the vertical axis in each panel should be multiplied by 07349times10 = 7349 This constant difference in scale does not impact the homogeneity analysis Error bars span approximate 95 confidence intervals

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

57

3 Certification measurements 31 Data

Originally the experiment consisted of 5 plates containing the 3 components (A B C) and 12 assays Ten of the assays were for genomic DNA identification (for the certified value) and two of the assays were for mitochondrial detection which is used for informational values (and are not included in this data set) Each of the plates contained samples from a new vial from different boxes (noted in the ldquovial usedrdquo portion the number references the box) The assay positions were alternated across the five plates as well For the plate design and which tubes were used see

Figure 6 Basic Plate Layout Design for Certification Measurements It was later determined that the homogeneity and stability data sets were obtained using a different set of pipettes than the certification data set and since all data was to be combined for certification another plate was obtained using the original pipette For the complete certification data see

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component Stability data was measured at different temperatures over several weeks using assays NEIF and ND6 Only the 4degC data was also used for certification For the complete stability data see

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

32 Model It was decided that all three sets of data should be used to derive the certified values In this way the uncertainty of the certified values would better reflect any potential heterogeneity in the samples as well as any additional unknown factors that may have effected the results Because two different pipettes were used to obtain different sets of data there was a significant bias in some of the measurements Specifically the ldquocertificationrdquo data from the first five plates was obtained using pipette set ldquo1rdquo and the remaining measurements were obtained using pipette set ldquo2rdquo This bias was introduced into the statistical model as part of the mean Possible effects of different assays and type of data (heterogeneity certification and stability) were accounted for in the statistical model in terms of a random effect Specifically the response for each component (A B C) was represented as

119910119910119894119894119894119894119894119894~1198731198731205741205741198941198941198941198941198941198941205901205901198941198942 where ~N() denotes ldquodistributed as a normal distribution of some mean and variancerdquo i = 123 denotes type of measurement heterogeneity certification and stability j = 1hellip10 denotes assay 2PR4 POTP NEIF R4Q5 D5 ND6 D9 HBB1 ND14 22C3 k = 1 2 denotes pipette set 120574120574119894119894119894119894119894119894 = 120572120572119894119894119894119894 + 120573120573119894119894 120572120572119894119894119894119894~119873119873(micro1198941198941205901205901205721205722) 120573120573119894119894 = 120573120573 for data type 1 and 120573120573119894119894 = 0 for data types 2 and 3 For a given 2372a component the variance 1205901205901205721205722 measures the between assay variability and 1205901205901198941198942 measures the repeatability variability for each data type The data was analyzed using Bayesian MCMC programmed in OpenBUGSC2 with non-informative Gaussian priors for the means and Gamma priors for the variance components

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

58

33 OpenBUGS code inits and data Certification code ModelBegin for(i in 13)mu[i]~dnorm(010E-5) sig[i]~dgamma(10E-510E-5) sigalph~dgamma(10E-510E-5) for(i in 110)alpha[1i]~dnorm(mu[1] sigalph) for(i in 12)alpha[2i]~dnorm(mu[2] sigalph) alpha[3i]~dnorm(mu[3]sigalph) b~dnorm(010E-5) beta[1]lt-b beta[2]lt-00 for(i in 1Ntot)mean[i]lt-alpha[typ[i]asy[i]]+beta[pip[i]] for(i in 1Ntot)lamb[i]~dnorm(mean[i]sig[typ[i]]) ModelEnd Certification Inits list(sig=c(111) sigalph=1) Certification Data for Component A list(Ntot=275) lamb[] asy[] cmp[] typ[] pip[] 1667 1 1 1 1 1655 2 1 1 1 1647 3 1 1 1 1594 4 1 1 1 1655 5 1 1 1 1629 6 1 1 1 1655 7 1 1 1 1644 8 1 1 1 1645 9 1 1 1 1545 10 1 1 1 1698 1 1 1 1 1685 2 1 1 1 1665 3 1 1 1 1658 4 1 1 1 1683 5 1 1 1 1647 6 1 1 1 1652 7 1 1 1 1611 8 1 1 1 1707 9 1 1 1 1603 10 1 1 1 1668 1 1 1 1 1587 2 1 1 1 1678 3 1 1 1 1628 4 1 1 1 1738 5 1 1 1 1667 6 1 1 1 1695 7 1 1 1 1757 8 1 1 1 1771 9 1 1 1 1762 10 1 1 1 1739 1 1 1 1 1702 2 1 1 1 1614 3 1 1 1 1762 4 1 1 1 1649 5 1 1 1 1678 6 1 1 1 1653 7 1 1 1

175 8 1 1 1 1764 9 1 1 1 1759 10 1 1 1 1638 1 1 1 1 1676 2 1 1 1 1616 3 1 1 1 1559 4 1 1 1 1608 5 1 1 1 1611 6 1 1 1 1517 7 1 1 1 1557 8 1 1 1 1604 9 1 1 1 1675 10 1 1 1 1667 1 1 1 1 1638 2 1 1 1 1624 3 1 1 1 1587 4 1 1 1 1652 5 1 1 1 1589 6 1 1 1 1607 7 1 1 1 1582 8 1 1 1 1625 9 1 1 1 1616 10 1 1 1 1888 1 1 1 1 1868 2 1 1 1 1953 3 1 1 1 1779 4 1 1 1 1924 5 1 1 1 1861 6 1 1 1 1824 7 1 1 1 1809 8 1 1 1 1864 9 1 1 1 193 10 1 1 1 1939 1 1 1 1 1815 2 1 1 1 1902 3 1 1 1 1871 4 1 1 1 2124 5 1 1 1

1787 6 1 1 1 1844 7 1 1 1 1799 8 1 1 1 1857 9 1 1 1 1942 10 1 1 1 1597 1 1 1 1 1533 2 1 1 1 1462 3 1 1 1 1493 4 1 1 1 1538 5 1 1 1 1492 6 1 1 1 1528 7 1 1 1 1548 8 1 1 1 1513 9 1 1 1 1562 10 1 1 1 159 1 1 1 1 1404 2 1 1 1 1533 3 1 1 1 1519 4 1 1 1 1586 5 1 1 1 1516 6 1 1 1 1538 7 1 1 1 1581 8 1 1 1 1514 9 1 1 1 1431 1 1 1 2 1407 2 1 1 2 1487 3 1 1 2 145 4 1 1 2 1394 5 1 1 2 1396 6 1 1 2 1414 7 1 1 2 1401 8 1 1 2 1412 9 1 1 2 1435 10 1 1 2 1411 1 1 1 2 1395 2 1 1 2 1398 3 1 1 2 1439 4 1 1 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

59

1422 5 1 1 2 1393 6 1 1 2 139 7 1 1 2 1382 8 1 1 2 146 9 1 1 2 141 10 1 1 2 1632 1 1 2 2 1531 1 1 2 2 1524 1 1 2 2 1454 1 1 2 2 1465 1 1 2 2 1512 1 1 2 2 1418 1 1 2 2 1496 1 1 2 2 1588 1 1 2 2 1574 1 1 2 2 1557 1 1 2 2 15 1 1 2 2 1672 1 1 2 2 1551 1 1 2 2 1573 1 1 2 2 1516 1 1 2 2 1612 1 1 2 2 1598 1 1 2 2 1575 1 1 2 2 1658 1 1 2 2 1547 1 1 2 2 1529 1 1 2 2 1525 1 1 2 2 154 1 1 2 2 1458 1 1 2 2 154 1 1 2 2 1491 1 1 2 2 148 1 1 2 2 156 1 1 2 2 1581 1 1 2 2 1513 1 1 2 2 1508 1 1 2 2 1476 1 1 2 2 1634 1 1 2 2 1568 1 1 2 2 152 1 1 2 2 1516 1 1 2 2 1535 1 1 2 2 1514 1 1 2 2 1568 1 1 2 2 1611 1 1 2 2 1528 1 1 2 2 1511 1 1 2 2 1549 1 1 2 2 1533 1 1 2 2 1427 1 1 2 2 1504 1 1 2 2 1491 1 1 2 2 1455 1 1 2 2

1501 1 1 2 2 1637 1 1 2 2 157 1 1 2 2 1633 1 1 2 2 1515 1 1 2 2 1804 1 1 2 2 1538 1 1 2 2 1617 1 1 2 2 1628 1 1 2 2 1672 1 1 2 2 1596 1 1 2 2 1586 1 1 2 2 1578 1 1 2 2 1532 1 1 2 2 1616 2 1 2 2 1615 2 1 2 2 1582 2 1 2 2 1627 2 1 2 2 1607 2 1 2 2 1511 2 1 2 2 1479 2 1 2 2 1514 2 1 2 2 1474 2 1 2 2 1594 2 1 2 2 1571 2 1 2 2 1626 2 1 2 2 149 2 1 2 2 1617 2 1 2 2 1634 2 1 2 2 1613 2 1 2 2 1529 2 1 2 2 1571 2 1 2 2 1652 2 1 2 2 1647 2 1 2 2 1537 2 1 2 2 164 2 1 2 2 1589 2 1 2 2 1152 2 1 2 2 1605 2 1 2 2 1633 2 1 2 2 1483 2 1 2 2 1557 2 1 2 2 1615 2 1 2 2 1438 2 1 2 2 1713 2 1 2 2 1611 2 1 2 2 1586 2 1 2 2 1516 2 1 2 2 1614 2 1 2 2 1627 2 1 2 2 1567 2 1 2 2 1495 2 1 2 2 1535 2 1 2 2 1545 2 1 2 2 1607 2 1 2 2

1513 2 1 2 2 1582 2 1 2 2 1601 2 1 2 2 1555 2 1 2 2 1607 2 1 2 2 1586 2 1 2 2 1524 2 1 2 2 1656 2 1 2 2 154 2 1 2 2 1454 2 1 2 2 1669 2 1 2 2 1631 2 1 2 2 1608 2 1 2 2 1497 2 1 2 2 1639 2 1 2 2 1575 2 1 2 2 1567 2 1 2 2 1528 2 1 2 2 1511 2 1 2 2 1627 2 1 2 2 1612 2 1 2 2 1536 2 1 2 2 1568 1 1 3 2 152 1 1 3 2 1486 1 1 3 2 1441 1 1 3 2 1706 1 1 3 2 144 1 1 3 2 15 1 1 3 2 1537 1 1 3 2 1387 1 1 3 2 1463 1 1 3 2 157 1 1 3 2 1429 1 1 3 2 1667 1 1 3 2 1632 1 1 3 2 1589 1 1 3 2 1474 2 1 3 2 1543 2 1 3 2 1555 2 1 3 2 1599 2 1 3 2 1494 2 1 3 2 1624 2 1 3 2 1577 2 1 3 2 1593 2 1 3 2 1479 2 1 3 2 1517 2 1 3 2 1538 2 1 3 2 159 2 1 3 2 1479 2 1 3 2 1479 2 1 3 2 1522 2 1 3 2 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

60

Table C-2 Example OpenBUGS Certification Summary for Component A

mean sd MC_error val25pc median val975pc start sample b 2584 02948 0005376 2001 2588 3158 100001 100000 mu[1] 1417 02706 0005366 1364 1416 147 100001 100000 mu[2] 1559 008368 4776E-4 1543 1559 1576 100001 100000 mu[3] 1533 01504 0001419 1504 1533 1563 100001 100000

Certification Data for Component B list(Ntot=278) lamb[] asy[] cmp[] typ[] pip[] 1869 1 2 1 1 185 2 2 1 1 1709 3 2 1 1 1856 4 2 1 1 1848 5 2 1 1 1828 6 2 1 1 1861 7 2 1 1 1825 8 2 1 1 1871 9 2 1 1 1776 10 2 1 1 196 1 2 1 1 1781 2 2 1 1 1707 3 2 1 1 1787 4 2 1 1 1829 5 2 1 1 1826 6 2 1 1 1858 7 2 1 1 1869 8 2 1 1 1773 9 2 1 1 1902 10 2 1 1 2174 1 2 1 1 2202 2 2 1 1 2172 3 2 1 1 223 5 2 1 1 2094 6 2 1 1 1949 7 2 1 1 1971 8 2 1 1 2123 9 2 1 1 2107 10 2 1 1 2231 1 2 1 1 2292 2 2 1 1 2216 3 2 1 1 2078 4 2 1 1 2192 5 2 1 1 2034 6 2 1 1 211 7 2 1 1 2123 8 2 1 1 2066 9 2 1 1 2046 10 2 1 1 181 1 2 1 1 1865 2 2 1 1 1806 3 2 1 1 1878 4 2 1 1 1824 5 2 1 1 1768 6 2 1 1 1811 7 2 1 1 1763 8 2 1 1 1865 9 2 1 1 1815 10 2 1 1

1806 1 2 1 1 1801 2 2 1 1 1786 3 2 1 1 1768 4 2 1 1 179 5 2 1 1 167 6 2 1 1 1749 7 2 1 1 1741 8 2 1 1 1781 9 2 1 1 1866 10 2 1 1 1784 1 2 1 1 1735 2 2 1 1 1778 3 2 1 1 1835 4 2 1 1 1794 5 2 1 1 17 6 2 1 1 1799 7 2 1 1 1765 8 2 1 1 1754 9 2 1 1 1792 10 2 1 1 1821 1 2 1 1 1813 2 2 1 1 1813 3 2 1 1 1749 4 2 1 1 1848 5 2 1 1 1711 6 2 1 1 1779 7 2 1 1 1769 8 2 1 1 1725 9 2 1 1 1814 10 2 1 1 1718 1 2 1 1 1753 2 2 1 1 1773 3 2 1 1 1777 4 2 1 1 1803 5 2 1 1 1718 6 2 1 1 1802 7 2 1 1 1834 8 2 1 1 1731 9 2 1 1 1842 10 2 1 1 1768 1 2 1 1 1698 2 2 1 1 1792 3 2 1 1 174 4 2 1 1 1795 5 2 1 1 1693 6 2 1 1 1743 7 2 1 1 1745 8 2 1 1 1714 9 2 1 1 1791 1 2 1 2

1746 2 2 1 2 1803 3 2 1 2 1767 4 2 1 2 1745 5 2 1 2 1824 6 2 1 2 1752 7 2 1 2 1813 8 2 1 2 1648 9 2 1 2 1818 10 2 1 2 177 1 2 1 2 182 2 2 1 2 1803 3 2 1 2 1739 4 2 1 2 1773 5 2 1 2 1729 6 2 1 2 1744 7 2 1 2 1754 8 2 1 2 1671 9 2 1 2 1741 10 2 1 2 179 1 2 2 2 1712 1 2 2 2 1693 1 2 2 2 1754 1 2 2 2 1757 1 2 2 2 1697 1 2 2 2 1722 1 2 2 2 1884 1 2 2 2 1713 1 2 2 2 1759 1 2 2 2 1749 1 2 2 2 1748 1 2 2 2 1632 1 2 2 2 1752 1 2 2 2 1871 1 2 2 2 1688 1 2 2 2 1789 1 2 2 2 1826 1 2 2 2 1751 1 2 2 2 1738 1 2 2 2 1742 1 2 2 2 1761 1 2 2 2 1909 1 2 2 2 1719 1 2 2 2 1755 1 2 2 2 1745 1 2 2 2 1687 1 2 2 2 1709 1 2 2 2 1625 1 2 2 2 1802 1 2 2 2 1734 1 2 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

61

1809 1 2 2 2 1782 1 2 2 2 1712 1 2 2 2 1668 1 2 2 2 1535 1 2 2 2 1835 1 2 2 2 1725 1 2 2 2 1769 1 2 2 2 178 1 2 2 2 1759 1 2 2 2 1678 1 2 2 2 1836 1 2 2 2 1743 1 2 2 2 1885 1 2 2 2 1704 1 2 2 2 1749 1 2 2 2 1739 1 2 2 2 1735 1 2 2 2 1714 1 2 2 2 1703 1 2 2 2 1772 1 2 2 2 1727 1 2 2 2 1738 1 2 2 2 174 1 2 2 2 1718 1 2 2 2 1701 1 2 2 2 1673 1 2 2 2 1887 1 2 2 2 1743 1 2 2 2 1825 1 2 2 2 1849 1 2 2 2 188 1 2 2 2 1768 1 2 2 2 1713 1 2 2 2 166 1 2 2 2 1775 2 2 2 2 1824 2 2 2 2 167 2 2 2 2 1683 2 2 2 2 1693 2 2 2 2 169 2 2 2 2 1864 2 2 2 2 1709 2 2 2 2 1764 2 2 2 2

1741 2 2 2 2 1797 2 2 2 2 1626 2 2 2 2 1724 2 2 2 2 177 2 2 2 2 1744 2 2 2 2 1716 2 2 2 2 1805 2 2 2 2 1827 2 2 2 2 1752 2 2 2 2 1746 2 2 2 2 1666 2 2 2 2 187 2 2 2 2 1718 2 2 2 2 1483 2 2 2 2 1632 2 2 2 2 1759 2 2 2 2 1723 2 2 2 2 1684 2 2 2 2 1887 2 2 2 2 1756 2 2 2 2 1816 2 2 2 2 1792 2 2 2 2 1614 2 2 2 2 1566 2 2 2 2 1741 2 2 2 2 1752 2 2 2 2 1697 2 2 2 2 1762 2 2 2 2 1801 2 2 2 2 1766 2 2 2 2 1747 2 2 2 2 1744 2 2 2 2 1692 2 2 2 2 186 2 2 2 2 1756 2 2 2 2 1896 2 2 2 2 1814 2 2 2 2 1703 2 2 2 2 1722 2 2 2 2 1688 2 2 2 2 1821 2 2 2 2 1758 2 2 2 2 1787 2 2 2 2

1784 2 2 2 2 179 2 2 2 2 1681 2 2 2 2 1685 2 2 2 2 1798 2 2 2 2 171 2 2 2 2 1638 2 2 2 2 1931 2 2 2 2 1786 2 2 2 2 1768 2 2 2 2 1764 2 2 2 2 1819 2 2 2 2 1666 1 2 3 2 1626 1 2 3 2 188 1 2 3 2 1653 1 2 3 2 1733 1 2 3 2 1551 1 2 3 2 1884 1 2 3 2 1718 1 2 3 2 1549 1 2 3 2 1723 1 2 3 2 194 1 2 3 2 164 1 2 3 2 1565 1 2 3 2 1787 1 2 3 2 1655 2 2 3 2 167 2 2 3 2 1673 2 2 3 2 1705 2 2 3 2 1755 2 2 3 2 1625 2 2 3 2 1708 2 2 3 2 1705 2 2 3 2 1862 2 2 3 2 1686 2 2 3 2 1723 2 2 3 2 1909 2 2 3 2 1859 2 2 3 2 1897 2 2 3 2 1972 2 2 3 2 END

Table C-3 Example OpenBUGS Certification Summary for Component B

mean sd MC_error val25pc median val975pc start sample b 09337 03387 0007504 02695 09371 1592 100001 100000 mu[1] 1764 03101 0007483 1704 1763 1825 100001 100000 mu[2] 1748 008014 4719E-4 1733 1748 1764 100001 100000 mu[3] 1735 02338 0003093 1688 1736 1781 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

62

Certification data for Component C list(Ntot=280) lamb[] asy[] cmp[] typ[] pip[] 163 1 3 1 1 1596 2 3 1 1 1574 3 3 1 1 159 4 3 1 1 161 5 3 1 1 1589 6 3 1 1 1569 7 3 1 1 1565 8 3 1 1 1564 9 3 1 1 1548 10 3 1 1 1706 1 3 1 1 1576 2 3 1 1 1538 3 3 1 1 1555 4 3 1 1 1607 5 3 1 1 1548 6 3 1 1 1572 7 3 1 1 1681 8 3 1 1 1492 9 3 1 1 1655 10 3 1 1 1499 1 3 1 1 1576 2 3 1 1 1545 3 3 1 1 1584 4 3 1 1 1577 5 3 1 1 1599 6 3 1 1 1587 7 3 1 1 152 8 3 1 1 1479 9 3 1 1 1512 10 3 1 1 1497 1 3 1 1 1611 2 3 1 1 156 3 3 1 1 1546 4 3 1 1 1529 5 3 1 1 1478 6 3 1 1 1504 7 3 1 1 1551 8 3 1 1 1438 9 3 1 1 1494 10 3 1 1 1452 1 3 1 1 1347 2 3 1 1 1395 3 3 1 1 1442 4 3 1 1 1448 5 3 1 1 1463 6 3 1 1 1402 7 3 1 1 1474 8 3 1 1 1466 9 3 1 1 1439 10 3 1 1 1424 1 3 1 1 1397 2 3 1 1 135 3 3 1 1 1459 4 3 1 1 1602 5 3 1 1 1435 6 3 1 1 1433 7 3 1 1 1481 8 3 1 1 1482 9 3 1 1

1568 10 3 1 1 147 1 3 1 1 1411 2 3 1 1 1512 3 3 1 1 1452 4 3 1 1 1567 5 3 1 1 148 6 3 1 1 1442 7 3 1 1 143 8 3 1 1 1413 9 3 1 1 1486 10 3 1 1 1444 1 3 1 1 1427 2 3 1 1 1439 3 3 1 1 1468 4 3 1 1 1524 5 3 1 1 1424 6 3 1 1 1459 7 3 1 1 1467 8 3 1 1 1411 9 3 1 1 148 10 3 1 1 1485 1 3 1 1 1422 2 3 1 1 1438 3 3 1 1 1439 4 3 1 1 1515 5 3 1 1 1409 6 3 1 1 1469 7 3 1 1 1453 8 3 1 1 1465 9 3 1 1 1363 10 3 1 1 1458 1 3 1 1 1459 2 3 1 1 1465 3 3 1 1 1468 4 3 1 1 1457 5 3 1 1 1412 6 3 1 1 1474 7 3 1 1 1428 8 3 1 1 1425 9 3 1 1 1393 10 3 1 1 1357 1 3 1 2 1447 2 3 1 2 1448 3 3 1 2 1409 4 3 1 2 1389 5 3 1 2 1415 6 3 1 2 1431 7 3 1 2 1381 8 3 1 2 1403 9 3 1 2 1435 10 3 1 2 1364 1 3 1 2 1361 2 3 1 2 1404 3 3 1 2 1411 4 3 1 2 1429 5 3 1 2 1417 6 3 1 2 138 7 3 1 2 1336 8 3 1 2 1387 9 3 1 2

1372 10 3 1 2 1493 1 3 2 2 1516 1 3 2 2 1451 1 3 2 2 144 1 3 2 2 1523 1 3 2 2 1437 1 3 2 2 1485 1 3 2 2 1433 1 3 2 2 1463 1 3 2 2 1534 1 3 2 2 1533 1 3 2 2 145 1 3 2 2 1537 1 3 2 2 1515 1 3 2 2 1511 1 3 2 2 1449 1 3 2 2 1388 1 3 2 2 1475 1 3 2 2 1468 1 3 2 2 1549 1 3 2 2 153 1 3 2 2 1442 1 3 2 2 1385 1 3 2 2 1513 1 3 2 2 1528 1 3 2 2 1486 1 3 2 2 1472 1 3 2 2 1493 1 3 2 2 1477 1 3 2 2 1405 1 3 2 2 1499 1 3 2 2 1458 1 3 2 2 1514 1 3 2 2 1525 1 3 2 2 1527 1 3 2 2 149 1 3 2 2 1568 1 3 2 2 1452 1 3 2 2 1421 1 3 2 2 1455 1 3 2 2 1525 1 3 2 2 1531 1 3 2 2 144 1 3 2 2 1487 1 3 2 2 151 1 3 2 2 1577 1 3 2 2 1506 1 3 2 2 152 1 3 2 2 1483 1 3 2 2 1522 1 3 2 2 1463 1 3 2 2 1478 1 3 2 2 1505 1 3 2 2 1454 1 3 2 2 1555 1 3 2 2 1497 1 3 2 2 1515 1 3 2 2 1494 1 3 2 2 1594 1 3 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

63

1471 1 3 2 2 1394 1 3 2 2 1445 1 3 2 2 1446 1 3 2 2 1499 1 3 2 2 1484 1 3 2 2 1436 1 3 2 2 1565 2 3 2 2 149 2 3 2 2 1485 2 3 2 2 1529 2 3 2 2 1513 2 3 2 2 1492 2 3 2 2 1475 2 3 2 2 1497 2 3 2 2 1503 2 3 2 2 1448 2 3 2 2 1488 2 3 2 2 1496 2 3 2 2 1499 2 3 2 2 1469 2 3 2 2 1515 2 3 2 2 1543 2 3 2 2 1452 2 3 2 2 1458 2 3 2 2 1435 2 3 2 2 1499 2 3 2 2 1499 2 3 2 2 1589 2 3 2 2 1562 2 3 2 2 1547 2 3 2 2 1505 2 3 2 2 1508 2 3 2 2 152 2 3 2 2

1481 2 3 2 2 1547 2 3 2 2 1501 2 3 2 2 1491 2 3 2 2 1465 2 3 2 2 147 2 3 2 2 1483 2 3 2 2 1479 2 3 2 2 148 2 3 2 2 1543 2 3 2 2 1448 2 3 2 2 1531 2 3 2 2 1478 2 3 2 2 1465 2 3 2 2 1454 2 3 2 2 1591 2 3 2 2 1466 2 3 2 2 157 2 3 2 2 1471 2 3 2 2 1506 2 3 2 2 1491 2 3 2 2 1461 2 3 2 2 1485 2 3 2 2 1495 2 3 2 2 1516 2 3 2 2 1474 2 3 2 2 1483 2 3 2 2 1519 2 3 2 2 1475 2 3 2 2 1466 2 3 2 2 1565 2 3 2 2 1491 2 3 2 2 1435 2 3 2 2 1459 2 3 2 2

1421 2 3 2 2 1468 2 3 2 2 1556 2 3 2 2 1424 1 3 3 2 1427 1 3 3 2 1381 1 3 3 2 1394 1 3 3 2 1473 1 3 3 2 133 1 3 3 2 1446 1 3 3 2 1424 1 3 3 2 133 1 3 3 2 1452 1 3 3 2 1512 1 3 3 2 1314 1 3 3 2 1525 1 3 3 2 1475 1 3 3 2 1542 1 3 3 2 1461 2 3 3 2 1443 2 3 3 2 1547 2 3 3 2 1394 2 3 3 2 1364 2 3 3 2 1395 2 3 3 2 1522 2 3 3 2 1483 2 3 3 2 1452 2 3 3 2 1482 2 3 3 2 1439 2 3 3 2 1458 2 3 3 2 155 2 3 3 2 1502 2 3 3 2 152 2 3 3 2 END

Table C-4 Example OpenBUGS Certification Summary for Component C

mean sd MC_error val25pc median val975pc start sample b 09626 01701 0002428 06285 09633 1293 100001 100000 mu[1] 1399 01566 0002413 1369 1399 143 100001 100000 mu[2] 1491 005701 2386E-4 148 1491 1503 100001 100000 mu[3] 1449 0132 0001181 1423 1449 1475 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

64

34 Results Table C-5 lists the results in CopiesnL for the three components

Table C-5 Parameter Values for Components A B and C

Posterior Distribution Summary Statistics Component Parameter Mean SD 25 50 975

A

β 260 029 203 260 318 μ1 1415 027 1362 1415 1467 μ2 1559 010 1539 1559 158 μ3 1533 016 1501 1533 1565

B

β 095 034 029 095 162 μ1 1762 031 1701 1762 1822 μ2 1748 010 1728 1748 1769 μ3 1735 024 1687 1735 1783

C

β 097 017 064 097 130 μ1 1399 016 1368 1399 1429 μ2 1491 008 1476 1491 1507 μ3 1449 014 1421 1449 1477

Since in each case the three data set types represent independent measurements the three estimates can be combined to obtain a consensus value In this way a between data set type variability is included in the final standard uncertainty Table C-6 lists the results in Copies per Nanoliter obtained using the NIST Consensus Builder (NICOB)C3

Table C-6 Certified Values for Components A B and C

Component Mean u(Mean) Uk=2(Mean) A 151 07 14 B 175 03 06 C 145 05 10

4 Mito-to-nuclear DNA ratios (mtDNAnDNA)

41 Data Measurements were also obtained on the same samples using three mitochondrial detection assays mtND1 mtBatz and mtKav and a single nDNA assay (genomic) to compute the ratios mtDNAnDNA For the complete mtDNAnDNA data see

Table 16 mtDNAnDNA Ratios

42 Model For each component the data was analyzed using the following random effects model to account for between assay variability

119903119903119894119894~119873119873120572120572119894119894 1205901205901198861198861198861198861198861198861198861198861198861198862

where 119894119894 denotes assay (for A and B 1 = mtND1 2 = mtBatz 3 = mtKav for C 1 = mtBatz 2 = mtKav) and 120572120572119894119894~119873119873micro120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119886119886119886119886119886119886119886119886119886119886

2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

65

43 OpenBUGS code inits and data Two versions of the same model are required one for the results from the three mtDNA assays available for components A and B mtND1 mtBatz and mtKav and the other for the results from the two mtDNA assays available for component C mtBatz and mtKav mtDNAnDNA code for components A and B ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 13)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for components A and B list(sigt=c(111)) Data for Component A list(Ndata=102) set[] rep[] asy[] rat[] 1 1 1 1931 1 2 1 187 1 1 2 1845 1 2 2 1764 2 1 1 181 2 2 1 1906 2 1 2 1785 2 2 2 1837 3 1 1 1743 3 2 1 1761 3 1 2 1829 3 2 2 1847 4 1 1 1716 4 2 1 180 4 1 2 174 4 2 2 1813 5 1 1 1923 5 2 1 1835 5 1 2 1935 5 2 2 1851 6 1 1 1757 6 2 1 1724 6 3 1 1768 6 1 2 1743 6 2 2 1671 6 3 2 1766 6 1 3 1692 6 2 3 1662 6 3 3 1767 7 1 1 1924 7 2 1 1944 7 3 1 1893 7 1 2 1786

7 2 2 1752 7 3 2 1743 7 1 3 153 7 2 3 1542 7 3 3 1523 7 1 2 158 7 2 2 1557 7 3 2 1592 7 1 3 2103 7 2 3 210 7 3 3 2049 7 1 1 1749 7 2 1 1769 7 3 1 1777 7 1 2 1965 7 2 2 1889 7 3 2 1947 7 1 3 1813 7 2 3 1864 7 3 3 1864 8 1 1 178 8 2 1 1776 8 3 1 1706 8 1 2 1804 8 2 2 1751 8 3 2 1738 8 1 3 1769 8 2 3 1786 8 3 3 1761 9 1 2 1584 9 2 2 1544 10 1 1 1947 10 2 1 1928 10 3 1 1922 10 1 1 1829

10 2 1 1895 10 3 1 1851 10 1 1 1674 10 2 1 1617 11 1 2 1667 11 2 2 1813 11 3 2 1708 11 4 2 1598 11 1 2 1566 11 2 2 1687 11 3 2 1663 11 4 2 1589 11 1 2 1643 11 2 2 1668 11 3 2 1604 11 4 2 1573 12 1 2 1719 12 2 2 1716 12 3 2 167 12 4 2 1666 12 1 2 1581 12 2 2 1573 12 3 2 1601 12 4 2 1612 12 1 2 1546 12 2 2 1542 12 3 2 1538 12 4 2 1587 13 1 1 1617 13 2 1 1599 13 1 1 1622 13 2 1 1579 13 1 1 1575 13 2 1 1528 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

66

Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A

mean sd MC_error val25pc median val975pc start sample mu 1730 1443 004419 1703 1730 1759 100001 100000

data for Component B list(Ndata=87) set[] rep[] asy[] rat[] 1 1 1 2461 1 2 1 2401 1 1 2 2555 1 2 2 2608 2 1 1 2229 2 2 1 2200 2 1 2 2108 2 2 2 2192 3 1 1 2127 3 2 1 2198 3 1 2 2455 3 2 2 2506 4 1 1 2211 4 2 1 2199 4 1 2 2380 4 2 2 2249 5 1 1 2105 5 2 1 1943 5 1 2 2136 5 2 2 2119 6 1 1 1990 6 2 1 1921 6 3 1 1982 6 1 2 1990 6 2 2 1924 6 3 2 1958 6 1 3 1903 6 2 3 1831

6 3 3 1919 7 1 1 2304 7 2 1 2341 7 3 1 2332 7 1 2 2048 7 2 2 2078 7 3 2 2067 7 1 3 2399 7 2 3 2479 7 3 3 2493 7 1 1 2065 7 2 1 2166 7 3 1 2164 7 1 2 2314 7 2 2 2280 7 3 2 2258 7 1 3 1867 7 2 3 1987 7 3 3 1971 8 1 1 2046 8 2 1 2057 8 3 1 2024 8 1 2 1998 8 2 2 2043 8 3 2 1952 8 1 3 2077 8 2 3 2117 8 3 3 2041 9 1 2 1890 9 2 2 1886

10 1 2 1928 10 2 2 1916 10 3 2 1915 10 1 2 1947 10 2 2 1946 10 3 2 1961 10 4 2 1909 10 1 2 1953 10 2 2 1952 10 3 2 1941 10 4 2 1881 11 1 2 1865 11 2 2 1863 11 3 2 1828 11 4 2 1828 11 1 2 1944 11 2 2 1964 11 3 2 1947 11 4 2 1895 11 1 2 1908 11 2 2 1909 11 3 2 1940 11 4 2 1839 12 1 1 1688 12 2 1 1783 12 1 1 1823 12 2 1 1631 12 1 1 1920 12 2 1 1887 END

Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B

mean sd MC_error val25pc median val975pc start sample mu 2045 2246 0111 1999 2045 2088 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

67

mtDNAnDNA code for component C ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 12)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for component C list(sigt=c(11)) mtDNAnDNA data for component C list(Ndata=66) set[] rep[] asy[] rat[] 1 1 1 3201 1 2 1 3299 2 1 1 3347 2 2 1 3287 3 1 1 3217 3 2 1 3384 4 1 1 3200 4 2 1 3363 5 1 1 3294 5 2 1 3013 6 1 1 2805 6 2 1 2825 6 3 1 2829 6 1 2 2709 6 2 2 2738 6 3 2 2727 7 1 1 2797 7 2 1 2642 7 3 1 2792 7 1 2 2804 7 2 2 2789

7 3 2 2842 7 1 1 2715 7 2 1 2606 7 3 1 2687 7 1 2 2799 7 2 2 2628 7 3 2 2559 7 1 1 2639 7 2 1 2535 7 3 1 2657 7 1 2 2953 7 2 2 2757 7 3 2 2802 8 1 1 2927 8 2 1 2776 8 3 1 2762 8 1 2 3044 8 2 2 2830 8 3 2 2825 9 1 1 2501 9 2 1 2522 10 1 1 2832 10 2 1 2840

10 3 1 2788 10 4 1 2725 10 1 1 2835 10 2 1 2790 10 3 1 2736 10 4 1 2679 10 1 1 2803 10 2 1 2773 10 3 1 2757 10 4 1 2752 11 1 1 2599 11 2 1 2508 11 3 1 2551 11 4 1 2609 11 1 1 2617 11 2 1 2483 11 3 1 2626 11 4 1 2602 11 1 1 2665 11 2 1 2535 11 3 1 2570 11 4 1 2650 END

Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C

mean sd MC_error val25pc median val975pc start sample mu 2774 2367 01176 2729 2774 2821 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

68

Table C-10 Results for mtDNAnDNA Posterior Distribution Summary Statistics

Component Mean Std Dev 25 50 975 A 1730 14 1703 1730 1758 B 2045 23 1994 2047 2087 C 2774 23 2729 2774 2821

5 ReferencesC1 Lunn DJ Spiegelhalter D Thomas A Best N Statistics in Medicine 2009283049ndash3082C2 Polson NG Scott JG Bayesian Analysis 20127887ndash902C3 NIST Consensus Builder Software and userrsquos guide freely available at

httpsconsensusnistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

69

Appendix D Performance in Commercial qPCR Chemistries The performance of the SRM 2372a materials in commercial qPCR chemistries was assessed using manufacturerrsquos recommended protocols Each panel in Figures D-1 to D-4 displays the average cycle threshold (Ct) as a function of DNA concentration for the 5-point dilutions of SRM 2372a components In all cases the external calibrant was SRM 2372 component A Error bars represent approximate 95 confidence intervals Table D-1 lists the estimated [DNA] for the three SRM 2372a components for all commercial assays evaluated This Table also lists the Component BComponent A (BA) and Component CComponent A (CA) ratios calculated from the [DNA] estimates

Figure D-1 Results for the Innogenomics Commercial qPCR assays

20

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantLong(IQ_h)

ABCStd

[A] = (384 plusmn 26) ngμL[B] = (441 plusmn 24) ngμL[C] = (516 plusmn 24) ngμL

Rsup2 = 0999 se(Ct) = 01016

17

18

19

20

21

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantShort(IQ_d)

ABCStd

[A] = (435 plusmn 28) ngμL[B] = (543 plusmn 12) ngμL[C] = (510 plusmn 28) ngμL

Rsup2 = 0999 se(Ct) = 007

24

25

26

27

28

29

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYLong(IQHY_h)

ABCStd

[A] = (387 plusmn 24) ngμL[B] = (448 plusmn 21) ngμL[C] = (559 plusmn 31) ngμL

Rsup2 = 0998 se(Ct) = 01314

16

18

20

22

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYShort(IQHY_d)

ABCStd

[A] = (353 plusmn 23) ngμL[B] = (267 plusmn 15) ngμL[C] = (390 plusmn 18) ngμL

Rsup2 = 1000 se(Ct) = 00420

22

24

26

28

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYY(IQHY_y)

ABCStd

[A] = (572 plusmn 13) ngμL[B] = 0 ngμL[C] = (137 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

70

Figure D-2 Results for the Thermo Fisher Commercial qPCR assays

21

22

23

24

25

26

27

28

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HumanHuman(QFH_h)

ABCStd

[A] = (464 plusmn 13) ngμL[B] = (529 plusmn 19) ngμL[C] = (472 plusmn 22) ngμL

Rsup2 = 1000 se(Ct) = 004

22

23

24

25

26

27

28

29

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoHuman(QFD_h)

ABCStd

[A] = (701 plusmn 53) ngμL[B] = (561 plusmn 11) ngμL[C] = (472 plusmn 11) ngμL

Rsup2 = 0994 se(Ct) = 01922

24

26

28

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoMale(QFD_y)

ABCStd

[A] = (706 plusmn 49) ngμL[B] = 0 ngμL[C] = (122 plusmn 02) ngμL

Rsup2 = 0999 se(Ct) = 007

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPLarge(QFHP_h)

ABCStd

[A] = (340 plusmn 18) ngμL[B] = (432 plusmn 05) ngμL[C] = (342 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 00720

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPSmall(QFHP_d)

ABCStd

[A] = (434 plusmn 15) ngμL[B] = (514 plusmn 10) ngμL[C] = (466 plusmn 07) ngμL

Rsup2 = 1000 se(Ct) = 005

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioLarge(QFT_h)

ABCStd

[A] = (362 plusmn 23) ngμL[B] = (430 plusmn 06) ngμL[C] = (378 plusmn 07) ngμL

Rsup2 = 0999 se(Ct) = 00920

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioSmall(QFT_d)

ABCStd

[A] = (468 plusmn 08) ngμL[B] = (505 plusmn 08) ngμL[C] = (523 plusmn 14) ngμL

Rsup2 = 1000 se(Ct) = 00318

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioY(QFT_y)

ABCStd

[A] = (372 plusmn 11) ngμL[B] = 0 ngμL[C] = (110 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 003

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

71

Figure D-3 Results for the Qiagen Commercial qPCR assays

Figure D-4 Results for the Promega Commercial qPCR assays

19

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresAutosomal(QPH_h)

ABCStd

[A] = (554 plusmn 21) ngμL[B] = (479 plusmn 05) ngμL[C] = (381 plusmn 09) ngμL

Rsup2 = 1000 se(Ct) = 00418

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresY(QPH_y)

ABCStd

[A] = (358 plusmn 10) ngμL[B] = 0 ngμL[C] = (118 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 007

16

17

18

19

20

21

22

23

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProHuman(QPP_h)

ABCStd

[A] = (525 plusmn 49) ngμL[B] = (480 plusmn 42) ngμL[C] = (467 plusmn 45) ngμL

Rsup2 = 0999 se(Ct) = 01019

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProDegradation(QPP_d)

ABCStd

[A] = (567 plusmn 18) ngμL[B] = (558 plusmn 08) ngμL[C] = (484 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 00410

15

20

25

30

35

-50 -40 -30 -20 -10 00 10

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProY(QPP_y)

ABCStd

[A] = (538 plusmn 27) ngμL[B] = (00 plusmn 00) ngμL[C] = (162 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 003

16

18

20

22

24

-05 00 05 10 15 20

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorAutosomal(P_h)

ABCStd

[A] = (759 plusmn 64) ngμL[B] = (775 plusmn 24) ngμL[C] = (1056 plusmn 17) ngμL

Rsup2 = 0995 se(Ct) = 01818

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorY(P_y)

ABCStd

[A] = (596 plusmn 51) ngμL[B] = 0 ngμL[C] = (117 plusmn 01) ngμL

Rsup2 = 0994 se(Ct) = 020

18

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantAutosomal(PQ_h)

ABCStd

[A] = (436 plusmn 13) ngμL[B] = (606 plusmn 20) ngμL[C] = (504 plusmn 10) ngμL

Rsup2 = 1000 se(Ct) = 00218

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantDegradation(PQ_d)

ABCStd

[A] = (420 plusmn 18) ngμL[B] = (529 plusmn 16) ngμL[C] = (530 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 00516

18

20

22

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantY(PQ_y)

ABCStd

[A] = (410 plusmn 20) ngμL[B] = 0 ngμL[C] = (139 plusmn 05) ngμL

Rsup2 = 0999 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

72

Table D-1 Summary [DNA] Results for the Commercial qPCR Assays Estimated [DNA] ngμL Ratios

A B C BA CA Assaya xb u(x)c xb u(x)c xb u(x)c yd u(y)e yd u(y)e

IQd 435 28 543 12 510 28 125 008 117 010 IQh 384 26 441 24 516 24 115 010 134 011

IQHYd 353 23 267 15 390 18 076 006 111 009 IQHYh 387 24 448 21 559 31 116 009 144 012

Ph 759 64 775 24 1056 16 102 009 139 012 PQd 420 18 529 16 530 06 126 007 126 006 PQh 436 13 606 20 504 10 139 006 116 004

QFDh 701 53 561 11 472 11 080 006 067 005 QFHh 464 13 529 19 472 22 114 005 102 006

QFHPd 434 15 514 10 466 07 118 005 107 004 QFHPh 340 18 432 05 342 03 127 007 101 005

QFTd 468 08 505 08 523 14 108 003 112 004 QFTh 362 23 430 06 378 07 118 008 104 007 QPHh 554 21 479 05 380 09 086 003 069 003 QPPd 567 18 558 08 484 11 098 003 085 003 QPPh 525 49 480 42 467 45 092 012 089 012

IQHYy 572 13 NDf 137 06

NRg

024 001 Py 596 51 NDf 117 01 020 002

PQy 410 20 NDf 139 05 034 002 QFDy 706 49 NDf 122 02 017 001 QFTy 372 11 NDf 110 06 030 002 QPHy 358 10 NDf 118 03 033 001 QPPy 538 27 0002 0001 162 11 030 003

a Assay codes are listed in Table 8 b Mean of dilution-adjusted results from five dilutions of the SRM 2372a component c Standard uncertainty of the mean estimated from the stdanrd deviation of the five dilutions d Ratio of the estimated [DNA] of component B or C relative to that of component A e Standard uncertainty of the ratio estimated from the standard uncertainties of the individual components f Not detected Component B was prepared using tissue from only female donors it is not expected to

respond to male-specific assays g Not relevant

  • Abstract
  • Keywords
  • Technical Information Contact for this SRM
  • Table of Contents
  • Table of Tables
  • Table of Figures
  • Purpose and Description
  • Warning SRM 2372a is a Human Source Material
  • Storage and Use
  • History and Background
    • Figure 1 Sales History of SRM 2372
      • Experimental Methods
        • Polymerase Chain Reaction Assays
          • Table 1 NIST-Developed Human Nuclear DNA Assays
          • Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays
          • Table 2 NIST-Optimized Human Mitochondrial DNA Assays
          • Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays
            • Droplet Digital Polymerase Chain Reaction Measurements
              • Nuclear DNA (nDNA)
                • Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays
                  • Mitochondrial DNA (mtDNA)
                    • Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays
                    • Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays
                        • Sample Preparation
                        • Absorbance Measurements
                        • SRM Unit Production
                        • Component Homogeneity
                          • Figure 4 Plate Layout Design for Homogeneity Measurements
                            • Component Stability
                              • Figure 5 Plate Layout Design for Stability Measurements
                                • Component Certification Measurements
                                  • Table 6 Tubes used for Certification Measurements
                                  • Figure 6 Basic Plate Layout Design for Certification Measurements
                                    • Mitochondrial DNA Measurements
                                      • Table 7 Run Parameters for Mitochondrial DNA
                                        • Performance in Commercial qPCR Chemistries
                                          • Table 8 Commercial qPCR Chemistries Tested with SRM 2372a
                                          • Figure 7 Plate Layout Design for Performance Assessments
                                              • Measurement Results
                                                • Absorbance Spectrophotometry
                                                  • Table 9 Absorbance at Selected Wavelengths
                                                  • Figure 8 Absorbance Spectra of SRM 2372a Components A B and C
                                                    • Droplet Volume
                                                    • Certification Measurements
                                                      • Homogeneity Results
                                                        • Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component
                                                        • Table 11 Within- and Between-Tube Relative Standard Deviations
                                                          • Stability Results
                                                            • Figure 9 ddPCR Stability Measurements as Function of Time
                                                            • Table 12 Stability Data as λacute Copy Number per Nanoliter of Component
                                                              • Ten-Assay Certification Results
                                                                • Figure 10 Comparison of Assay Performance
                                                                • Table 13 Certification Data as λacute Copy Number per Nanoliter of Component
                                                                  • Estimation of Copy Number Concentration
                                                                    • Table 14 Measured Copy Number Concentrations
                                                                      • Conversion of Copy Number to Mass Concentration
                                                                        • Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution
                                                                            • Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)
                                                                              • Figure 11 mtDNAnDNA for Components A B and C
                                                                              • Table 16 mtDNAnDNA Ratios
                                                                              • Table 17 mtDNAnDNA Summary Results
                                                                                • Performance in Commercial qPCR Chemistries
                                                                                  • Figure 12 qPCR Autosomal Quantitation Performance Summary
                                                                                  • Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary
                                                                                      • Qualification of ddPCR as Fit-For-Purpose
                                                                                        • Figure 14 Absorbance Spectra of SRM 2372 Components A and B
                                                                                        • Digital Polymerase Chain Reaction Assays
                                                                                        • Heat Treatments (Melting of the DNA into Single Strands)
                                                                                        • Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a
                                                                                          • Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a
                                                                                            • Quantitative Evaluation of ssDNA Content
                                                                                              • ddPCR Measurement of ssDNA Proportion
                                                                                              • Proof of Principle
                                                                                                • Table 19 Comparison of Snap-Cooling to Slow-Cooling
                                                                                                  • Time at Denaturation Temperature
                                                                                                    • Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures
                                                                                                      • Temperature
                                                                                                        • Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a
                                                                                                          • DNA Concentration in the ddPCR Solution
                                                                                                            • Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution
                                                                                                              • Equivalence of Assays
                                                                                                                • Table 23 Comparison of Assays
                                                                                                                    • ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials
                                                                                                                      • Table 24 Evaluation of Limiting Ratio
                                                                                                                          • Conclusions and Recommendations
                                                                                                                            • Recommended Certification Values
                                                                                                                              • Table 25 Recommended Values and 95 Uncertainties for Certification
                                                                                                                                • Use of SRM 2372a for Value-Assigning In-House Reference Materials
                                                                                                                                  • Certificate of Analysis
                                                                                                                                  • References
                                                                                                                                  • Appendix A DNA Extraction Method
                                                                                                                                    • Procedure
                                                                                                                                    • Reagents
                                                                                                                                    • Reference
                                                                                                                                      • Appendix B Spectrophotometric Measurement Protocol
                                                                                                                                        • Spectrophotometer Calibration
                                                                                                                                          • Establish Baseline
                                                                                                                                          • Enable Verification of Wavelength Calibration
                                                                                                                                          • Enable Verification of Absorbance Calibration
                                                                                                                                            • For ldquonativerdquo dsDNA Samples
                                                                                                                                              • Verify Optical Balance of Cuvettes
                                                                                                                                                • For NaOH denatured ssDNA Samples
                                                                                                                                                  • Prepare a 04 molL NaOH Stock Solution
                                                                                                                                                  • Prepare the Sample Solutions
                                                                                                                                                  • Measure Sample Solutions
                                                                                                                                                    • After All Sample Measurements are Complete
                                                                                                                                                      • Enable Verification of Spectrophotometer Stability
                                                                                                                                                      • Clean Up
                                                                                                                                                        • Evaluation
                                                                                                                                                        • Reference
                                                                                                                                                          • Appendix C SRM 2372a Statisticianrsquos report
                                                                                                                                                            • 1 Introduction
                                                                                                                                                            • 11 OpenBUGS Analysis
                                                                                                                                                            • 12 OpenBUGS Exemplar Output
                                                                                                                                                            • 2 Homogeneity assessment
                                                                                                                                                            • 21 Data
                                                                                                                                                            • 22 Model
                                                                                                                                                            • 23 OpenBUGS code and data
                                                                                                                                                              • Table C-1 Example OpenBUGS Homogeneity Summary
                                                                                                                                                                • 24 Results
                                                                                                                                                                  • Figure C-1 Homogeneity Assessment
                                                                                                                                                                    • 3 Certification measurements
                                                                                                                                                                    • 31 Data
                                                                                                                                                                    • 32 Model
                                                                                                                                                                    • 33 OpenBUGS code inits and data
                                                                                                                                                                      • Table C-2 Example OpenBUGS Certification Summary for Component A
                                                                                                                                                                      • Table C-3 Example OpenBUGS Certification Summary for Component B
                                                                                                                                                                      • Table C-4 Example OpenBUGS Certification Summary for Component C
                                                                                                                                                                        • 34 Results
                                                                                                                                                                          • Table C-5 Parameter Values for Components A B and C
                                                                                                                                                                          • Table C-6 Certified Values for Components A B and C
                                                                                                                                                                            • 4 Mito-to-nuclear DNA ratios (mtDNAnDNA)
                                                                                                                                                                            • 41 Data
                                                                                                                                                                            • 42 Model
                                                                                                                                                                            • 43 OpenBUGS code inits and data
                                                                                                                                                                              • Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A
                                                                                                                                                                              • Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B
                                                                                                                                                                              • Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C
                                                                                                                                                                              • Table C-10 Results for mtDNAnDNA
                                                                                                                                                                                • 5 References
                                                                                                                                                                                  • Appendix D Performance in Commercial qPCR Chemistries
                                                                                                                                                                                    • Table D-1 Summary [DNA] Results for the Commercial qPCR Assays
Page 5: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

ii

Table of Contents Purpose and Description 1 Warning SRM 2372a is a Human Source Material 1 Storage and Use 1 History and Background 2 Experimental Methods 4

Polymerase Chain Reaction Assays 4 Droplet Digital Polymerase Chain Reaction Measurements 7

Nuclear DNA (nDNA) 7 Mitochondrial DNA (mtDNA) 8

Sample Preparation 9 Absorbance Measurements 10 SRM Unit Production 10 Component Homogeneity 10 Component Stability 11 Component Certification Measurements 12 Mitochondrial DNA Measurements 13 Performance in Commercial qPCR Chemistries 13

Measurement Results 15 Absorbance Spectrophotometry 15 Droplet Volume 16 Certification Measurements 16

Homogeneity Results 16 Stability Results 18 Ten-Assay Certification Results 20 Estimation of Copy Number Concentration 22 Conversion of Copy Number to Mass Concentration 22

Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA) 23 Performance in Commercial qPCR Chemistries 25

Qualification of ddPCR as Fit-For-Purpose 27 Digital Polymerase Chain Reaction Assays 28 Heat Treatments (Melting of the DNA into Single Strands) 28 Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a 29 Quantitative Evaluation of ssDNA Content 30

ddPCR Measurement of ssDNA Proportion 30 Proof of Principle 31 Time at Denaturation Temperature 33 Temperature 35 DNA Concentration in the ddPCR Solution 36 Equivalence of Assays 38

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials 40 Conclusions and Recommendations 43

Recommended Certification Values 43 Use of SRM 2372a for Value-Assigning In-House Reference Materials 44

Certificate of Analysis 44 References 44

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

iii

Appendix A DNA Extraction Method 46 Procedure 46 Reagents 46 Reference 47

Appendix B Spectrophotometric Measurement Protocol 48 Spectrophotometer Calibration 48

Establish Baseline 48 Enable Verification of Wavelength Calibration 48 Enable Verification of Absorbance Calibration 48

For ldquonativerdquo dsDNA Samples 48 Verify Optical Balance of Cuvettes 48

For NaOH denatured ssDNA Samples 49 Prepare a 04 molL NaOH Stock Solution 49 Prepare the Sample Solutions 49 Measure Sample Solutions 50

After All Sample Measurements are Complete 50 Enable Verification of Spectrophotometer Stability 50 Clean Up 50

Evaluation 50 Appendix C SRM 2372a Statisticianrsquos report 51

1 Introduction 51 11 OpenBUGS Analysis 51 12 OpenBUGS Exemplar Output 51 2 Homogeneity assessment 52 21 Data 52 22 Model 52 23 OpenBUGS code and data 52 24 Results 55 3 Certification measurements 57 31 Data 57 32 Model 57 33 OpenBUGS code inits and data 58 34 Results 64 4 Mito-to-nuclear DNA ratios (mtDNAnDNA) 64 41 Data 64 42 Model 64 43 OpenBUGS code inits and data 65 5 References 68

Appendix D Performance in Commercial qPCR Chemistries 69

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

iv

Table of Tables Table 1 NIST-Developed Human Nuclear DNA Assays 5 Table 2 NIST-Optimized Human Mitochondrial DNA Assays 6 Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays 8 Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays 8 Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays 8 Table 6 Tubes used for Certification Measurements 12 Table 7 Run Parameters for Mitochondrial DNA 13 Table 8 Commercial qPCR Chemistries Tested with SRM 2372a 14 Table 9 Absorbance at Selected Wavelengths 15 Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component 17 Table 11 Within- and Between-Tube Relative Standard Deviations 18 Table 12 Stability Data as λacute Copy Number per Nanoliter of Component 19 Table 13 Certification Data as λacute Copy Number per Nanoliter of Component 21 Table 14 Measured Copy Number Concentrations 22 Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution 23 Table 16 mtDNAnDNA Ratios 24 Table 17 mtDNAnDNA Summary Results 25 Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a 29 Table 19 Comparison of Snap-Cooling to Slow-Cooling 32 Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures 34 Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 36 Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution 37 Table 23 Comparison of Assays 39 Table 24 Evaluation of Limiting Ratio 41 Table 25 Recommended Values and 95 Uncertainties for Certification 43 Table C-1 Example OpenBUGS Homogeneity Summary 55 Table C-2 Example OpenBUGS Certification Summary for Component A 60 Table C-3 Example OpenBUGS Certification Summary for Component B 61 Table C-4 Example OpenBUGS Certification Summary for Component C 63 Table C-5 Parameter Values for Components A B and C 64 Table C-6 Certified Values for Components A B and C 64 Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A 66 Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B 66 Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C 67 Table C-10 Results for mtDNAnDNA 68 Table D-1 Summary [DNA] Results for the Commercial qPCR Assays 72

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

i

Table of Figures Figure 1 Sales History of SRM 2372 3 Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays 6 Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays 7 Figure 4 Plate Layout Design for Homogeneity Measurements 11 Figure 5 Plate Layout Design for Stability Measurements 11 Figure 6 Basic Plate Layout Design for Certification Measurements 12 Figure 7 Plate Layout Design for Performance Assessments 14 Figure 8 Absorbance Spectra of SRM 2372a Components A B and C 16 Figure 9 ddPCR Stability Measurements as Function of Time 18 Figure 10 Comparison of Assay Performance 20 Figure 11 mtDNAnDNA for Components A B and C 24 Figure 12 qPCR Autosomal Quantitation Performance Summary 26 Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary 26 Figure 14 Absorbance Spectra of SRM 2372 Components A and B 27 Figure 15 Comparison of Snap-Cooling to Slow-Cooling 32 Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures 33 Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 35 Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution 37 Figure 19 Comparison of Assays 38 Figure 20 Evaluation of Limiting Ratio 40 Figure 21 Degrees of Equivalence 42 Figure C-1 Homogeneity Assessment 56 Figure D-1 Results for the Innogenomics Commercial qPCR assays 69 Figure D-2 Results for the Thermo Fisher Commercial qPCR assays 70 Figure D-3 Results for the Qiagen Commercial qPCR assays 70 Figure D-4 Results for the Promega Commercial qPCR assays 71

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

1

Purpose and Description Standard Reference Material (SRM) 2372a is intended for use in the value assignment of human genomic deoxyribonucleic acid (DNA) forensic quantitation materials A unit of SRM 2372a consists of three well-characterized human genomic DNA materials solubilized in 10 mmolL 2-amino-2-(hydroxymethyl)-13 propanediol hydrochloride (Tris HCl) and 01 mmolL ethylenediaminetetraacetic acid disodium salt (disodium EDTA) pH 80 buffer (TE-4) The three component genomic DNA materials labeled A B and C are respectively derived from a single male donor a single female donor and 13 mixture of a male and a female donor A unit of the SRM consists of one 05 mL tube of each component each tube containing approximately 55 microL of DNA solution Each of these tubes is labeled and is sealed with a color-coded screw cap NIST is guided by and adheres to the ethical principles set forth in the Belmont Report1 SRM 2372a was developed after an appropriate human subject research determination by the NIST Human Subjects Protections Office

Warning SRM 2372a is a Human Source Material The three SRM 2372a components are human genomic DNA extracts prepared at NIST from buffy coat white blood cells from single-source anonymous donors All buffy coats were purchased from Interstate Blood Bank (Memphis TN) The supplier reported that each donor unit used in the preparation of this product was tested by FDA-licensed tests and found to be negative for human immunodeficiency virus (HIV-1 RNA and Anti-HIV 12) hepatitis B (HBV DNA and HBsAg) hepatitis C (HCV RNA and Anti-HCV) West Nile virus (WNV RNA) and syphilis Additionally each donor was tested according to FDA guidelines for Trypanosoma cruzi (Chagas) and found to be negativenon-reactive However no known test method can offer complete assurance that infectious agents are absent from this material Accordingly this human blood-based product should be handled at the Biosafety Level 12 SRM 2372a components and derived solutions should be disposed of in accordance with local state and federal regulations

Storage and Use Until required for use SRM 2372a should be stored in the dark between 2 degC and 8 degC The SRM 2372a component tubes should be mixed briefly and centrifuged (without opening the tube cap) prior to removing sample aliquots for analysis For the certified values to be applicable materials should be withdrawn immediately after opening the tubes and processed without delay Certified values do not apply to any material remaining in recapped tubes The certification only applies to the initial use and the same results are not guaranteed if the remaining material is used later

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

2

History and Background SRM 2372a is the second member of the SRM 2372 Human DNA Quantitation Standard series The supply of the initial material SRM 2372 was exhausted in May 2017 SRM 2372a is derived from commercially obtained buffy coat cells from different donors than those used in SRM 2372 A series of interlaboratory studies coordinated by what was then the Biotechnology Division of the National Institute of Standards and Technology (NIST) during the 1990s demonstrated that the DNA quantitation technologies then in use by the forensic human identification community adequately met the needs of the DNA typing tools then available34 However our 2004 study demonstrated that the increasingly sensitive and informative polymerase chain reaction (PCR)-based multiplex typing tools required more reliable determinations of DNA input quantity for best performance5 Quantitative PCR (qPCR) methods for DNA quantitation require calibration to an external standard of defined DNA concentration ([DNA]) expressed in units of nanograms of DNA per microliter of extract solution The true [DNA] of a commercial secondary standard may vary from its nominal concentration depending on supplier and lot number To enable providers of secondary standards to more reliably evaluate their materials and forensic testing laboratories to validate their use of commercial quantitation methods and supplies in 2007 we delivered the human DNA calibration standard SRM 2372 Human DNA Quantitation Standard6 The quantity of DNA in each of the three components of SRM 2372 was originally assigned based on the absorbance of double-stranded DNA (dsDNA) at 260 nm7 However by 2012 the absorbance properties of the materials had increased beyond the certified uncertainty limits due to slow but progressive changes in DNA tertiary structure Sales of the SRM were restricted until the DNA quantities were reassigned based on measurements of single-stranded DNA (ssDNA) in early 20138-10 Figure 1 displays the sales history of this relatively high-selling SRM While SRM 2372 was fit for the practical purpose of harmonizing DNA quantitation measurements to an accessible and reproducible reference material the conversion of both dsDNA and ssDNA absorbance to mass [DNA] is through conventional proportionality constants that are not metrologically traceable to the modern reference system for measurements the International System of Units (SI) Responding to requests from the forensic and clinical communities for calibration standards that provide traceability to the SI we and others have established droplet digital polymerase chain reaction (ddPCR) as a potentially primary measurement process that can provide traceable results11-18 The copy numbers of human nuclear DNA (nDNA) in the reaction mixture per partition symbolized as λ were determined by ddPCR using ten PCR assays optimized for use with our ddPCR platform These assays target loci on eight chromosomes with three assays on widely separated locations on one chromosome (see Table 1)

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

3

Figure 1 Sales History of SRM 2372

Each small black dot represents the sale of one unit of SRM 2372 The red lines summarize the average rate of sales as originally certified (2007 to 2012) and after re-certification (2013 to 2017) The magenta labels state the average yearly sales rate during these two periods

The nDNA copy number concentration human haploid genome equivalents (HHGE) per nanoliter component solution and symbolized here as λacute is calculated as

copy number [DNA] = λ = λ(VF) HHGEnL

where V is the average droplet volume and F is the total dilution factor the volume fraction of the component solution in the ddPCR reaction mixture The manufacturer of the ddPCR platform used a Bio-Rad QX200 Droplet Digital PCR System (Hercules CA) system does not provide metrologically traceable droplet volume information This lack of information prevented metrologically traceable conversion of the number per partition measurements to the desired number per reaction volume units In consequence NIST developed a measurement method that provides metrologically traceable droplet volumes19 We show elsewhere that the mass [DNA] nanograms of human DNA per microliter component can be traceably estimated as

mass [DNA] = 3031timesλacute ngmicroL

where the constant is the product of the number of nucleotide basepairs (bp) in the human reference genome the average bp molecular weight Avogadrorsquos number of bp per mole and

2372

2372(Recertif ied)

050

010

00

1713 unitsyear

1416 unitsyear

2007 2009 2011 2013 2015 2017

Uni

ts S

old

Calendar Year

Human DNA Quantitation StandardSRM 2372

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

4

unit scaling factors18 The three components of SRM 2372a are certified for both copy number and mass [DNA] Mitochondrial DNA (mtDNA) analysis within the forensic community is useful when the nDNA may be degraded or is in low-to-undetectable quantity The challenges of mtDNA testing include contamination of the standard due to routine laboratory use degradation of plasmids or synthetic oligomers or improper commutability of the standard relative to casework samples Commonly cell line materials are used as standards for qPCR with human genomic DNA which may lead to an incorrect mtDNA-to-nDNA ratio (mtDNAnDNA) for unknown samples Non-certified values of mtDNAnDNA for the three components are provided to bridge the current gap between well characterized mitochondrial DNA quantification assays and the availability of commercial standards With the understanding that the forensic DNA community primarily uses qPCR as a method for DNA quantification the performance of the three SRM 2372a materials was evaluated in eleven commercial qPCR chemistries currently used employed within the forensic DNA community

Experimental Methods Production and value-assignment of any certified reference material (CRM) is a complex process In addition to acquiring processing and packaging the materials included in the CRM the materials must be sufficiently homogenous across the units stable over time within its packaging and well-characterized to deliver results that are fit-for-purpose within the intended measurement community20 The process has been particularly complicated for SRM 2372a because we needed to determine the metrological properties of the required ddPCR measurement systems before we could trust the results they provide The following sections describe the methods and processes used to produce and characterize the SRM 2372a materials including sample preparation SRM unit production PCR assays ddPCR measurement processes spectrophotometric absorbance measurements homogeneity testing stability testing ddPCR nDNA and mtDNAnDNA measurements and performance in commercial qPCR assessments Many of these processes were mutually dependent therefore the sections are presented in a ldquodefinitionalrdquo order that sequentially introduces terminology Polymerase Chain Reaction Assays

Ten PCR assays were developed for characterization of SRM 2372a Each assay was confirmed to target one locus per HHGE using the National Center for Biotechnology Informationrsquos BLASTblastn system21 Primers were purchased from Eurofins Operon (Huntsville AL) TaqMan probes labeled with 6-carboxyfluorescein (FAM) were purchased from Thermo Fisher (Waltham MA) FAM-labeled Blackhole Quencher and Blackhole Quencher+ probes were purchased from LGC Biosearch Technologies (Novato CA) Table 1 details the ten nDNA assays used to certify copy number and mass [DNA] of the SRM 2372a components FAM-labeled TaqMan probes were purchased from Thermo Fisher

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

5

and used in monoplex reactions VIC (a proprietary dye)-labeled probes were purchased from the same source and were used when nDNA and mtDNA were analyzed in the same reaction mixture Figure 2 displays exemplar droplet patterns for the assays All assays were developed at NIST and have been optimized for ddPCR18 Table 2 details three mitochondrial DNA (mtDNA) assays used to estimate mtDNAnDNA in the SRM 2372a components Figure 3 displays exemplar droplet patterns for the assays These assays were optimized for ddPCR from real-time quantitative PCR (qPCR) assays from the literature All primers probes were diluted with TE-4 buffer purchased from Affymetrix-USB Products (Affymetrix Inc Cleveland OH) to a nominal 5 micromol working solution for all single-plex assays (individual nDNA and mtDNA assays) For duplex assays (combined nDNA and mtDNA) primer and probe working solutions were diluted to a nominal 10 micromol

Table 1 NIST-Developed Human Nuclear DNA Assays

Assay Target

Chromosome Band Accession Primers and Probe a

Amplicon Length bp

NEIF Gene EIF5B

Chr 2 p111-q111 NC_00000212

F gccaaacttcagccttctcttc R ctctggcaacatttcacactaca PB+ tcatgcagttgtcagaagctg

67

2PR4 Gene RPS27A

Chr 2 p16 NC_00000212

F cgggtttgggttcaggtctt R tgctacaatgaaaacattcagaagtct PB tttgtctaccacttgcaaagctggccttt

97

POTP STR TPOX

Chr 2 p253 NC_00000212

F ccaccttcctctgcttcacttt R acatgggtttttgcctttgg PT caccaactgaaatatg

60

NR4Q Gene DCK

Chr 4 q133-q211 NC_00000412

F tggtgggaatgttcttcagatga R tcgactgagacaggcatatgtt PB+ tgtatgagaaacctgaacgatggt

83

D5 STR D5S2500

Chr 5 q112 NC_00000510

F ttcatacaggcaagcaatgcat R cttaaagggtaaatgtttgcagtaatagat PT ataatatcagggtaaacaggg

75

ND6 STR D6S474

Chr 6 q21-22 NC_00000612

F gcatggctgagtctaaagttcaaag R gcagcctcagggttctcaa PB+ cccagaaccaaggaagatggt

82

D9 STR D9S2157

Chr 9 q342 NC_00000912

F ggctttgctgggtactgctt R ggaccacagcacatcagtcact PT cagggcacatgaat

60

HBB1 Gene HBB

Chr 11 p155 NC_00001110

F gctgagggtttgaagtccaactc R ggtctaagtgatgacagccgtacct PT agccagtgccagaagagccaagga

76

ND14 STR D14S1434

Chr 14 q3213C NC_0000149

F tccaccactgggttctatagttc R ggctgggaagtcccacaatc PB+ tcagactgaatcacaccatcag

109

22C3 Gene PMM1

Chr 22 q132 NC_00002210

F cccctaagaggtctgttgtgttg R aggtctggtggcttctccaat PB caaatcacctgaggtcaaggccagaaca

78

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

6

a F Forward primer R Reverse primer PB Blackhole quencher probe (FAM labeled) PB+ Blackhole Plus quencher probe (FAM labeled) PT TaqMan MGB probe (FAM labeled)

Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays

Table 2 NIST-Optimized Human Mitochondrial DNA Assays

Assay Location Primers and Probe a Amplicon Length bp Ref

mtND1 34853504 F ccctaaaacccgccacatct

69 22 35333553 R gagcgatggtgagagctaaggt 35063522 PT ccatcaccctctacatc

mtBatz 84468473 F aatattaaacacaaactaccacctacct

79 23 85038524 R tggttctcagggtttgttataa 84758491 PT cctcaccaaagcccata

mtKav 1318813308 F ggcatcaaccaaccacaccta

105 24 1336913392 R attgttaaggttgtggatgatgga 1331013325 PT cattcctgcacatctg

a F Forward primer R Reverse primer PT TaqMan MGB probe (FAM labeled)

2PR4

POTP

NEI

F

R4Q

5

D5

ND6

D9

HBB1

ND1

4

22C3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

7

Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays

Droplet Digital Polymerase Chain Reaction Measurements

Nuclear DNA (nDNA) All nDNA ddPCR measurements used in this study followed the same measurement protocol

bull Droplets were generated on the Auto Droplet Generator using the Bio-Rad ldquoddPCR ldquoSupermix for Probes (No dUTP)rdquo (Bio-Rad cat 186-3024 Control 64079080) and ldquoDroplet Generation Oil for Probesrdquo (Bio-Rad cat 186-4110 Control 64046051) Note desoxyuridine 5-triphosphatase (dUTP) is an enzyme sometimes included in PCR reaction mixtures to minimize PCR carryover contamination

bull Table 3 details the composition of the ddPCR mastermix setup for a given assay and sample where ldquomastermix refers to reaction mixture of manufacturerrsquos proprietary PCR reagent ldquoSupermixrdquo forward and reverse PCR primers for a given assay probe for that assay the sample DNA and sometimes additional water to produce the desired value dilution factor F This setup was used for all assays and samples

bull Non-Template Controls (NTCs) for each assay were included in every analysis bull Droplets were thermal cycled on a ProFlex thermal cycler (Applied Biosystems) for

95 degC for 10 min followed by 60 cycles of 94 degC for 05 min and 61 degC for 1 min then 98 degC for 10 min before a 4 degC hold until the plate was removed

bull Droplets were read on the QX200 Droplet Reader and analysis was performed using the QuantaSoft Analysis Software version 1740917

mtND1 mtBatz mtKav

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

8

bull The numbers of positive and negative droplets at the end of 60 cycles were determined and were exported into a spreadsheet for further analysis Assay-specific intensity thresholds were determined by visual inspection for each assay for each measurement session

Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays

Concentration nDNA Mastermix Microliter

per Reaction 2X ldquoSupermix for Probes (no dUTPs)rdquo 125

5 μmolL Forward Primer 1875 5 μmolL Reverse Primer 1875 5 μmolL Probe (FAM) 125

PCR Grade Water 50 14 Diluted Sample 25 Total Volume 250

Mitochondrial DNA (mtDNA) All the mtDNA measurements used in this study followed the nDNA protocol described above but with different mastermix setups The duplex mtDNA measurements used the setup described in Table 4 The monoplex mtDNA measurements used the setup described in Table 5

Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 125

10 μmolL Forward Primer 09375 10 μmolL Reverse Primer 09375 10 μmolL Probe (VIC) 0625

PCR Grade Water 95 nDNA Mastermix from Table 3 05 Total Volume 250

Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 120

10 μmolL Forward Primer 09 10 μmolL Reverse Primer 09 10 μmolL Probe (FAM) 06

PCR Grade Water 72 Diluted Sample (See Table 7) 24

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

9

Total Volume 240 Sample Preparation

The buffy coat white blood cells used to prepare SRM 2372a materials were purchased from Interstate Blood Bank These materials were from four single-source anonymous individuals supplied to us labeled as Female1 Female2 Male1 and Male2 In our laboratories each donor buffy coat bag was aliquoted (5 mL per aliquot) into sterile 50 mL conical tubes and stored at 4 degC prior to extraction The modified salt-out manual extraction protocol described in Appendix A was performed for each aliquot with rehydration of the DNA in TE-4 buffer From 150 mL to 250 mL of solubilized DNA for each component was accumulated through the extraction of the 5 mL aliquots The DNA was stored in perfluoroalkoxy polymer (PFA) containers that had been bleach sterilized thoroughly rinsed with deionized water rinsed with ethanol and air-dried The minimum volume for production of 2000 units was 120 mL Preliminary estimates of the mass [DNA] of the re-solubilized components were obtained using a Nanodrop microvolume spectrophotometer Dilutions of the component materials were made until the absorbance at 260 nm was approximately 1 corresponding to a nominal mass [DNA] of 50 ngμL of dsDNA7 After addition of the TE-4 buffer the containers were placed on an orbital shaker and gently rotated for several hours prior to refrigerated storage at 4 ordmC The four single-source components were initially screened with the nDNA ddPCR measurement process using four nDNA assays 2PR4 NR4Q ND6 and 22C3 Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate The gravimetric 1rarr4 dilution was prepared by weighing an empty PFA vial and recording the weight then pipetting 30 microL TE-4 into the vial and recording the weight then pipetting 10 microL DNA and recording the final weight The dilution factor was calculated in a spreadsheet These screening measurements verified that the individual component materials were fit-for-purpose The four single-source components were evaluated with the nDNA ddPCR measurement process using the ten nDNA assays Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate Quantitative information was derived from the average of the ten nDNA assays to prepare the component C mixture Thirty milliliters of sample Male2 (estimated at 452 ngμL mass [DNA]) were combined with 80 mL of sample Female2 (estimated at 507 ngμL mass [DNA]) and 10 mL of TE-4 buffer to create a 1 male nDNA to 3 female nDNA producing a 14 malehuman mixture After addition of the TE-4 buffer the component C container was gently rotated on the orbital shaker for several hours prior to refrigerated storage at 4 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

10

Absorbance Measurements

In February 2017 absorbance spectra of the recently prepared fully diluted nominally dsDNA and ssDNA solutions of the three components were acquired with a BioCary 100 spectrophotometer using the measurement protocol described in Appendix B In December 2017 additional spectra were acquired for the remaining bulk solution of components A and B and for SRM 2372 components A and B These solutions had been stored at 4 degC in well-sealed PFA containers There was no remaining bulk component C solution SRM Unit Production

In the morning of the day a component was transferred into the sterile component tubes (Sarstedt Biosphere SC micro tube-05 mL Newton NC) used to package the SRM the container holding the diluted component was removed from the refrigerator and placed on a slow orbital shaker for 2 h The tubes were opened and placed under the laminar flow hood in rows of 5 within 80-hole tube racks for a total of 30 tubes per rack Tubes were placed in every second row of the racks to facilitate filling with an Eppendorf Repeater Xstream pipette (Eppendorf North America Inc Hauppauge NY) fitted with a 1-mL positive displacement tip set to dispense 55 microL per tube The component container was removed from the shaker a magnetic stir-bar was added the container was placed on a magnetic stir plate in the laminar flow hood and the material was stirred gently Filling proceeded until 2200 vials were filled with pipet tip replacement after filling every 200 tubes The filling process consisted of two individuals manually filling each vial within a biosafety cabinet (one physically pipetting each sample the other verifying that each tube was filled) and four individuals tightly closing the lids to each tube on a sterilized and covered bench When all tubes in a rack were filled the rack was moved out of the hood the tubes checked for proper filling their lids closed and the tubes transferred to 100-unit pre-labeled storage boxes As the 100-unit boxes were filled they were transferred to the labeling room where SRM labels were applied by hand and colored inserts were added to the lids red for component A white for component B and blue for component C After all tubes for a given component were labeled the 22 boxes were placed in a locked refrigerator Filling of each of the components took 60 min to 90 min All units were transferred to the refrigerator within 4 hours of beginning the filling process All units of the three materials were allowed to equilibrate for a minimum of two weeks prior to beginning homogeneity analysis Component Homogeneity

The ddPCR measurement protocol described above was used to assess the nDNA homogeneity of the three components One tube from the approximate center of each of the 22 storage boxes was assayed using the NEIF and ND6 assays (see Table 1) A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Three technical replicates were obtained for every tube A total

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

11

of six plates were used two plates per component one with the NEIF assay and one with the ND6 assay Figure 4 details the layout of the samples

Figure 4 Plate Layout Design for Homogeneity Measurements

Component Stability

Stability of the materials was evaluated at 4 degC room temperature (RT asymp 22 degC) and 37 degC Tubes from the homogeneity study were divided among the three temperatures for the stability study Additionally two never-opened tubes (one from box 8 and one from box 12 of each component) were held at each temperature Two tubes at each temperature for each component were opened sampled and resealed 4 times over the course of 8 weeks Measurements were made at week 1 week 3 week 5 and week 8 The measurement on week 8 was from an unopened tube (from box 8) for each of the components at all temperatures The data from the homogeneity study was used as timepoint 0 in the stability study anchored to 4 degC by the storage recommendations A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used for both the NEIF and ND6 assays All temperatures components and both assays fit onto one plate for timepoints 1 to 5 Twelve NTCs placed in rows G and H were included for each assay Figure 5 details the sample layout

Figure 5 Plate Layout Design for Stability Measurements

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

ND6 ND6 ND6

NEIFNEIFNEIFComponent A

Component A

Component B

Component B

Component C

Component C

1 2 3 4 5 6 7 8 9 10 11 12A A-7 A-7 B-7 B-7 C-7 C-7 A-7 A-7 B-7 B-7 C-7 C-7B A-10 A-10 B-10 B-10 C-10 C-10 A-10 A-10 B-10 B-10 C-10 C-10C A-9 A-9 B-9 B-9 C-9 C-9 A-9 A-9 B-9 B-9 C-9 C-9D A-12 A-12 B-12 B-12 C-12 C-12 A-12 A-12 B-12 B-12 C-12 C-12E A-8 A-8 B-8 B-8 C-8 C-8 A-8 A-8 B-8 B-8 C-8 C-8F A-11 A-11 B-11 B-11 C-11 C-11 A-11 A-11 B-11 B-11 C-11 C-11G NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTCH NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

NEIF

4C

RT

37C

ND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

12

Component Certification Measurements

After establishing homogeneity and stability measurements intended to be used to certify nDNA values and inform mtDNA values were started using randomly selected SRM tubes from the boxes listed in Table 7 The ten nDNA assays described in Table 1 and two of the three mtDNA assays described in Table 2 were run simultaneously on the same sample plate A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used in each of the six measurement campaigns

Table 6 Tubes used for Certification Measurements

Component Plate A B C

dd20170711 Box 1 Box 1 Box 1 dd20170712a Box 10 tube a Box 10 tube a Box 10 tube a dd20170712b Box 10 tube b Box 10 tube b Box 10 tube b dd20170713 Box 20 Box 20 Box 20 dd20170714 Box 5 Box 5 Box 5 dd20170801 Box 5 Box 5 Box 5

Every sample was analyzed in duplicate with every assay in all six plates Component A was always located in rows B and C component B in rows D and E component C in rows F and G and NTCs in rows A and H Figure 6 displays the basic layout of the ten nDNA and two mtDNA assays In addition to the plate column assignment scheme shown the three assay groupings were also ordered as Group 2 Group 3 Group 1 and Group 3 Group 1 Group 2 to address potential positional bias

Group 1 Group 2 Group 3 2PR4 POTP NEIF NR4Q D5 ND6 D9 HBB1 ND14 22C3 mtND1 mtBatz 1 2 3 4 5 6 7 8 9 10 11 12 A NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Comp A B Comp A C Comp B D Comp B E Comp C F Comp C G

H NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Figure 6 Basic Plate Layout Design for Certification Measurements

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

13

Mitochondrial DNA Measurements

A ldquoduplexed reactionrdquo was used to determine mtDNAnDNA where 1 μL of one of the nDNA mastermix was added to 49 μL of the mtDNA mastermix resulting in a 1rarr50 volumetric dilution of the nDNA The nDNA was targeted with a FAM-labeled probe while the mtDNA was targeted with a VIC labeled probe The primer and probe concentrations were increased from 5 micromol to 10 micromol in the duplexed reactions Additional testing of the mtDNA assays were performed in monoplex with FAM-labeled probes to determine the additional sources of bias between sampling assay setup and dilution factors Table 7 outlines the different methods examined to determine mtDNAnDNA for the three components

Table 7 Run Parameters for Mitochondrial DNA Droplet

Generation nDNA mtDNA

Plate Component Assay Dilution Reps Assay Dilution Reps dd20170824 A manualDG HBB1 1rarr4 3 mtND1 1rarr200 9 dd20170825 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170829 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170831 AB autoDG HBB1 1rarr3 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr6 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr12 2 mtND1 1rarr10rarr100 3 dd20170907 ABC manualDG HBB1 1rarr3 2 mtBatz 1rarr10rarr100 3 dd20170913 AB manualDG NEIF 1rarr4 3 mtND1 mtBatz mtKav 1rarr10rarr100 3 dd20170913 C manualDG NEIF 1rarr4 3 mtBatz mtKav 1rarr10rarr100 3 Performance in Commercial qPCR Chemistries

Table 8 lists the 11 qPCR chemistries that were commercially available in late 2017 All three of the SRM 2372a components were evaluated with these chemistries using component A of SRM 2372 as the reference standard For each component of SRM 2372a four tubes were pulled from box 22 and combined into one tube A 1rarr5 dilution of all samples was prepared followed by three serial 1rarr2 dilutions A five-point serial dilution of each SRM 2372a component was amplified in triplicate for each chemistry This consisted of the neat material the 1rarr5 dilution then the three serial 1rarr2 dilutions A six-point serial dilution of SRM 2372 Component A was also created for the standard curve for the commercial qPCR chemistries This consisted of the neat material followed by a 1rarr5 dilution then followed by four 1rarr2 dilutions serially The six-point serial dilution series of SRM 2372 Component A was also amplified in triplicate These commercial chemistries were amplified in individual plates on an Applied Biosystems 7500 HID Real-Time PCR Instrument (Foster City CA) following the manufacturersrsquo recommended protocols for reaction mix setup and cycling conditions Figure 7 details the plate layout for the qPCR plates Data was analyzed with the HID Real-Time PCR Analysis Software v12 and further analyzed in a spreadsheet

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

14

Table 8 Commercial qPCR Chemistries Tested with SRM 2372a

Manufacturer Commercial Kit Target Codea

Innogenomics

InnoQuant Long IQh Short IQd

InnoQuantHY Long IQHYh Short IQHYd Y IQHYy

Thermo Fisher Quantifiler Duo Human QFDh Male QFDy

Quantifiler HP Large QFHPh Small QFHPd

Quantifiler Human Human QFHh

Quantifiler Trio Large QFTh Small QFTd Y QFTy

Qiagen

Quantiplex Autosomal QPh

Quantiplex Hyres Autosomal QPHh Y QPHy

Quantiplex Pro Human QPPh Degradation QPPd Y QPPy

Promega

Plexor Autosomal Ph Y Py

PowerQuant Autosomal PQh Degradation PQd Y PQy

a Assays are coded with a short acronym derived from the kit name followed by the following codes for the type of target ldquohrdquo autosomal non-degradation described as ldquoAutosomalrdquo ldquoHumanrdquo or ldquoLongrdquo ldquodrdquo autosomal degradation described as ldquoDegradationrdquo or ldquoShortrdquo ldquoyrdquo Y-chromosome described as ldquoMalerdquo or ldquoYrdquo

1 2 3 4 5 6 7 8 9 10 11 12

A

B 2372-A_57 2372-A_57 2372-A_57 2372a-neat_A

2372a-neat_A

2372a-neat_A

2372a-neat_B

2372a-neat_B

2372a-neat_B

2372a-neat_C

2372a-neat_C

2372a-neat_C

C 2372-A_11

4 2372-A_11

4 2372-A_11

4 2372a-15_A

2372a-15_A

2372a-15_A

2372a-15_B

2372a-15_B

2372a-15_B

2372a-15_C

2372a-15_C

2372a-15_C

D 2372-A_5

7 2372-A_5

7 2372-A_5

7 2372a-110_A

2372a-110_A

2372a-110_A

2372a-110_B

2372a-110_B

2372a-110_B

2372a-110_C

2372a-110_C

2372a-110_C

E 2372-A_2

85 2372-A_2

85 2372-A_2

85 2372a-120_A

2372a-120_A

2372a-120_A

2372a-120_B

2372a-120_B

2372a-120_B

2372a-120_C

2372a-120_C

2372a-120_C

F 2372-A_1

425 2372-A_1

425 2372-A_1

425 2372a-140_A

2372a-140_A

2372a-140_A

2372a-140_B

2372a-140_B

2372a-140_B

2372a-140_C

2372a-140_C

2372a-140_C

G 2372-A_0

7125 2372-A_0

7125 2372-A_0

7125

H NTC NTC NTC

Figure 7 Plate Layout Design for Performance Assessments

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

15

Measurement Results After preparing the SRM 2372a bulk solutions their approximate 50 ngmicroL mass [DNA] were confirmed spectrophotometrically The bulk solutions were then packaged into individual tubes The materials contained in the tubes were characterized for homogeneity stability nDNA copy number [DNA] and mtDNAnDNA ratio The following sections detail the results of these measurements Absorbance Spectrophotometry

By widely accepted convention an aqueous DNA solution having an absorbance of 10 at 260 nm corresponds to a mass [DNA] of 50 ngmicroL for dsDNA and (37 to 40) ngmicroL for ssDNA7 Absorbance at (230 270 280 and 330) nm are traditionally used in the assessment of DNA quality725 Due to minor baseline shifts during acquisition of some spectra it was necessary to bias-adjusted all spectra to have absorbance of 00 at 330 nm Table 9 lists the resulting bias-adjusted values for selected wavelengths and the conventional mass [DNA] conversion of the absorbance at 260 nm The close agreement between the mass concentration values for the nominally dsDNA and the converted-to-ssDNA spectra confirm that the bulk solutions were correctly prepared to be asymp 50 ngμL dsDNA The slight increase in absorbance over the 10 months between acquiring the spectra suggests that like SRM 2372 the tertiary structure of the dsDNA in the SRM 2372a materials may be changing slowly with time Figure 8 displays the spectra

Table 9 Absorbance at Selected Wavelengths

Wavelength nm Form Date Component 230 260 270 280 ngμL dsDNA 2142017 A 0394 0931 0758 0488 466

B 0480 1112 0905 0583 556 C 0451 1006 0815 0524 503

ssDNA 2142017 A 0390 0614 0540 0316 454 B 0457 0723 0634 0372 535 C 0411 0641 0563 0330 474

dsDNA 12142017 A 0402 0956 0775 0503 478 B 0484 1128 0914 0589 564

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

16

Figure 8 Absorbance Spectra of SRM 2372a Components A B and C

dsDNA is denatured to ssDNA by diluting with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to DNA mass concentration

Droplet Volume

Droplet volumes for the two lots of ldquoSupermixrdquo used in the studies described in this report were determined using NISTrsquos Special Test Method 11050S-D as described in Dagata et al19 The lot used for most of the ddPCR analyses produced (07349 plusmn 00085) nL droplets where the ldquoplusmnrdquo value is the combined standard uncertainty The lot used for analyses performed after October 11 2017 produced (07376 plusmn 00085) nL droplets NIST measurements have demonstrated that droplet volumes are not influenced by the presence or absence of DNA in the sample solution Measurements repeated over time on an earlier lot of the Supermix product used in these studies suggest that droplet volumes remain constant over the shelf-life of the lot However droplet volumes for different lots of this product as well as for different Supermix products have been observed to differ by several percent We intend to document these measurements in a future report Other researchers have demonstrated that non-constant droplet volume distributions may contribute to ddPCR measurement imprecision and bias2627 However these effects become significant only with distributions wider than observed using Special Test Method 11050S-D19 and at copydroplet values larger than the λ le 05 copydroplet typically used in our measurements Certification Measurements

Homogeneity Results Table 10 lists the λacute nDNA copy number values adjusted for both droplet volume and sample dilution from the homogeneity measurements The data was analyzed using the three-level Gaussian hierarchical ANOVA model described in Appendix C Table 11 lists the values of the within- and between-tube variance components expressed as percent relative standard deviations At the 95 confidence level all components are homogeneous

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372a-A2141721417-NaOH121417

220 260 300 340

Wavelength nm

2372a-B2141721417-NaOH121517

220 260 300 340

Wavelength nm

2372a-C21417

21417-NaOH

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

17

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 NEIF A-1-01 1632 1529 1511 B-1-01 1790 1909 1885 C-1-01 1493 1385 1510 ND6 1616 1640 1582 NAe 1870 1860 1565 1562 1466

NEIF A-1-02 1531 1525 1549 B-1-02 1712 1719 1704 C-1-02 1516 1513 1577 ND6 1615 1589 1601 1775 1718 1756 1490 1547 1570

NEIF A-1-03 1524 1540 1533 B-1-03 1693 1755 1749 C-1-03 1451 1528 1506 ND6 1582 1152 1555 1824 1483 1896 1485 1505 1471

NEIF A-1-04 1454 1458 1427 B-1-04 1754 1745 1739 C-1-04 1440 1486 1520 ND6 1627 1605 1607 1670 1632 1814 1529 1508 1506

NEIF A-1-05 1465 1540 1504 B-1-05 1757 1687 1735 C-1-05 1523 1472 1483 ND6 1607 1633 1586 1683 1759 1703 1513 1520 1491

NEIF A-1-06 1512 1491 1491 B-1-06 1697 1709 1714 C-1-06 1437 1493 1522 ND6 1511 1483 1524 1693 1723 1722 1492 1481 1461

NEIF A-1-07 1418 1480 1455 B-1-07 1722 1625 1703 C-1-07 1485 1477 1463 ND6 1479 1557 1656 1690 1684 1688 1475 1547 1485

NEIF A-1-08 1496 1560 1501 B-1-08 1884 1802 1772 C-1-08 1433 1405 1478 ND6 1514 1615 1540 1864 1887 1821 1497 1501 1495

NEIF A-1-09 1588 1581 1637 B-1-09 1713 1734 1727 C-1-09 1463 1499 1505 ND6 1474 1438 1454 1709 1756 1758 1503 1491 1516

NEIF A-1-10 1574 1513 1570 B-1-10 1759 1809 1738 C-1-10 1534 1458 1454 ND6 1594 1713 1669 1764 1816 1787 1448 1465 1474

NEIF A-1-11 1557 1508 1633 B-1-11 1749 1782 1740 C-1-11 1533 1514 1555 ND6 1571 1611 1631 1741 1792 1784 1488 1470 1483

NEIF A-1-12 1500 1476 1515 B-1-12 1748 1712 1718 C-1-12 1450 1525 1497 ND6 1626 1586 1608 1797 1614 1790 1496 1483 1519

NEIF A-1-13 1672 1634 1804 B-1-13 1632 1668 1701 C-1-13 1537 1527 1515 ND6 1490 1516 1497 1626 1566 1681 1499 1479 1475

NEIF A-1-14 1551 1568 1538 B-1-14 1752 1535 1673 C-1-14 1515 1490 1494 ND6 1617 1614 1639 1724 1741 1685 1469 1480 1466

NEIF A-1-15b 1793 1722 1733 B-1-15 1871 1835 1887 C-1-15 1511 1568 1594 ND6 1818 1753 1762 1770 1752 1798 1515 1543 1565

NEIF A-1-16 1573 1520 1617 B-1-16 1688 1725 1743 C-1-16 1449 1452 1471 ND6 1634 1627 1575 1744 1697 1710 1543 1448 1491

NEIF A-1-17 1516 1516 1628 B-1-17 1789 1769 1825 C-1-17 1388 1421 1394 ND6 1613 1567 1567 1716 1762 1638 1452 NAe 1435

NEIF A-1-18 1612 1535 1672 B-1-18 1826 1780 1849 C-1-18 1475 1455 1445 ND6 1529 1495 1528 1805 1801 1931 1458 1531 1459

NEIF A-1-19 1598 1514 1596 B-1-19 1751 1759 1880 C-1-19 1468 1525 1446 ND6 1571 1535 1511 1827 1766 1786 1435 1478 1421

NEIF A-1-20 1575 1568 1586 B-1-20 1738 1678 1768 C-1-20 1549 1531 1499 ND6 1652 1545 1627 1752 1747 1768 1499 1465 1468

NEIF A-1-21 1658 1611 1578 B-1-21 1742 1836 1713 C-1-21 1530 1440 1484 ND6 1647 1607 1612 1746 1744 1764 1499 1454 NAe

NEIF A-1-22 1547 1528 1532 B-1-22 1761 1743 1660 C-1-22 1442 1487 1436 ND6 1537 1513 1536 1666 1692 1819 1589 1591 1556

NEIF A-1-15c 1456 1448 1523 ND6 1525 1402 1436

NEIF A-2-15d 1679 1599 1674 ND6 1581 1550 1584

a Tube coded as Component-Tube-Box b These results were identified as a probable error in the preparation or volumetric delivery of the sample

solution into the microfluidic device and were not used in the homogeneity evaluations c Repeat assessment of the solution in sample A-1-15 d Results for a second tube from box 15 analyzed at the same time as the repeat assessment of A-1-15 e Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

18

Table 11 Within- and Between-Tube Relative Standard Deviations

Within-Tube Component NEIF ND6 Between-Tube

A 28 40 33 B 27 38 25 C 23 18 19

Stability Results Table 12 lists the λacute volume- and dilution-adjusted nDNA copy number values from the stability measurements Figure 9 summarizes these results The results for the tubes held at 4 degC room temperature (RT asymp 22 degC) and 37 degC are very similar for all three components The absence of between-temperature differences identifies the between-week differences as attributable to variation in the measurement processes rather than sample instability These results indicate that the SRM 2372a genomic copy number is thermally stable from 4 degC to 37 degC over an extended period

Figure 9 ddPCR Stability Measurements as Function of Time

Each dot-and-bar symbol represents the mean λacute and the approximate 95 expanded uncertainty for all results for a given component at a given storage temperature The thick black horizontal lines represent an approximate 95 confidence interval on the λacute values estimated from the homogeneity data

13

14

15

16

17

18

19

0 1 2 3 4 5 6 7 8

λacute H

aplo

id n

DNA

Copi

es μL

Weeks

A

0 1 2 3 4 5 6 7 8

Weeks

B

4 degC 22 degC 37 degC

0 1 2 3 4 5 6 7 8

Weeks

C

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

19

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Week Temp Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3

1 4 degC NEIF A-10 1568 1441 B-10 1666 1653 C-10 1424 1394 ND6 A-10 1474 1494 B-10 1655 1705 C-10 1461 1394

NEIF A-07 1500 1463 B-07 1718 1640 C-07 1446 1452 ND6 A-07 1479 1479 B-07 1708 1686 C-07 1522 1482

RT (22 degC) NEIF A-12 1674 1608 B-12 1805 1777 C-12 1506 1495 ND6 A-12 1522 1507 B-12 1759 1788 C-12 1478 1461 NEIF A-09 1635 1539 B-09 1791 1793 C-09 1411 NAb ND6 A-09 1543 1519 B-09 1661 1682 C-09 1482 1496

37 degC NEIF A-11 1614 1637 B-11 1748 1744 C-11 1515 1464 ND6 A-11 1494 1502 B-11 1740 1697 C-11 1495 1535 NEIF A-08 1592 1614 B-08 1691 1725 C-08 1528 1500 ND6 A-08 1581 1595 B-08 1727 1699 C-08 1458 1483 3 4 degC NEIF A-10 1520 1706 B-10 NAb 1733 C-10 1427 1473 ND6 A-10 1543 1624 B-10 1670 1755 C-10 1443 1364 NEIF A-07 1537 1570 B-07 1549 1565 C-07 1424 1512 ND6 A-07 1517 1479 B-07 1705 1723 C-07 1483 1439

RT (22 degC) NEIF A-12 1521 1607 B-12 1817 1842 C-12 1566 1471 ND6 A-12 1644 1597 B-12 1845 1786 C-12 1504 1511 NEIF A-09 1432 1626 B-09 1825 1856 C-09 1392 1387 ND6 A-09 1539 1528 B-09 1734 1737 C-09 1535 1465

37 degC NEIF A-11 1651 1685 B-11 1759 1779 C-11 1437 1440 ND6 A-11 1680 1594 B-11 1806 1806 C-11 1471 1509 NEIF A-08 1580 1613 B-08 1740 1791 C-08 1582 1590 ND6 A-08 1592 1524 B-08 1769 1694 C-08 1468 1520 5 4 degC NEIF A-10 1486 1440 B-10 1626 1551 C-10 1381 1330 ND6 A-10 1555 1577 B-10 1673 1625 C-10 1547 1395 NEIF A-07 1387 1429 B-07 1723 1787 C-07 1330 1314 ND6 A-07 1538 1522 B-07 1862 1909 C-07 1452 1458

RT (22 degC) NEIF A-12 1601 1631 B-12 1718 1752 C-12 1546 1541 ND6 A-12 1699 1677 B-12 1772 1721 C-12 1490 1471 NEIF A-09 1381 1484 B-09 1675 1694 C-09 1371 1376 ND6 A-09 1625 1615 B-09 1904 1815 C-09 1468 1472

37 degC NEIF A-11 1456 1517 B-11 1720 1704 C-11 1505 1493 ND6 A-11 1650 1658 B-11 1148 1152 C-11 1527 1600 NEIF A-08 1590 1666 B-08 1745 1723 C-08 1469 1540 ND6 A-08 1235 1244 B-08 1809 1749 C-08 1508 1529 8 4 degC NEIF A-08a 1667 1632 1589 B-08a 1880 1884 1940 C-08a 1525 1475 1542 ND6 A-08a 1599 1593 1590 B-08a 1859 1897 1972 C-08a 1550 1502 1520

RT (22 degC) NEIF A-08b 1693 1666 1634 B-08b 1769 1884 1771 C-08b 1595 1557 1545 ND6 A-08b 1644 1620 1647 B-08b 1862 1738 1811 C-08b 1611 1585 1535

37 degC NEIF A-08c 1620 1576 1608 B-08c 1934 1952 1956 C-08c 1595 1600 1569 ND6 A-08c 1576 1553 1532 B-08c 1889 1844 1889 C-08c 1569 1514 1581

a Tube coded as Component-Tube-Box b Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

20

Ten-Assay Certification Results The original intention was to certify the copy number [DNA] of the three components with results from five independently prepared sets of samples analyzed over four days Each sample of the three components was from a new tube from different boxes and was evaluated with 10 nDNA assays The assay positions were alternated across these five ldquo10-assay certificationrdquo plates The λ results for these five plates are systematically somewhat larger than those for the homogeneity and stability data On review we realized that a different workspace and its suite of pipettes had been used to prepare the 1rarr10 dilution of the 1rarr4 diluted components into the reaction mastermix The ldquohighrdquo workstation was used for stability and homogeneity measurements while the ldquolowrdquo workstation was used for certification measurements Gravimetric testing of the pipettes established that the volumes delivered by these alternate pipettes was sufficiently biased to account for the observed λ differences A sixth ldquocertificationrdquo plate was prepared using the original (and now re-verified) pipettes The λ results or this plate agreed well with the homogeneity and stability data Table 13 lists the λacute volume- and dilution-adjusted nDNA copy number values for all six of the ten-assay datasets Figure 10 displays the extent of agreement among the ten assays While there are small-but-consistent differences between the assays there is only a 4 difference between the assays providing smallest and largest λ

Figure 10 Comparison of Assay Performance

The blue dot-and-bars represent the mean and its combined standard uncertainty for the A B and C components for each assay relative to the over-all grand mean for each component The black dot-and-bars represent combination of three component results for each assay The solid horizontal line represents the median difference of the assay means from the grand mean the dotted horizontal lines bound an approximate 95 confidence region about the median

2PR4POTPNEIFR4Q5 D5ND6 D9 HBB1ND14 22C3-10

-8

-6

-4

-2

0

2

4

6

8

10

Rela

tive

Diffe

renc

e fr

om M

ean

Assays

Component MeansAssay MeansMedianU95

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

21

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

7112017 2PR4 1667 1698 1869 1960 1630 1706 POTP 1655 1685 1850 1781 1596 1576 NEIF 1647 1665 1709 1707 1574 1538 R4Q5 1594 1658 1856 1787 1590 1555 D5 1655 1683 1848 1829 1610 1607 ND6 1629 1647 1828 1826 1589 1548 D9 1655 1652 1861 1858 1569 1572 HBB1 1644 1611 1825 1869 1565 1681 ND14 1645 1707 1871 1773 1564 1492 22C3 1545 1603 1776 1902 1548 1655

7122017 2PR4 1668 1739 2174 2231 1499 1497 POTP 1587 1702 2202 2292 1576 1611 NEIF 1678 1614 2172 2216 1545 1560 R4Q5 1628 1762 NA 2078 1584 1546 D5 1738 1649 2230 2192 1577 1529 ND6 1667 1678 2094 2034 1599 1478 D9 1695 1653 1949 2110 1587 1504 HBB1 1757 1750 1971 2123 1520 1551 ND14 1771 1764 2123 2066 1479 1438 22C3 1762 1759 2107 2046 1512 1494

7122017 2PR4 1638 1667 1810 1806 1452 1424 POTP 1676 1638 1865 1801 1347 1397 NEIF 1616 1624 1806 1786 1395 1350 R4Q5 1559 1587 1878 1768 1442 1459 D5 1608 1652 1824 1790 1448 1602 ND6 1611 1589 1768 1670 1463 1435 D9 1517 1607 1811 1749 1402 1433 HBB1 1557 1582 1763 1741 1474 1481 ND14 1604 1625 1865 1781 1466 1482 22C3 1675 1616 1815 1866 1439 1568

7132017 2PR4 1888 1939 1784 1821 1470 1444 POTP 1868 1815 1735 1813 1411 1427 NEIF 1953 1902 1778 1813 1512 1439 R4Q5 1779 1871 1835 1749 1452 1468 D5 1924 2124 1794 1848 1567 1524 ND6 1861 1787 1700 1711 1480 1424 D9 1824 1844 1799 1779 1442 1459 HBB1 1809 1799 1765 1769 1430 1467 ND14 1864 1857 1754 1725 1413 1411 22C3 1930 1942 1792 1814 1486 1480

7142017 2PR4 1597 1590 1718 1768 1485 1458 POTP 1533 1404 1753 1698 1422 1459 NEIF 1462 1533 1773 1792 1438 1465 R4Q5 1493 1519 1777 1740 1439 1468 D5 1538 1586 1803 1795 1515 1457 ND6 1492 1516 1718 1693 1409 1412 D9 1528 1538 1802 1743 1469 1474 HBB1 1548 1581 1834 1745 1453 1428 ND14 1513 1514 1731 1714 1465 1425 22C3 1562 NA 1842 NA 1363 1393

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

22

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

101217 2PR4 1431 1411 1791 1770 1357 1364 POTP 1407 1395 1746 1820 1447 1361 NEIF 1487 1398 1803 1803 1448 1404 R4Q5 1450 1439 1767 1739 1409 1411 D5 1394 1422 1745 1773 1389 1429 ND6 1396 1393 1824 1729 1415 1417 D9 1414 1390 1752 1744 1431 1380 HBB1 1401 1382 1813 1754 1381 1336 ND14 1412 1460 1648 1671 1403 1387 22C3 1435 1410 1818 1741 1435 1372

Estimation of Copy Number Concentration It was decided that all three sets of results (homogeneity stability and certification) would be used to derive the certified values with the λ of the first five ldquoten-assayrdquo plates adjusted by a data-driven estimate of the volumetric bias This bias was incorporated into the statistical model providing better estimates of the λ means while capturing all of the variance components (homogeneity stability between-assay and between data sets) as random effects using Bayesian MCMC programmed in OpenBUGS28 Table 14 lists the estimated values and their uncertainties See Appendix C for details

Table 14 Measured Copy Number Concentrations

Genomic Copies per Nanoliter of Solution Component Mean a u(Mean) b U95(Mean) c CV95

d A 151 07 14 93 B 175 03 06 34 C 145 05 10 69

a Mean the experimentally determined most probable estimate of the true value b u(Mean) the standard uncertainty of the Mean c Mean plusmnU95(Mean) the interval which is believed with about 95 confidence to contain the true value d CV95 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the counting unit one29 2) the validity of the Poisson endpoint transformation for digital PCR endpoint assays when applied to samples providing le 05 copies per droplet 3) the determination that the copies are dispersed as dsDNA and 4) calibrated mean droplet volume measurements made at NIST during sample dilution and mastermix preparation Conversion of Copy Number to Mass Concentration As described in Reference 18 metrologically valid conversion from copies per nanoliter to nanograms per microliter requires quantitative estimates for the number of nucleotide bases per reference human genome and the average nucleotide mass Using the values estimated in that report and asserting that each target copy corresponds to one HHGE the required transformation is

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

23

[DNA] ngmicroL

= [DNA] HHGE

nLtimes

3301 ngHHGE

The combined standard uncertainty is

119906119906 [DNA] ng

microL = 1199061199062 [DNA] HHGE

nL + 0531100

times[DNA] HHGE

nL 2

Both the measured and the conversion factor uncertainties are associated with ldquolargerdquo degrees of freedom Therefor the approximate 95 confidence uncertainty is estimated

11988011988095 [DNA] ng

microL = 2 times 119906119906

[DNA] ngmicroL

Table 15 lists the transformation of the Table 14 copy number values These values agree well with the conventional 260 nm absorbance values averaging (4 plusmn 5) larger

Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution

ngmicroL Component Value a u(Measurement)b u(Conversion)c U95(Value)d CV95

e A 498 23 03 47 93 B 578 10 03 21 36 C 479 17 03 33 70

a Estimated true value b Standard uncertainty component from the ddPCR measurements c Standard uncertainty estimated from the relative uncertainty of the conversion factor d Value plusmnU95(Value) is believed with about 95 confidence to contain the true value e 100U95(Value)Value the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the measured estimates of copies per nanoliter and 2) determinations made by us from internationally recognized literature resources of the average number of nucleotide bases per HHGE and the mean molecular weights of the bases18 Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)

While not intended to provide certified values mtDNAnDNA was measured as part of 13 experiments over 5 months Measurements were made using dilution-adjusted λ for the mtND1 mtBatz andor mtKav mitochondrial assays referenced against the dilution-adjusted λ of either the NEIF or HBB1 nDNA assays By chance both DNAs in the component C mixture have a binding site mutation that silences the mtND1 probe Figure 11 graphically summarizes the results by assay and component Table 16 lists the data Table 17 summarizes results from a random effects model that accounts for between-assay variability See Appendix C for details

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

24

Due to the limited number of mtDNA assays used in these measurements these results are metrologically traceable only to the assays used

Figure 11 mtDNAnDNA for Components A B and C

Each dot-and-bar symbol represents a mean mtDNAnDNA and its approximate 95 expanded uncertainty The dotted horizontal lines represent approximate 95 confidence intervals on the ratios

Table 16 mtDNAnDNA Ratios

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

20170711a 1 mtND1 1931 1870 2461 2401 1 mtBatz 1845 1764 2555 2608 3201 3299

20170711b 1 mtND1 1810 1906 2229 2200 1 mtBatz 1785 1837 2108 2192 3347 3287

20170712 1 mtND1 1743 1761 2127 2198 1 mtBatz 1829 1847 2455 2506 3217 3384

20170713 1 mtND1 1716 1800 2211 2199 1 mtBatz 1740 1813 2380 2249 3200 3363

20170714 1 mtND1 1923 1835 2105 1943 1 mtBatz 1935 1851 2136 2119 3294 3013

20170814 1 mtND1 1780 1776 1706 2046 2057 2024 1 mtBatz 1804 1751 1738 1998 2043 1952 2927 2776 2762 1 mtKav 1769 1786 1761 2077 2117 2041 3044 2830 2825

20170824 1 mtND1 1947 1928 1922 2 mtND1 1829 1895 1851 3 mtND1 1674 1617

20170825 1 mtBatz 1667 1813 1708 1598 1928 1916 1915 2832 2840 2788 2725 2 mtBatz 1566 1687 1663 1589 1947 1946 1961 1909 2835 2790 2736 2679 3 mtBatz 1643 1668 1604 1573 1953 1952 1941 1881 2803 2773 2757 2752

20170829 1 mtBatz 1719 1716 1670 1666 1865 1863 1828 1828 2599 2508 2551 2609 2 mtBatz 1581 1573 1601 1612 1944 1964 1947 1895 2617 2483 2626 2602 3 mtBatz 1546 1542 1538 1587 1908 1909 1940 1839 2665 2535 2570 2650

20170831 1 mtND1 1617 1599 1688 1783

ND1

ND1

ND1Ka

v

Kav

Kav

Batz

Batz

Batz

A

B

C

160

210

260

mtD

NAn

DNA

Ratio

Assay

Bind

ing s

ite m

utat

ion

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

25

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

2 mtND1 1622 1579 1823 1631 3 mtND1 1575 1528 1920 1887

20170907 1 mtBatz 1584 1544 1890 1886 2501 2522 20170913 1 mtND1 1757 1724 1768 1990 1921 1982

1 mtBatz 1743 1671 1766 1990 1924 1958 2805 2825 2829 1 mtKav 1692 1662 1767 1903 1831 1919 2709 2738 2727

20171016 1 mtND1 1924 1944 1893 2304 2341 2332 1 mtBatz 1786 1752 1743 2048 2078 2067 2797 2642 2792 1 mtKav 1530 1542 1523 2399 2479 2493 2804 2789 2842 2 mtBatz 1580 1557 1592 2715 2606 2687 2 mtKav 2103 2100 2049 2799 2628 2559 3 mtND1 1749 1769 1777 2065 2166 2164 3 mtBatz 1965 1889 1947 2314 2280 2258 2639 2535 2657 3 mtKav 1813 1864 1864 1867 1987 1971 2953 2757 2802

Table 17 mtDNAnDNA Summary Results

mtDNAnDNA Component Value a u(Value)b U95(Value)c CV95

d A 1737 16 38 17 B 2058 22 44 21 C 2791 23 47 17

a Estimated true value b Standard uncertainty c Value plusmnU95(Value) is believed with about 95 confidence to contain the true value d 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) Performance in Commercial qPCR Chemistries

Traditional qPCR assays only quantify relative to a reference standard Thus qPCR assays do not provide absolute [DNA] for the individual materials but they can provide [DNA] of one component relative to another Figure 12 summarizes the BA and CA ratios for the commercial autosomal (Human) targets from the eleven assays evaluated (see Table 8) The Y-chromosome (Male) assays cannot be summarized in this manner as the (all female) component B does not include a Y chromosome Figure 13 summaries the mass [DNA] results for the (all male) component A and (male-female mixture) component C Since this material was prepared as a nominal 13 malefemale mixture the expected malehuman (total) ratio is 14 sjc As displayed in these figures there is intrinsic variability between the SRM 2372a components and between the commercial qPCR chemistries examined This reflects the difference in the targets and amplification chemistries for each of the commercial kits Many of the commercial chemistries on the market to date contain multi-copy targets for both autosomal and Y-chromosomal quantification to increase the sensitivity of the assays

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

26

Figure 12 qPCR Autosomal Quantitation Performance Summary

Open circles denote the BA CA ratios the 17 commercial qPCR assays evaluated each labeled with the code listed in Table 8 The solid black circle represents ratios as estimated by ddPCR The red lines bound the approximate 95 confidence intervals on the ddPCR-estimated ratios

Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary

Open squares denote the ngμL values for SRM 2372a Components A (male) and C (mixture of male and female) for the seven commercial Y-chromosome qPCR assays evaluated each labeled with the code listed in Table 8 The dotted red lines represent malehuman ratios of 13 14 and 15

Appendix D provides a graphical analysis of all the data generated in the performance assessments for each of the targets amplified in the eleven commercial qPCR chemistries and the [DNA] values for all of the assays

dd

IQ_h

IQ_d

IQHY_h

IQHY_d

QFD_h

QFH_h QFHP_h

QFHP_dQFT_h

QFT_d

QPH_h

QPP_hQPP_d

P_h

PQ_h

PQ_d

06

08

10

12

14

07 09 11 13

CA

BA

IQHY_y

QFD_y

QFT_y

QPH_y

QPP_y

P_y

PQ_y

1413

15

10

12

14

16

18

30 40 50 60 70

[DNA

] C n

gμL

[DNA]A nguL

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

27

Qualification of ddPCR as Fit-For-Purpose We have documented the requirements for establishing metrological traceability of ddPCR human genomic assay results expressed as copies per nanoliter of sample in a series of three peer-reviewed papers151617 Through the experimental evidence they are based upon these papers demonstrate that ddPCR can accurately enumerate the number of independently migrating genomic copies in an aqueous extract Our most recent paper establishes the metrological traceability of these ddPCR results when transformed into the nanogram per microliter units used by the forensic and clinical communities18 However our previous studies have not addressed a fundamental requirement for correctly interpreting digital PCR measurements In broad simplification end-point assay systems count the number of independently partitioning entities regardless of the number of copies the entities contain With assays that target just one site per HHGE entities that migrate as ssDNA will yield counts that are twice the value had they migrated as dsDNA30 Should some originally dsDNA be converted over time to ssDNA ddPCR measurements of the number of independently migrating entities would increase although the number of HHGE would be unchanged Measurements of mass [DNA] provided by absorbance spectrophotometry and ddPCR reflect different properties absorbance at 260 nm is sensitive to the spatial configuration of the nucleotides as well as their number while ddPCR counts independently migrating entities The SRM 2372a materials have absorbance spectra shown in Figure 8 consistent with being predominantly dsDNA The SRM 2372 materials now have the absorbance spectra shown in Figure 14 which are consistent with them being entirely ssDNA

Figure 14 Absorbance Spectra of SRM 2372 Components A and B dsDNA is denatured to ssDNA by diluting the source material with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to mass [DNA]

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372-A200720122012-NaOH121417

0

4

6

0

4

6

220 260 300 340

Wavelength nm

2372-B200720122012-NaOH121517

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

28

To ensure the validity of ddPCR-based value assignments it is thus necessary to evaluate the nature of the partitioning entities Should a solution of what is thought to be dsDNA contain an appreciable fraction of ssDNA then the ddPCR results would require adjustment31 To avoid further delay in making SRM 2372a available to the forensic community the following sections detail a series of experiments that have enabled us to conclude with high confidence that the SRM 2372a components have a dsDNA conformation and that for this material ldquoentitiesrdquo can be considered dsDNA ldquocopiesrdquo The remaining stock of the SRM 2372 components also have a dsDNA conformation however the strands in the old materials are much less robustly connected than they are in the new materials Digital Polymerase Chain Reaction Assays

Three nDNA assays were used HBB1 NEIF and ND6 The HBB1 and NEIF assays reproducibly yield very similar results for DNA extracts from many different donors The HBB1 target is at the tip of the short (p) arm of chromosome 11 The NEIF target is in the centromere region of chromosome 2 The ND6 assay was selected for use because it typically yields the lowest results of the ten nDNA assays used to certify the SRM 2372a components (see Figure 10 below) The ND6 target is about halfway down the long (q) arm of chromosome 6 Heat Treatments (Melting of the DNA into Single Strands)

In all experiments the non-heated control and the heat-treated diluted solutions were stored in the same type of PCR reaction tube The starting point for the heat treatment was to determine if there was a difference between snap-cooling on ice and slower cooling at 25 degC Two tubes of the diluted material were placed in a GeneAmp 9700 thermal cycler (Applied Biosystems) and cycled up to 98 degC and held for 15 min After 15 min one tube was removed and placed on ice while the remaining tube was cooled at 1 degCs in the thermal cycler to 25 degC then placed on ice until prepared for ddPCR Additional heat treatments included cycling samples on a GeneAmp 9700 thermal cycler

a) from 98 degC for 5 min then cooling to 25 degC for 2 min During the 25 degC hold a randomly selected tube was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 45 min at 98 degC

b) from 94 degC for 1 min then cooling to 25 degC for 1 min During the 1 min hold at 25 degC a random sample was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 9 min at 94 degC

c) from 90 degC for 1 min then cooling to 25 degC for 1 min for a total of 20 cycles Samples were pulled at 25 degC after 1 min 3 min 5 min 10 min and 20 min and placed on ice

After the above experiments heat treatments were performed using a ProFlex thermal cycler This thermal cycler has six temperature zones that can be le 5 degC apart This enables heating

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

29

tubes at the same time but at different temperatures After treatment tubes were cooled at 5 degCs to 4 degC in the thermal cycler While designed for use with PCR plates a MicroAmp 96-well tray for the VeriFlex block enabled its use with single PCR reaction tubes Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a

As a first step in evaluating whether the DNA in the SRM materials disperse as dsDNA or ssDNA a direct comparison of the five materials was made using the HBB1 and NEIF assays Table 18 reports summary statistics for the ddPCR measurements reported as λ copiesdroplet and the estimated mass [DNA] values from the ddPCR results18 from the conventional dsDNA relationship7 and from the conventional ssDNA relationship after denaturing with 04 molL NaOH89 While the mass [DNA] values differ somewhat between the three methods there is no evidence for the SRM materials segregating exclusively or even mostly as ssDNA The extent of the hypochromic interactions among the nucleotides does not accurately reflect the extent to which the two strands of dsDNA have separated

Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a

λ copies per droplet λ copies per droplet ngmicroL Sample Assay Nrep

a Mean CVb ncmbc Mean CVb ddPCRd dsDNAe ssDNAf

2372-A HBB1 4 03645 19 8 03644 20 566g 524 57 2372-A NEIF 4 03644 23 2372-B HBB1 4 03513 23 8 03487 35 580h 536 61 2372-B NEIF 4 03461 47 2372a-A HBB1 4 02603 27 8 02649 30 474 465 454 2372a-A NEIF 4 02694 24 2372a-B HBB1 4 02908 34 8 02891 37 517 556 534 2372a-B NEIF 4 02873 45 2372a-C HBB1 4 02757 28 8 02711 28 485 502 474 2372a-C NEIF 4 02665 16

a Number of technical replicates for the given assay b CV = 100(standard deviation)mean c Combined number of technical replicates ignoring potential differences between the assays d Estimated from the ddPCR λ measurements listed in this Table18 e Estimated from the conventional relationship between the mass [DNA] of dsDNA and absorbance at

260 nm7 using absorbance spectra acquired shortly after solutions were produced f Estimated from the relationship between the mass [DNA] of ssDNA and absorbance at 260 nm after 11

dilution with 04 molL NaOH89 g Values adjusted to measured (132 plusmn19) volume loss of archived units due to evaporation h Values adjusted to measured (71 plusmn25) volume loss of archived due to evaporation

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

30

Quantitative Evaluation of ssDNA Content

Absorbance spectrophotometry is routinely used to quantitatively assess DNA tertiary structure3233 On storage ldquoat physiological pH the intrastrand repulsion between the negatively charged phosphate groups forces the double helix into more rigid rodlike conformationrdquo34 suggesting a mechanism for the observed slow changes in the SRM 2372 materials As demonstrated in Table 18 absorbance at 260 nm does not necessarily correlate with ddPCR behavior Yet ldquofurthermore the repulsion between phosphate groups on opposite strands tends to separate the complementary strandsrdquo34 Therefore a method that is independent of hypochromism is needed to adequately evaluate the proportion (if any) of ldquotruerdquo (completely disjoint) ssDNA in the SRM materials Circular dichroism spectroscopy is sensitive to the conformational structures of DNA in solution35 and different DNA configurations migrate differentially in agarose gel electrophoresis7 Both techniques have been used to qualitatively discriminate ssDNA from dsDNA in ldquopurerdquo single-sequence materials but neither technique appears to be able to quantify small proportions of ssDNA in a complex multi-chromosome genomic material Wilson and Ellison have proposed a real-time chamber digital PCR (cdPCR) technique that can estimate excess proportions of ssDNA copies from the crossing-threshold distribution31 However their method is insufficiently sensitive for low ssDNA proportions It is well known that heating DNA beyond the melting temperature separates mammalian dsDNA into ssDNA36 albeit we have shown using cdPCR that boiling dsDNA for some minutes also renders some target sites non-amplifiable15 In the following section we document a ddPCR-based approach for estimating the proportion if any of ssDNA in a sample that involves less destructive heating While the method is not yet optimized the results from these preliminary studies demonstrate that the SRM 2372 and 2372a components disperse into the ddPCR droplets as dsDNA ddPCR Measurement of ssDNA Proportion Given two otherwise identical aliquots of a given DNA material one subjected to some treatment and the other untreated the ratio between the average number of copies per droplet of treated material and that of its untreated sibling λtλu quantitatively estimates the change in the number of independently dispersing entities Should the treatment reduce the number of accessible amplifiable targets (AAT) λtλu will be less than 1 Reduction in AAT may result from damage to the target itself andor conformational changes to the fragment containing the target that render the target inaccessible or non-amplifiable16 Assuming no denaturation of dsDNA to ssDNA either because of an ineffective treatment or the material having started as ldquopurerdquo ssDNA if the treatment does not itself damage targets then λtλu will equal 1 Should treatment convert at least some dsDNA to ssDNA without damaging targets the number of AAT will increase and λtλu will be greater than 1 Assuming complete conversion of ldquopurerdquo dsDNA to ldquopurerdquo ssDNA without other damage then λtλu will equal 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

31

Values greater than 1 but less than 2 indicate some conversion of dsDNA to ssDNA but also some combination of (a) incomplete denaturation of dsDNA to ssDNA (b) partial renaturation of ssDNA to dsDNA after treatment but before ddPCR analysis (c) damage to targets during the process or ndash of greatest interest ndash (d) the original sample material was an impure mixture of dsDNA and ssDNA The following discussions graphically summarize sets of λtλu followed by tabular listings of the relevant experimental conditions ddPCR measurements and the numerical λtλu values The information in these tables includes

Plate Index Experiment identifier indicates date performed in YYYYMMDD format Material Sample material SRM 2372-A 2372-B 2372a-A 2372a-B or 2372a-C Assay ddPCR assay HBB1 NEIF or ND6 degC Denaturation temperature in degC Min Length of time the treated material was held at denaturation temperature in

minutes Cool Method used to cool treated material from denaturation temperature Snap

Slow R1 or R5 where ldquoSnaprdquo indicates cooling on ice ldquoSlowrdquo cooling at room temperature to 25 degC ldquoR1rdquo cooling in a thermal cycler to 25 degC at 1 degCs and ldquoR5rdquo cooling in a thermal cycler to 4 degC at 5 degCs

F Dilution factor expressed as (volume Material) per (final volume of diluted solution) 12 14 or 18

n Number of technical replicates Mean Mean of technical replicates CV Coefficient of Variation expressed as a percentage 100 Standard Deviation

Mean

λtλu the copies per droplet result of the treated solution λt divided by the copies per droplet result of its untreated sibling λu

u(λtλu) Standard uncertainty of λtλu 119906119906(120582120582t 120582120582ufrasl ) = 120582120582t 120582120582ufrasl CV(120582120582t)100

2

+ CV(120582120582u)100

2

Proof of Principle Figure 15 summarizes two replicate sets of measurements performed to assess whether the SRM 2372 and 2372a materials respond similarly whether the denatured materials renature if allowed to cool slowly and whether results are (reasonably) repeatable Table 19 details the experimental conditions The λtλu values for the four materials are similar all much greater than 1 but not approaching 2 Most of the ratios are on or close to the ldquoslow cooling is the same as snap coolingrdquo line suggesting that renaturation is not a significant concern While the replicate results for the SRM 2372-A and 2372a-A materials do not overlap as closely as those for SRM 2372-B and 2372a-B similar treatment conditions produce similar results

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

32

Figure 15 Comparison of Snap-Cooling to Slow-Cooling

Symbols mark the means of replicate measurements made using the same dilutions on successive days with bars representing plusmn one standard combined uncertainty Results are labeled by material ldquoArdquo and ldquoBrdquo for SRM 2372 components and ldquoaArdquo ldquoaBrdquo and ldquoaCrdquo for SRM 2372a components The diagonal black line represents equality between the snap- and slow-cooling treatments

Table 19 Comparison of Snap-Cooling to Slow-Cooling

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171218 2372-A HBB1 14 4 03830 29

dd20171218 2372-A HBB1 98 15 Slow 14 2 05644 18 1474 0051 dd20171218 2372-A HBB1 98 15 Snap 14 4 05852 15 1528 0051

dd20171218 2372a-A HBB1 14 4 03029 18

dd20171218 2372a-A HBB1 98 15 Slow 14 4 04430 21 1463 0040 dd20171218 2372a-A HBB1 98 15 Snap 14 4 04454 33 1471 0055

dd20171218 2372-B HBB1 14 4 03594 31

dd20171218 2372-B HBB1 98 15 Slow 14 4 05806 18 1615 0058 dd20171218 2372-B HBB1 98 15 Snap 14 4 05825 13 1620 0055

dd20171218 2372a-B HBB1 14 4 03383 15

dd20171218 2372a-B HBB1 98 15 Slow 14 4 05043 38 1491 0061 dd20171218 2372a-B HBB1 98 15 Snap 14 4 05254 20 1553 0039

dd20171219 2372-A HBB1 14 4 03357 23

dd20171219 2372-A HBB1 98 15 Slow 14 4 05340 16 1591 0046 dd20171219 2372-A HBB1 98 15 Snap 14 4 05340 12 1591 0042

A

aA

B

aB

A

aA

B

aB

14

15

16

17

14 15 16 17

λ tλ

u Sn

ap C

oole

d

λtλu Slow Cooled

Rep 1Rep 2Snap = Slow

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

33

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171219 2372a-A HBB1 14 4 02735 24

dd20171219 2372a-A HBB1 98 15 Slow 14 4 04151 03 1518 0037 dd20171219 2372a-A HBB1 98 15 Snap 14 4 03918 21 1433 0045 dd20171219 2372-B HBB1 14 4 03346 22

dd20171219 2372-B HBB1 98 15 Slow 14 4 05476 17 1637 0045 dd20171219 2372-B HBB1 98 15 Snap 14 4 05480 29 1638 0059 dd20171219 2372a-B HBB1 14 4 03158 19

dd20171219 2372a-B HBB1 98 15 Slow 14 4 04798 16 1519 0038 dd20171219 2372a-B HBB1 98 15 Snap 14 4 04856 17 1538 0039

Time at Denaturation Temperature Figure 16 summarizes λtλu for the SRM 2372-B material as a function of denaturation temperature and the cumulative time the solution is held at that temperature Table 20 details the experimental conditions

Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures

The symbols represent the ratio of the mean result of three technical replicates of the untreated and each of the treated solutions the bars represent combined standard uncertainties The lines denote regression fits to the function λtλu = atimese-bt where t is the cumulative time Values and the standard uncertainties of the coefficients are listed in the legend

12

14

16

18

20

0 5 10 15 20

λ t λ

u

Cummulative Treatment Time min

degC a u (a ) b u (b )90 1804 0039 00198 0000494 1788 0029 00099 0002998 1855 0019 00067 00019

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

34

For all three of the denaturation temperatures studied the relationship between λtλu and cumulative time t can be modeled as

λtλu = ae-bt

where a is the extrapolated value of λtλu at t = 0 and b is the rate of change in AAT per minute As expected the rate of loss due to the treatment increases with increasing treatment temperature and cumulative time at that temperature37 Given the much longer treatment time for the 98 degC series the agreement among the three t = 0 parameter (a) values is remarkable This suggests that the first 1 min of treatment efficiently separates dsDNA into ssDNA The pooled relative standard uncertainty of these regression parameter is 17 somewhat smaller than the 25 pooled relative standard uncertainties of the individual λtλu However the plusmnstandard uncertainty interval of the 1-min λtλu results for the 94 degC and 90 degC both overlap the extrapolated t = 0 value suggesting that reliable results can be obtained using a single treatment of short duration

Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171228 2372-B HBB1 14 3 03650 17 dd20171228 2372-B HBB1 98 5 R1 14 3 06157 12 1687 0035 dd20171228 2372-B HBB1 98 10 R1 14 3 05493 08 1505 0028 dd20171228 2372-B HBB1 98 15 R1 14 3 05019 13 1375 0029 dd20171228 2372-B HBB1 98 20 R1 14 3 04583 21 1256 0034 dd20171228 2372-B HBB1 98 25 R1 14 3 04154 32 1138 0041 dd20171228 2372-B HBB1 98 30 R1 14 3 03784 14 1037 0023 dd20171228 2372-B HBB1 98 35 R1 14 3 03408 10 0934 0019 dd20171228 2372-B HBB1 98 40 R1 14 3 02980 01 0816 0014 dd20171228 2372-B HBB1 98 40 R1 14 3 02815 08 0771 0015

dd20171229 2372-B HBB1 14 3 03708 27

dd20171229 2372-B HBB1 94 1 R1 14 3 06598 26 1779 0056 dd20171229 2372-B HBB1 94 2 R1 14 3 06553 32 1767 0063 dd20171229 2372-B HBB1 94 3 R1 14 3 06301 20 1699 0045 dd20171229 2372-B HBB1 94 4 R1 14 3 06252 14 1686 0037 dd20171229 2372-B HBB1 94 5 R1 14 3 06577 14 1773 0039 dd20171229 2372-B HBB1 94 6 R1 14 3 06156 07 1660 0030 dd20171229 2372-B HBB1 94 7 R1 14 3 06118 10 1650 0033 dd20171229 2372-B HBB1 94 8 R1 14 3 06256 12 1687 0035 dd20171229 2372-B HBB1 94 9 R1 14 3 05999 10 1618 0032

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

35

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180104 2372-B HBB1 14 4 03713 25 dd20180104 2372-B HBB1 90 1 R1 14 3 06916 22 1863 0052 dd20180104 2372-B HBB1 90 3 R1 14 3 06465 06 1741 0032 dd20180104 2372-B HBB1 90 5 R1 14 3 06431 20 1732 0045 dd20180104 2372-B HBB1 90 10 R1 14 3 06014 25 1620 0049 dd20180104 2372-B HBB1 90 15 R1 14 3 06024 10 1623 0032 dd20180104 2372-B HBB1 90 20 R1 14 3 06003 14 1617 0035

Temperature Figure 17 summarizes the effect of denaturation temperature using a single treatment of 05 min on the 2372-B and 2372a-B materials Table 21 details the experimental conditions

Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

The symbols represent the ratio of the mean result of four technical replicates of the untreated and three replicates of the treated solutions the bars represent combined standard uncertainties In all cases the duration of the treatment was 05 min The lines denote regression fits to the linear function λtλu = a +bT where Tis the denaturation temperature

The SRM 2372 and 2372a materials respond very differently to changes in denaturation temperature The older spectrophotometrically ssDNA material suffers somewhat greater loss in λtλu as temperature increases The new spectrophotometrically dsDNA material suffers much greater loss in λtλu as temperature decreases

14

15

16

17

18

19

20

82 86 90 94 98

λ tλ

u

Temperature degC

2372-B

2372a-B

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

36

Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180109 2372-B HBB1 14 4 03577 25 dd20180109 2372-B HBB1 100 05 R5 14 3 06314 16 1765 0051

dd20180104a 2372-B HBB1 14 4 03577 05

dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032 dd20180104a 2372-B HBB1 94 05 R5 14 3 06459 25 1806 0046 dd20180104a 2372-B HBB1 90 05 R5 14 2 06591 08 1842 0017 dd20180104a 2372-B HBB1 88 05 R5 14 3 06589 05 1842 0013 dd20180104a 2372-B HBB1 84 05 R5 14 3 06729 10 1881 0022 dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

We hypothesize that in the older SRM 2372 material the between-strand bond is relatively fragile and can be completely severed at relatively low temperature while avoiding much temperature-dependent AAT loss In contrast the between-strand bond in the SRM 2372a material requires higher temperature to completely disassociate the strands This suggests that there is not a sample-independent ldquooptimumrdquo combination of denaturation temperature and treatment duration However 98 degC for 05 min appears to provide comparable results for these two materials DNA Concentration in the ddPCR Solution Figure 18 summarizes the effect of [DNA] in the ddPCR solution with SRM 2372-B at a denaturation temperature of 98 degC and treatment duration of 05 min Table 22 details the experimental conditions

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

37

Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution

The solid black symbols represent the ratio of the mean result of four technical replicates of the untreated and treated solutions with the λtλu values shown on the left-hand vertical axis the bars represent combined standard uncertainties The solid blue circles represent λu and the open red circles λt with values shown on the right-hand vertical axis at this graphical scale the standard deviations of the technical replicates are covered by the symbol The diagonal blue and red lines represent the log-linear relationships between just the 18 and 14 λ values

The primary effect of [DNA] appears to be the loss of linearity in λt with copiesdroplet values greater than about 1 copiesdroplet There may be some advantage to use somewhat greater dilutions than the routine 14 however the small increase in λtλu may be offset by the increased variability with small λu This suggests the use of dilutions that give λu values of (01 to 04) copiesdroplet

Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180117 2372-B HBB1 12 4 07127 11 dd20180117 2372-B HBB1 98 05 R5 12 4 12331 14 1730 0031

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 3 06151 15 1843 0040

dd20180117 2372-B HBB1 18 4 01549 18 dd20180117 2372-B HBB1 98 05 R5 18 4 02897 20 1870 0051

12 14 18

170

175

180

185

190

02

04

06

08

101214

λ T

arge

t D

ropl

et

λ tλ

u

Dilution Factor

λ₊λᵤλᵤλ₊

λtλu

λu

λt

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

38

Equivalence of Assays Figure 19 summarizes the agreement between the HBB1 NEIF and ND6 assays at a denaturation temperature of 98 degC and treatment duration of 05 min Table 23 details the experimental conditions

Figure 19 Comparison of Assays

The symbols compared the λtλu results from the HBB1 and either the NEIF or ND6 assays Each λtλu value is the ratio of the mean result of four technical replicates of an untreated and three or four replicates of the treated solutions The bars represent combined standard uncertainties The diagonal line represents equality between the assays

There is little or no difference between the λtλu results from the HBB1 NEIF and ND6 assays suggesting that the chromosomal location of the assay target has little influence Since the NEIFHBB1 comparison includes results for both the SRM 2372 and 2372a materials this equivalence is not material-specific

14

16

18

20

14 16 18 20

λ tλ

u NE

IF o

r ND6

λtλu HBB1

NEIFHBB1

ND6HBB1

2nd Assay = HBB1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

39

Table 23 Comparison of Assays

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu)

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045 dd20180105 2372a-B ND6 94 05 R5 14 3 05526 36 1711 0074 dd20180105 2372a-B ND6 90 05 R5 14 3 05326 18 1649 0051 dd20180105 2372a-B ND6 88 05 R5 14 3 05077 22 1572 0052 dd20180105 2372a-B ND6 86 05 R5 14 3 05038 23 1560 0053 dd20180105 2372a-B ND6 84 05 R5 14 3 04906 10 1519 0041

dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

40

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials

Figure 20 summarizes the currently available λtλu results obtained by denaturing at 98 degC for 05 min rapidly cooling in the thermal cycler with dilutions that provide asymp 03 copiesdroplet The figure also displays the consensus result and its 95 confidence region estimated using the Hierarchical Bayes method provided in the NIST Consensus Builder (NICOB) system3839 Table 24 details the experimental conditions

Figure 20 Evaluation of Limiting Ratio

The symbols represent the ratio of the mean result of three or four technical replicates of the untreated and treated solutions the bars represent combined standard uncertainties The thick horizontal line represents a consensus value covering all displayed results the dotted horizontal lines bracket the 95 confidence interval about the consensus value

The Hierarchical Bayes consensus result for these values is 183 with a 95 confidence interval of plusmn003 All but three of the 16 λtλu values are within one standard uncertainty of this consensus value with most of the λtλu means lying within the 95 confidence interval This suggests that there are no statistically significant differences between the SRM 2372 and 2372a materials with regard to ssDNA proportion

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

17

18

19

20

λ tλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

41

Table 24 Evaluation of Limiting Ratio

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180104a 2372-B HBB1 14 4 03577 05 dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032

dd20180108 2372-B HBB1 14 3 03463 09 dd20180108 2372-B HBB1 98 05 R5 14 3 06562 14 1895 0031

dd20180109 2372-B HBB1 14 3 03577 25 dd20180109 2372-B HBB1 98 05 R5 14 3 06495 21 1816 0059

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 4 06151 15 1843 0040

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

42

The agreement among the measured λtλu is more directly visualized using estimated unilateral ldquoDegrees of Equivalencerdquo d Each d is the absolute difference between the measured and consensus values

119889119889 = |120582120582119905119905 120582120582119906119906frasl minus consensus|

and is associated with a 95 confidence interval U95(d) that combines the uncertainty estimates of the measured and consensus values The U95(d) calculation details are specific to the consensus estimation method39 Values of d plusmn U95(d) that include 0 indicate that the λtλu result does not differ from the majority at about a 95 level of confidence Figure 21 displays the d for the λtλu results for the Hierarchical Bayes consensus value39 By this figure of merit all the λtλu can be considered as members of a single population

Figure 21 Degrees of Equivalence

The symbols represent the Degree of Equivalence d for the λtλu results relative to the Hierarchical Bayes consensus value The bars represent 95 level of confidence uncertainty intervals d plusmn U95(d) The horizontal line marks the d = 0 ldquono differencerdquo value Results with d plusmn U95(d) that include this line are not significantly different from the majority

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

-024

-018

-012

-006

000

006

012

018

024

Degr

ees o

f Equ

ival

ence

λtλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

43

Conclusions and Recommendations With high confidence all tubes of the SRM 2372a solution have the same nDNA copy number content within measurement repeatability Additionally within measurement repeatability all tubes of the SRM 2372a have the same mtDNAnDNA With high confidence the nDNA copy number content of the SRM 2372a solution is thermally stable from 4 degC to 37 degC over an extended period The solution should not be shipped or stored below 4 degC With high confidence the ddPCR measurements used to establish the entity and mass concentrations and the mtDNAnDNA ratios are adequately precise and unbiased With high confidence the DNA in the SRM 2372a solutions is now dsDNA and can be expected to remain dsDNA for at least ten years when stored at 4 degC Recommended Certification Values

Since the component materials are believed to be homogeneous are of very similar composition and were prepared and processed similarly the observed differences in the measured entity concentration uncertainties are likely artifacts of the measurement process To limit over-interpretation of these differences we recommend that the copy per nanoliter values of the three components be assigned same 95 relative uncertainty interval (CV95) based on the largest of the measured CV95 values (93 ) rounded up to 10 Since the ngμL transformation does not appreciably inflate this relative uncertainty we recommend that the mass concentration values also be assigned CV95 of 10 Similarly we recommend that the mtDNAnDNA ratios be rounded to integer values all three components be assigned the same CV95 and that this uncertainty be based on the largest (21 ) of the measured CV95 rounded up to 25 Table 25 lists the values that we recommend be used in the Certificate of Analysis

Table 25 Recommended Values and 95 Uncertainties for Certification

Component Units Value U95(Value) a A Copies per nanoliter 151 15 B Copies per nanoliter 175 18 C Copies per nanoliter 145 15 A ngμL 498 50

B ngμL 578 58 C ngμL 479 48 A mtDNAnDNA 174 4

B mtDNAnDNA 206 5 C mtDNAnDNA 279 7

a The interval Value plusmn U95(Value) is believed with about 95 confidence to contain the true value of the measurand

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

44

Use of SRM 2372a for Value-Assigning In-House Reference Materials

The within-component and between-assay variability observed for the commercial qPCR assays is not unexpected since different qPCR methods exploit qualitatively different molecular targets Given the intrinsic variability of the human genome the prospect of developing a human DNA source material that responds equally to all current and future qPCR assays is improbable All three components should be used to elucidate potential bias when using these materials to value assign in-house DNA reference and control materials for qPCR assays

Certificate of Analysis In accordance with ISO Guide 31 2000 a NIST SRM certificate is a document containing the name description and intended purpose of the material the logo of the US Department of Commerce the name of NIST as a certifying body instructions for proper use and storage of the material certified property value(s) with associated uncertainty(ies) method(s) used to obtain property values the period of validity if appropriate and any other technical information deemed necessary for its proper use A Certificate is issued for an SRM certified for one or more specific physical or engineering performance properties and may contain NIST reference information or both values in addition to certified values A Certificate of Analysis is issued for an SRM certified for one or more specific chemical properties Note ISO Guide 31 is updated periodically check with ISO for the latest version [httpswwwnistgovsrmsrm-definitions] For the most current version of the Certificate of Analysis for NIST SRM 2372a Human DNA Quantification Standard please visit httpswww-snistgovsrmorsview_detailcfmsrm=2372a

References

1 National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research Department of Health Education and Welfare (1978) Belmont Report Ethical Principles and Guidelines for the Protection of Human Subjects of Research DHEW Publication No (OS) 78-0012 httpvideocastnihgovpdfohrp_belmont_reportpdf

2 CDCNIH Biosafety in Microbiological and Biomedical Laboratories 5th ed HHS publication No (CDC) 21-1112 Chosewood LC Wilson DE Eds US Government Printing Office Washington DC (2009) httpwwwcdcgovbiosafetypublicationsbmbl5indexhtm

3 Duewer DL Kline MC Redman JW Newall PJ Reeder DJ NIST Mixed Stain Studies 1 and 2 Interlaboratory Comparison of DNA Quantification Practice and Short Tandem Repeat Multiplex Performance with Multiple-Source Samples J Forensic Sci 200146(5)1199-1210 PMID 11569565

4 Kline MC Duewer DL Redman JW Butler JM NIST Mixed Stain Study 3 DNA quantitation accuracy and its influence on short tandem repeat multiplex signal intensity Anal Chem 200375(10)2463-2469 httpsdoiorg101021ac026410i

5 Kline MC Duewer DL Redman JW Butler JM Results from the NIST 2004 DNA Quantitation Study J Forensic Sci 200550(3)571-578 PMID 15932088

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

45

6 Kline MC Duewer DL Travis JC Smith MV Redman JW Vallone PM Decker AE Butler JM Production and Certification of Standard Reference Material 2372 Human DNA Quantitation Standard Anal Bioanal Chem 2009394(4)1183-1192 httpsdoiorg101007s00216-009-2782-0

7 Sambrook J and Russell DW (2001) Molecular Cloning a Laboratory Manual Cold Spring Harbor Laboratory Press Cold Spring Harbor New York

8 EU-JRC Protocol NK603 ndash Community Reference Laboratory Event-specific method for the quantification of maize line NK603 using real-time PCR Section 33 Spectrophotometric Measurement of DNA Concentration httpgmo-crljrceceuropaeusummariesNK603-WEB-ProtocolValidationpdf

9 ISO 215712005(E) Foodstuffs ndash Methods of analysis for the detection of genetically modified organisms and derived products ndash Nucleic acid extraction Annex B Methods for quantitation of extracted DNA pp 34-36 International Organization for Standardization Geneva (2005)

10 Certificate of Analysis Standard Reference Material 2372 Human DNA Quantitation Standard (2013) Gaithersburg MD USA Standard Reference Materials Program National Institute of Standards and Technology httpswww-snistgovsrmorsview_certcfmsrm=2372

11 Vogelstein B Kinzler KW Digital PCR Proc Natl Acad Sci USA 1999969236-9241 12 Hindson BJ Ness KD Masquelier DA Belgrader P Heredia NJ Makarewicz AJ Bright IJ Lucero

MY Hiddessen AL Legler TC Kitano TK Hodel MR Petersen JF Wyatt PW Steenblock ER Shah PH Bousse LJ Troup CB Mellen JC Wittmann DK Erndt NG Cauley TH Koehler RT So AP Dube S Rose KA Montesclaros L Wang S Stumbo DP Hodges SP Romine S Milanovich FP White HE Regan JF Karlin-Neumann GA Hindson CM Saxonov S Colston BW High-Throughput Droplet Digital PCR System for Absolute Quantitation of DNA Copy Number Anal Chem 2011838604-8610 httpsdoiorg101021ac202028g

13 Pinheiro LB Coleman VA Hindson CM Herrmann J Hindson BJ Bhat S Emslie KR Evaluation of a Droplet Digital Polymerase Chain Reaction Format for DNA Copy Number Quantification AnalChem 2012841003-1011 httpsdoiorg101021ac202578x

14 Corbisier P Pinheiro L Mazoua S Kortekaas AM Chung PY Gerganova T Roebben G Emons H Emslie K DNA copy number concentration measured by digital and droplet digital quantitative PCR using certified reference materials Anal Bioanal Chem 2015407(7)1831-1840 httpsdoiorg101007s00216-015-8458-z

15 Duewer DL Kline MC Romsos EL Real-time cdPCR opens a window into events occurring in the first few PCR amplification cycles Anal Bioanal Chem 2015407(30)9061-9069 httpsdoiorg101007s00216-015-9073-8

16 Kline MC Romsos EL Duewer DL Evaluating Digital PCR for the Quantification of Human Genomic DNA Accessible Amplifiable Targets Anal Chem 201688(4)2132-2139 httpsdoiorg101021acsanalchem5b03692

17 Kline MC Duewer DL Evaluating Droplet Digital Polymerase Chain Reaction for the Quantification of Human Genomic DNA Lifting the Traceability Fog Anal Chem 201789(8)4648-4654 httpsdoiorg101021acsanalchem7b00240

18 Duewer DL Kline MC Romsos EL Toman B Evaluating Droplet Digital PCR for the quantification of human genomic DNA Conversion from Copies per Nanoliter to Nanogram Genomic DNA per Microliter Anal Bioanal Chem 2018in press httpsdoiorg101007s00216-018-0982-1

19 Dagata JA Farkas N Kramer JA Method for measuring the volume of nominally 100 μm diameter spherical water in oil emulsion droplets NIST Special Publication 260-184 2016 httpsdoiorg106028NISTSP260-184

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

20 ISO 170342016 General requirements for the competence of reference material producers International Organization for Standardization Geneva (2016)

21 NCBI Standard Nucleotide BLAST httpsblastncbinlmnihgovBlastcgiPAGE_TYPE=BlastSearch

22 Timken MD Swango KL Orrego C Buoncristiani MR A Duplex Real-Time qPCR Assay for the Quantification of Human Nuclear and Mitochondrial DNA in Forensic Samples Implications for Quantifying DNA in Degraded Samples J Forensic Sci 200550(5)1044-60 PMID 16225209

23 Walker JA Hedges DJ Perodeau BP Landry KE Stoilova N Laborde ME Shewale J Sinha SK Batzer MA Multiplex polymerase chain reaction for simultaneous quantitation of human nuclear mitochondrial and male Y-chromosome DNA application in human identification Anal Biochem 200533789ndash97 httpsdoiorg101016jab200409036

24 Kavlick MF Lawrence HS Merritt RT Fisher C Isenberg A Robertson JM Budowle B Quantification of Human Mitochondrial DNA Using Synthesized DNA Standards J Forensic Sci 201156(6) 1457ndash1463 httpsdoiorg101111j1556-4029201101871x

25 Stulnig TM Amberger A Exposing contaminating phenol in nucleic acid preparations BioTechniques 199416(3) 403-404

26 Majumdar N Wessel T Marks J Digital PCR Modeling for Maximal Sensitivity Dynamic Range and Measurement Precision PLoS ONE 201510(3)e0118833 httpsdoi101371journalpone0118833

27 Tellinghuisen J Partition Volume Variability in Digital Polymerase Chain Reaction Methods Polydispersity Causes Bias but Can Improve Precision Anal Chem 2016881283-12187 httpsdoiorg101021acsanalchem6b03139

28 Lunn DJ Spiegelhalter D Thomas A Best N (2009) Statistics in Medicine 283049ndash3082 29 De Biegravevre P Dybkaer R Fajgelj A Hibbert DB Metrological traceability of measurement results in

chemistry Concepts and implementation Pure Appl Chem 201183(10)1873-1935 httpsdoiorg101007s00769-011-0805-y

30 Huggett JF Foy CA Benes V Emslie K Garson JA Haynes R Hellemans J Kubista M Mueller RD Nolan T Pfaffl MW Shipley GL Vandesompele J Wittwer CT Bustin SA The Digital MIQE Guidelines Minimum Information for Publication of Quantitative Digital PCR Experiments Clin Chem 201359(6)892-902 httpsdoiorg101373clinchem2013206375

31 Wilson PJ Ellison SLR Extending digital PCR analysis by modelling quantification cycle data BMC Bioinfo 201617421 httpsdoiorg101186s12859-016-1275-3

32 Wang X Son A Effects of pretreatment on the denaturation and fragmentation of genomic DNA for DNA hybridization Environ Sci Processes Impacts 2013152204 httpsdoiorg101039c3em00457k

33 Nwokeoji AO Kilby PM Portwood DE Dickman MJ Accurate Quantification of Nucleic Acids Using Hypochromicity Measurements in Conjunction with UV Spectrophotometry Anal Chem 201789(24)13567-13574 https101021acsanalchem7b04000

34 Farkas DH Kaul KL Wiedbrauk DL Kiechle FL Specimen collection and storage for diagnostic molecular pathology investigation Arch Pathol Lab Med 1996120(6)591-6

35 Kypr J Kejnovskaacute I Renčiuk D Vorličkovaacute M Circular dichroism and conformational polymorphism of DNA Nucleic Acids Res 200937(6)1713-1725 httpsdoiorg101093nargkp026

36 Strachan T Read AP (1999) Human Molecular Genetics 2nd Ed John Wiley amp Sons NY

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

37 Bhat S McLaughlin JL Emslie KR Effect of sustained elevated temperature prior to amplification on template copy number estimation using digital polymerase chain reaction Analyst 2011136724ndash32 httpsdoiorg101039c0an00484g

38 NIST Consensus Builder (NICOB) httpsconsensusnistgov 39 Koepke A Lafarge T Toman B Possolo A NIST Consensus Builder Userrsquos Manual

httpsconsensusnistgovNISTConsensusBuilder-UserManualpdf

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

Appendix A DNA Extraction Method The following method has been adapted from Miller et al (1988)A1 Procedure

1 Transfer approximately 5 mL of buffy coat into a 50 mL sterile polypropylene centrifuge tube and add 30 mL Red Blood Cell Lysis Buffer

2 Mix by gentle inversion and let stand at room temperature for 15 min

3 Centrifuge at approximately 209 rads (2000 rpm) for 20 min

4 Discard supernatant

5 Add 9 mL of Nucleic Lysis Buffer and re-suspend the pelleted cells by agitating with a plastic transfer pipette

6 Add 300 microL of 20 sodium dodecyl sulfate (SDS) and 18 mL of Protease K Solution (freshly prepared 2 mg to 5 mg Protease K per 1 mL Protease K Buffer)

7 Mix by inversion followed by incubation overnight or longer at 37 ordmC to 40 ordmC in a shaking water bath The length of incubation is determined by the time required to solubilize the cells present in the tube

8 Add 4 mL saturated ammonium acetate solution (in deionized water stored at 2 ordmC to 8 ordmC) Shake vigorously for 15 s

9 Centrifuge at approximately 471 rads (4500 rpm) for 10 min

10 Pipette supernatant into a 50 mL sterile polypropylene centrifuge tube and add two volumes of room temperature absolute ethanol Mix by gentle inversion until the DNA aggregates and floats to the top of the tube

11 Spool the aggregated DNA from the current tube into a fresh tube of 80 ethanol using a sterile inoculation loop At this step multiple tubes of precipitated DNA may be pooled to make a large lot of the purified material Mix by gentle inversion

12 Transfer the aggregated DNA to a fresh dry 50 mL polypropylene centrifuge tube Remove as much ethanol as possible by squeezing the DNA against the wall of the tube with a sterile inoculation loop Air-dry the material in a laminar flow hood

13 Resolubilize DNA in TE-4 buffer and store material at 4 ordmC Reagents

Red Blood Cell Lysis Buffer 144 mmolL ammonium chloride 1 mmolL sodium bicarbonate in deionized (DI) water Store at 2 ordmC to 8 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

Nucleic Lysis Buffer 10 mmolL Tris-HCl 400 mmolL sodium chloride 20 mmolL disodium ethylenediaminetetraacetic acid in DI Water adjust pH to 82 Store at 2 ordmC to 8 ordmC

Protease K Buffer

20 mmolL disodium ethylenediaminetetraacetic acid 10 (mass per volume) Sodium Dodecyl Sulfate (SDS) in DI water Store at 2 ordmC to 8 ordmC

Protease K

Sigma-Aldrich cat P6556 Tritirachium album lyophilized powder Store at ndash20 ordmC

Ammonium Acetate

192 molL ammonium acetate (saturated) in DI water Store at 2 ordmC to 8 ordmC

Absolute Ethanol (200 proof)

Store at room temperature Reference

A1 Miller SA Dykes DD Polesky HF A simple salting out procedure for extracting DNA from human nucleated cells Nucleic Acids Res 198816(3)1215 httpsdoiorg101093nar1631215

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

48

Appendix B Spectrophotometric Measurement Protocol The following is for use with a scanning recording spectrophotometer The procedures include the spectrophotometer calibration measurement of the ldquonativerdquo dsDNA and measurement of the NaOH denatured ssDNA The procedure for evaluating solutions after treatment with NaOH is adapted from references89 Spectrophotometer Calibration

Establish Baseline Set the spectrophotometer to acquire data from 220 nm to 345 nm in steps of 01 nm Zero the instrument with nothing (air) in the sample and reference cells Acquire an air-air baseline of the instrument with nothing (air) in the sample and reference cells If the trace has significant structure or excessive noise take corrective action If the trace is acceptable proceed with data collection Enable Verification of Wavelength Calibration SRM 2034 is a solution of holmium oxide in 10 perchloric acid contained in a sealed quartz cuvette delivering an intrinsic UVVis wavelength standard for absorption spectrophotometryB1 Evaluation of the location of peak maxima in the spectra of this solution enables verification of the spectrophotometers wavelength calibration Place SRM 2034 in the sample cell Create and name a file to hold the verification spectra Perform three replicate SRM 2034-to-air scans saving each scan to the file Remove SRM 2034 from the spectrophotometer and store appropriately Enable Verification of Absorbance Calibration SRM 2031 consists of three metal-on-fused silica filters held in cuvette-sized filter holders plus an empty filter holder The three filters are certified for transmittance and transmittance density (-log10(transmittance) effectively absorbance) at several wavelengths with nominal wavelength-independent transmittances of 10 30 and 90 Evaluation of the absorbance of each of these three filters at the certified 250 nm 280 nm and 340 nm wavelengths enables verification of the spectrophotometers absorbance calibration Place the empty SRM 2031 filter holder in the reference cell Remove the protective covers from the 10 transmittance filter and place in the sample cell Perform three replicate scans saving each scan to file Remove the filter from the spectrophotometer replace the protective covers and store appropriately Repeat for the 30 and 90 filters Close the file containing the verification spectra For ldquonativerdquo dsDNA Samples

Verify Optical Balance of Cuvettes The measurements of the DNA-containing samples are made using a 100 microL quartz cuvette in the sample cell compared to an optically matched cuvette in the reference cell The optical equivalence of these two cuvettes can be verified by filling both cuvettes with appropriate volumes of a sample solution matrix (TE-4 buffer) that does not contain DNA Fill two dry

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

49

clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes followed by a methanol rinse in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquonativerdquo dsDNA remove and clean both the reference cuvette and the sample cuvette For NaOH denatured ssDNA Samples

Prepare a 04 molL NaOH Stock Solution To a 15 mL disposable centrifuge tube add 016 g NaOH pellets add deionized (DI) water to the 10 mL line and mix until all the NaOH pellets have dissolved Centrifuge at 524 rads (5000 rpm) for 5 min For every 9 samples to be analyzed aliquot 1 mL of this 04 molL NaOH stock solution into a 15 mL screw top tube Use these tubes to prepare the NaOH blanks and DNA samples In a 15 mL screw cap vial mix 200 microL TE-4 buffer with 200 microL 04 molL sodium hydroxide solution to achieve a final NaOH concentration of 02 molL Vortex 5 s then centrifuge at 1257 rads (12 000 rpm) for 1 min Fill two dry clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Prepare the Sample Solutions For each unit of material to be measured (Tube or vial)

a) Vortex the unit for 5 s b) Centrifuge at 1257 rads (12 000 rpm) for 1 min c) Remove a 100 microL aliquot and place in a new labeled vial d) Add 100 microL of the freshly prepared 04 molL NaOH solution (with the same pipet) e) Vortex 5 s f) Centrifuge at 1257 rads (12 000 rpm) for 1 min

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

50

Measure Sample Solutions Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes rinse with methanol in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the NaOH denatured DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquodenaturedrdquo ssDNA close the file holding the spectra for this series Remove and clean both the reference cuvette and the sample cuvette After All Sample Measurements are Complete

Enable Verification of Spectrophotometer Stability The stability of the spectrophotometer over the course of the data collection can be evaluated by comparing spectra for SRM 2034 and 2031 taken at the start of data collection with spectra taken at the end of data collection Repeat the wavelength and absorbance verification measurements described above collecting the spectra in a suitably named file Clean Up Transfer all relevant data files from the spectrophotometers storage to a transfer device Clean the spectrophotometer work area as necessary and restore the SRMs and cuvettes to their appropriate storage locations Evaluation

Export the files to a spreadsheet Average the replicate spectra Subtract the average absorbance at A320 from the absorbance at A260 resulting in the corrected absorbance at A260 The conventional spectrophotometric estimate of the mass concentration of the dsDNA in the sample is achieved by multiplying the corrected absorbance by 50 The conventional estimate for the NaOH-treated ssDNA is achieved by multiplying the corrected absorbance by 37times2 where the ldquotimes2rdquo adjusts for the 1rarr2 volumetric dilution of the sample Note reliable spectrophotometric measurements require corrected A260 to be greater than 005 Reference

B1 Travis JC Acosta JC Andor G Bastie J Blattner P Chunnilall CJ Crosson SC Duewer DL Early EA Hengstberger F Kim C-S Liedquist L Manoocheri F Mercader F Metzdorf J Mito A Monardh LAG Nevask S Nilsson M Noeumll M Rodriguez AC Ruiacutez A Schirmacher A Smith MV Valencia G van Tonder N Zwinkels J Intrinsic Wavelength Standard Absorption Bands in Holmium Oxide Solution for UVvisible Molecular Absorption Spectrophotometry J Phys Chem Ref Data 200534(1)41-56

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

51

Appendix C SRM 2372a Statisticianrsquos report Extracted from Blaza Toman

November 30 2017 1 Introduction Candidate material for SRM 2372a is composed of three components labeled A B and C All components are human genomic DNA and prepared at NIST from buffy coat white blood cells from single-source anonymous donors Component A and B are single source (malefemale respectively) and Component C is a 1 male to 3 female mixture of two single-source donors Droplet digital PCR (ddPCR) assays were used to obtain measurements of the concentration of copies per microliter of sample Ten NIST-developed human nuclear DNA and three literature-derived human mitochondrial DNA PCR assays were used in this study

11 OpenBUGS Analysis All parameters in the following sections were estimated using Markov Chain Monte Carlo (MCMC) implemented in the OpenBUGS freewareC1 OpenBUGS is a software application for the Bayesian analysis of complex statistical models

12 OpenBUGS Exemplar Output All following sections list the OpenBUGs model (ldquocoderdquo) initialization values (ldquoinitsrdquo) if they are needed and the data used to evaluate the various data sets discussed in the body of this report homogeneity stability nDNA certification and mtDNAnDNA value assignment The provided information in each of the following sections is sufficient to produce estimates but the estimates may not exactly reproduce the exemplar results provided The MCMC paradigm does not produce ldquoanalyticalrdquo values but approaches them asymptotically as the number o model updates increases All exemplar results presented here are based upon the following bull 10-fold ldquothinningrdquo to minimize autocorrelation there were 10 MCMC estimates for every

model update used in parameter estimation bull a ldquoburn inrdquo of 5 000 updates before beginning parameter estimation and bull 10 000 model updates

The following statistics are provided in the tables listing the exemplar results

mean arithmetic mean of the parameterrsquos model update values sd standard deviation of the parameterrsquos model update values MC_error Monte Carlo standard deviation of the mean meanradic(number of model updates) val25pc 25th percentile of the model update values median 50th percentile of the model update values val975pc 975th percentile of the model update values startnumber of ldquoburn inrdquo updates sample number of model update values

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

52

2 Homogeneity assessment 21 Data

Samples of the material of all three components were drawn from 22 different boxes Two assays (NEIF ND6) were used to obtain 3 replicates of the measurement in copies per nanoliter for each sample and component For the complete homogeneity data see

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

22 Model The data were analyzed using a three-level Gaussian hierarchical ANOVA model

119910119910119894119894119894119894119894119894~119873119873120572120572119894119894119894119894119894119894 120590120590119908119908119894119894119905119905ℎ119894119894119894119894 119887119887119887119887119887119887 1198941198941198941198942

where ~N() denotes ldquodistributed as a normal distribution of some mean μ and variancerdquo 119894119894 denotes component (1 for A 2 for B 3 for C) 119895119895 denotes assay (1 for NEIF 2 for ND6) 119896119896 denotes box (1hellip22) and 120572120572119894119894119894119894119894119894~119873119873micro119894119894119894119894120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119887119887119887119887119887119887119894119894

2

All variance components were given a Half Cauchy prior as described in Reference C2 Note The Half Cauchy prior is specified by ldquo~dnorm(0 00016)I(0001)rdquo The ldquoIrdquo notation is for censoring It cuts off any random draws from the Normal distribution that are lower than 0001 An informative prior is needed when there are less than four data per distribution and the half Cauchy is a good general-purpose prior

23 OpenBUGS code and data Homogeneity code ModelBegin for(i in 13)xiNsb[i] ~ dnorm(0 00016)I(0001) chSqNsb[i] ~ dgamma(0505) sigb[i] lt- xiNsb[i]sqrt(chSqNsb[i]) for(i in 13)for(j in 12)mu[ij]~dnorm(010E-5) for(i in 13) muA[i]lt-mean(mu[i])sigbP[i]lt-100(muA[i]sqrt(sigb[i])) for(i in 13)for(j in 12)xiNs[ij] ~ dnorm(0 00016)I(0001) chSqNs[ij]~dgamma(0505) sigt[ij]lt-xiNs[ij]sqrt(chSqNs[ij]) sigtP[ij]lt-100(mu[ij]sqrt(sigt[ij])) for(i in 13)for(j in 12)for(k in 122)alpha[ijk]~dnorm(mu[ij]sigb[i]) for(idx in 1393)lamb[idx]~dnorm(alpha[cmp[idx]asy[idx]box[idx]]sigt[cmp[idx]asy[idx]]) ModelEnd Inits None required

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

53

Homogeneity Data lamb[] asy[] box[] cmp[] rep[] 1632 1 1 1 1 1529 1 1 1 2 1511 1 1 1 3 1616 2 1 1 1 1640 2 1 1 2 1582 2 1 1 3 1531 1 2 1 1 1525 1 2 1 2 1549 1 2 1 3 1615 2 2 1 1 1589 2 2 1 2 1601 2 2 1 3 1524 1 3 1 1 1540 1 3 1 2 1533 1 3 1 3 1582 2 3 1 1 1152 2 3 1 2 1555 2 3 1 3 1454 1 4 1 1 1458 1 4 1 2 1427 1 4 1 3 1627 2 4 1 1 1605 2 4 1 2 1607 2 4 1 3 1465 1 5 1 1 1540 1 5 1 2 1504 1 5 1 3 1607 2 5 1 1 1633 2 5 1 2 1586 2 5 1 3 1512 1 6 1 1 1491 1 6 1 2 1491 1 6 1 3 1511 2 6 1 1 1483 2 6 1 2 1524 2 6 1 3 1418 1 7 1 1 1480 1 7 1 2 1455 1 7 1 3 1479 2 7 1 1 1557 2 7 1 2 1656 2 7 1 3 1496 1 8 1 1 1560 1 8 1 2 1501 1 8 1 3 1514 2 8 1 1 1615 2 8 1 2 1540 2 8 1 3 1588 1 9 1 1 1581 1 9 1 2 1637 1 9 1 3 1474 2 9 1 1 1438 2 9 1 2 1454 2 9 1 3 1574 1 10 1 1 1513 1 10 1 2 1570 1 10 1 3 1594 2 10 1 1 1713 2 10 1 2 1669 2 10 1 3

1557 1 11 1 1 1508 1 11 1 2 1633 1 11 1 3 1571 2 11 1 1 1611 2 11 1 2 1631 2 11 1 3 1500 1 12 1 1 1476 1 12 1 2 1515 1 12 1 3 1626 2 12 1 1 1586 2 12 1 2 1608 2 12 1 3 1672 1 13 1 1 1634 1 13 1 2 1804 1 13 1 3 1490 2 13 1 1 1516 2 13 1 2 1497 2 13 1 3 1551 1 14 1 1 1568 1 14 1 2 1538 1 14 1 3 1617 2 14 1 1 1614 2 14 1 2 1639 2 14 1 3 1456 1 15 1 1 1448 1 15 1 2 1523 1 15 1 3 1525 2 15 1 1 1402 2 15 1 2 1436 2 15 1 3 1573 1 16 1 1 1520 1 16 1 2 1617 1 16 1 3 1634 2 16 1 1 1627 2 16 1 2 1575 2 16 1 3 1516 1 17 1 1 1516 1 17 1 2 1628 1 17 1 3 1613 2 17 1 1 1567 2 17 1 2 1567 2 17 1 3 1612 1 18 1 1 1535 1 18 1 2 1672 1 18 1 3 1529 2 18 1 1 1495 2 18 1 2 1528 2 18 1 3 1598 1 19 1 1 1514 1 19 1 2 1596 1 19 1 3 1571 2 19 1 1 1535 2 19 1 2 1511 2 19 1 3 1575 1 20 1 1 1568 1 20 1 2 1586 1 20 1 3 1652 2 20 1 1 1545 2 20 1 2 1627 2 20 1 3 1658 1 21 1 1

1611 1 21 1 2 1578 1 21 1 3 1647 2 21 1 1 1607 2 21 1 2 1612 2 21 1 3 1547 1 22 1 1 1528 1 22 1 2 1532 1 22 1 3 1537 2 22 1 1 1513 2 22 1 2 1536 2 22 1 3 1790 1 1 2 1 1909 1 1 2 2 1885 1 1 2 3 1870 2 1 2 2 1860 2 1 2 3 1712 1 2 2 1 1719 1 2 2 2 1704 1 2 2 3 1775 2 2 2 1 1718 2 2 2 2 1756 2 2 2 3 1693 1 3 2 1 1755 1 3 2 2 1749 1 3 2 3 1824 2 3 2 1 1483 2 3 2 2 1896 2 3 2 3 1754 1 4 2 1 1745 1 4 2 2 1739 1 4 2 3 1670 2 4 2 1 1632 2 4 2 2 1814 2 4 2 3 1757 1 5 2 1 1687 1 5 2 2 1735 1 5 2 3 1683 2 5 2 1 1759 2 5 2 2 1703 2 5 2 3 1697 1 6 2 1 1709 1 6 2 2 1714 1 6 2 3 1693 2 6 2 1 1723 2 6 2 2 1722 2 6 2 3 1722 1 7 2 1 1625 1 7 2 2 1703 1 7 2 3 1690 2 7 2 1 1684 2 7 2 2 1688 2 7 2 3 1884 1 8 2 1 1802 1 8 2 2 1772 1 8 2 3 1864 2 8 2 1 1887 2 8 2 2 1821 2 8 2 3 1713 1 9 2 1 1734 1 9 2 2 1727 1 9 2 3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

54

1709 2 9 2 1 1756 2 9 2 2 1758 2 9 2 3 1759 1 10 2 1 1809 1 10 2 2 1738 1 10 2 3 1764 2 10 2 1 1816 2 10 2 2 1787 2 10 2 3 1749 1 11 2 1 1782 1 11 2 2 1740 1 11 2 3 1741 2 11 2 1 1792 2 11 2 2 1784 2 11 2 3 1748 1 12 2 1 1712 1 12 2 2 1718 1 12 2 3 1797 2 12 2 1 1614 2 12 2 2 1790 2 12 2 3 1632 1 13 2 1 1668 1 13 2 2 1701 1 13 2 3 1626 2 13 2 1 1566 2 13 2 2 1681 2 13 2 3 1752 1 14 2 1 1535 1 14 2 2 1673 1 14 2 3 1724 2 14 2 1 1741 2 14 2 2 1685 2 14 2 3 1871 1 15 2 1 1835 1 15 2 2 1887 1 15 2 3 1770 2 15 2 1 1752 2 15 2 2 1798 2 15 2 3 1688 1 16 2 1 1725 1 16 2 2 1743 1 16 2 3 1744 2 16 2 1 1697 2 16 2 2 1710 2 16 2 3 1789 1 17 2 1 1769 1 17 2 2 1825 1 17 2 3 1716 2 17 2 1 1762 2 17 2 2 1638 2 17 2 3 1826 1 18 2 1 1780 1 18 2 2 1849 1 18 2 3 1805 2 18 2 1 1801 2 18 2 2 1931 2 18 2 3 1751 1 19 2 1 1759 1 19 2 2 1880 1 19 2 3 1827 2 19 2 1 1766 2 19 2 2

1786 2 19 2 3 1738 1 20 2 1 1678 1 20 2 2 1768 1 20 2 3 1752 2 20 2 1 1747 2 20 2 2 1768 2 20 2 3 1742 1 21 2 1 1836 1 21 2 2 1713 1 21 2 3 1746 2 21 2 1 1744 2 21 2 2 1764 2 21 2 3 1761 1 22 2 1 1743 1 22 2 2 1660 1 22 2 3 1666 2 22 2 1 1692 2 22 2 2 1819 2 22 2 3 1493 1 1 3 1 1385 1 1 3 2 1510 1 1 3 3 1565 2 1 3 1 1562 2 1 3 2 1466 2 1 3 3 1516 1 2 3 1 1513 1 2 3 2 1577 1 2 3 3 1490 2 2 3 1 1547 2 2 3 2 1570 2 2 3 3 1451 1 3 3 1 1528 1 3 3 2 1506 1 3 3 3 1485 2 3 3 1 1505 2 3 3 2 1471 2 3 3 3 1440 1 4 3 1 1486 1 4 3 2 1520 1 4 3 3 1529 2 4 3 1 1508 2 4 3 2 1506 2 4 3 3 1523 1 5 3 1 1472 1 5 3 2 1483 1 5 3 3 1513 2 5 3 1 1520 2 5 3 2 1491 2 5 3 3 1437 1 6 3 1 1493 1 6 3 2 1522 1 6 3 3 1492 2 6 3 1 1481 2 6 3 2 1461 2 6 3 3 1485 1 7 3 1 1477 1 7 3 2 1463 1 7 3 3 1475 2 7 3 1 1547 2 7 3 2 1485 2 7 3 3 1433 1 8 3 1

1405 1 8 3 2 1478 1 8 3 3 1497 2 8 3 1 1501 2 8 3 2 1495 2 8 3 3 1463 1 9 3 1 1499 1 9 3 2 1505 1 9 3 3 1503 2 9 3 1 1491 2 9 3 2 1516 2 9 3 3 1534 1 10 3 1 1458 1 10 3 2 1454 1 10 3 3 1448 2 10 3 1 1465 2 10 3 2 1474 2 10 3 3 1533 1 11 3 1 1514 1 11 3 2 1555 1 11 3 3 1488 2 11 3 1 1470 2 11 3 2 1483 2 11 3 3 1450 1 12 3 1 1525 1 12 3 2 1497 1 12 3 3 1496 2 12 3 1 1483 2 12 3 2 1519 2 12 3 3 1537 1 13 3 1 1527 1 13 3 2 1515 1 13 3 3 1499 2 13 3 1 1479 2 13 3 2 1475 2 13 3 3 1515 1 14 3 1 1490 1 14 3 2 1494 1 14 3 3 1469 2 14 3 1 1480 2 14 3 2 1466 2 14 3 3 1511 1 15 3 1 1568 1 15 3 2 1594 1 15 3 3 1515 2 15 3 1 1543 2 15 3 2 1565 2 15 3 3 1449 1 16 3 1 1452 1 16 3 2 1471 1 16 3 3 1543 2 16 3 1 1448 2 16 3 2 1491 2 16 3 3 1388 1 17 3 1 1421 1 17 3 2 1394 1 17 3 3 1452 2 17 3 1 1435 2 17 3 3 1475 1 18 3 1 1455 1 18 3 2 1445 1 18 3 3 1458 2 18 3 1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

55

1531 2 18 3 2 1459 2 18 3 3 1468 1 19 3 1 1525 1 19 3 2 1446 1 19 3 3 1435 2 19 3 1 1478 2 19 3 2 1421 2 19 3 3 1549 1 20 3 1

1531 1 20 3 2 1499 1 20 3 3 1499 2 20 3 1 1465 2 20 3 2 1468 2 20 3 3 1530 1 21 3 1 1440 1 21 3 2 1484 1 21 3 3 1499 2 21 3 1

1454 2 21 3 2 1442 1 22 3 1 1487 1 22 3 2 1436 1 22 3 3 1589 2 22 3 1 1591 2 22 3 2 1556 2 22 3 3 END

Table C-1 Example OpenBUGS Homogeneity Summary

mean sd MC_error val25pc median val975pc start sample mu[11] 15460 0122 000123 1522 1546 1570 5001 10000 mu[12] 15640 0137 000141 1537 1564 1591 5001 10000 mu[21] 17490 0113 000118 1727 1749 1772 5001 10000 mu[22] 17480 0129 000125 1722 1748 1774 5001 10000 mu[31] 14870 0073 000073 1473 1487 1501 5001 10000 mu[32] 14960 0069 000063 1482 1496 1509 5001 10000 muA[1] 15550 0092 000092 1536 1555 1573 5001 10000 muA[2] 17490 0085 000086 1732 1749 1765 5001 10000 muA[3] 14910 0050 000046 1481 1491 1501 5001 10000 sigbP[1] 3269 04985 000514 2376 3244 4324 5001 10000 sigbP[2] 2548 04385 000461 1733 2533 3472 5001 10000 sigbP[3] 1850 0292 000260 1329 1832 2477 5001 10000

sigtP[11] 2774 0298 000260 2268 2750 3442 5001 10000 sigtP[12] 4036 0429 000438 3290 4005 4986 5001 10000 sigtP[21] 2664 0301 000321 2155 2640 3313 5001 10000 sigtP[22] 3807 0395 000377 3116 3772 4662 5001 10000 sigtP[31] 2313 0250 000209 1880 2291 2873 5001 10000 sigtP[32] 1812 0195 000170 1473 1797 2243 5001 10000

24 Results The material appears to be homogeneous as can be seen in Figure C-1 Based on these results the between-tube variability is of the same order as within-tube variability

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

56

Component A Assay NEIF Component A Assay ND6 box plot gamma[11]

[111][112][113]

[114]

[115][116]

[117]

[118]

[119]

[1110][1111]

[1112]

[1113]

[1114]

[1115]

[1116][1117]

[1118]

[1119][1120]

[1121]

[1122]

114

box plot gamma[12]

[121][122]

[123]

[124][125]

[126]

[127][128]

[129]

[1210]

[1211][1212]

[1213]

[1214]

[1215]

[1216]

[1217]

[1218][1219]

[1220][1221]

[1222]

115

Component B Assay NEIF Component B Assay ND6

box plot gamma[21]

[211]

[212]

[213][214]

[215][216]

[217]

[218]

[219]

[2110][2111]

[2112]

[2113][2114]

[2115]

[2116]

[2117]

[2118]

[2119]

[2120]

[2121]

[2122]

129

box plot gamma[22]

[221]

[222][223]

[224][225][226][227]

[228]

[229]

[2210][2211]

[2212]

[2213]

[2214]

[2215]

[2216][2217]

[2218]

[2219]

[2220][2221][2222]

128

Component C Assay NEIF Component C Assay ND6

box plot gamma[31]

[311]

[312]

[313][314]

[315][316]

[317]

[318]

[319][3110]

[3111]

[3112]

[3113]

[3114]

[3115]

[3116]

[3117]

[3118]

[3119]

[3120]

[3121]

[3122]109

box plot gamma[32]

[321][322]

[323]

[324][325]

[326]

[327][328][329]

[3210]

[3211]

[3212]

[3213][3214]

[3215]

[3216]

[3217]

[3218]

[3219]

[3220][3221]

[3222]

11

Figure C-1 Homogeneity Assessment

Note These results were obtained using the 1-4 gravimetric dilution-adjusted λ but uncorrected for droplet volume

or the 1-10 volumetric dilution That is the vertical axis in each panel should be multiplied by 07349times10 = 7349 This constant difference in scale does not impact the homogeneity analysis Error bars span approximate 95 confidence intervals

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

57

3 Certification measurements 31 Data

Originally the experiment consisted of 5 plates containing the 3 components (A B C) and 12 assays Ten of the assays were for genomic DNA identification (for the certified value) and two of the assays were for mitochondrial detection which is used for informational values (and are not included in this data set) Each of the plates contained samples from a new vial from different boxes (noted in the ldquovial usedrdquo portion the number references the box) The assay positions were alternated across the five plates as well For the plate design and which tubes were used see

Figure 6 Basic Plate Layout Design for Certification Measurements It was later determined that the homogeneity and stability data sets were obtained using a different set of pipettes than the certification data set and since all data was to be combined for certification another plate was obtained using the original pipette For the complete certification data see

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component Stability data was measured at different temperatures over several weeks using assays NEIF and ND6 Only the 4degC data was also used for certification For the complete stability data see

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

32 Model It was decided that all three sets of data should be used to derive the certified values In this way the uncertainty of the certified values would better reflect any potential heterogeneity in the samples as well as any additional unknown factors that may have effected the results Because two different pipettes were used to obtain different sets of data there was a significant bias in some of the measurements Specifically the ldquocertificationrdquo data from the first five plates was obtained using pipette set ldquo1rdquo and the remaining measurements were obtained using pipette set ldquo2rdquo This bias was introduced into the statistical model as part of the mean Possible effects of different assays and type of data (heterogeneity certification and stability) were accounted for in the statistical model in terms of a random effect Specifically the response for each component (A B C) was represented as

119910119910119894119894119894119894119894119894~1198731198731205741205741198941198941198941198941198941198941205901205901198941198942 where ~N() denotes ldquodistributed as a normal distribution of some mean and variancerdquo i = 123 denotes type of measurement heterogeneity certification and stability j = 1hellip10 denotes assay 2PR4 POTP NEIF R4Q5 D5 ND6 D9 HBB1 ND14 22C3 k = 1 2 denotes pipette set 120574120574119894119894119894119894119894119894 = 120572120572119894119894119894119894 + 120573120573119894119894 120572120572119894119894119894119894~119873119873(micro1198941198941205901205901205721205722) 120573120573119894119894 = 120573120573 for data type 1 and 120573120573119894119894 = 0 for data types 2 and 3 For a given 2372a component the variance 1205901205901205721205722 measures the between assay variability and 1205901205901198941198942 measures the repeatability variability for each data type The data was analyzed using Bayesian MCMC programmed in OpenBUGSC2 with non-informative Gaussian priors for the means and Gamma priors for the variance components

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

58

33 OpenBUGS code inits and data Certification code ModelBegin for(i in 13)mu[i]~dnorm(010E-5) sig[i]~dgamma(10E-510E-5) sigalph~dgamma(10E-510E-5) for(i in 110)alpha[1i]~dnorm(mu[1] sigalph) for(i in 12)alpha[2i]~dnorm(mu[2] sigalph) alpha[3i]~dnorm(mu[3]sigalph) b~dnorm(010E-5) beta[1]lt-b beta[2]lt-00 for(i in 1Ntot)mean[i]lt-alpha[typ[i]asy[i]]+beta[pip[i]] for(i in 1Ntot)lamb[i]~dnorm(mean[i]sig[typ[i]]) ModelEnd Certification Inits list(sig=c(111) sigalph=1) Certification Data for Component A list(Ntot=275) lamb[] asy[] cmp[] typ[] pip[] 1667 1 1 1 1 1655 2 1 1 1 1647 3 1 1 1 1594 4 1 1 1 1655 5 1 1 1 1629 6 1 1 1 1655 7 1 1 1 1644 8 1 1 1 1645 9 1 1 1 1545 10 1 1 1 1698 1 1 1 1 1685 2 1 1 1 1665 3 1 1 1 1658 4 1 1 1 1683 5 1 1 1 1647 6 1 1 1 1652 7 1 1 1 1611 8 1 1 1 1707 9 1 1 1 1603 10 1 1 1 1668 1 1 1 1 1587 2 1 1 1 1678 3 1 1 1 1628 4 1 1 1 1738 5 1 1 1 1667 6 1 1 1 1695 7 1 1 1 1757 8 1 1 1 1771 9 1 1 1 1762 10 1 1 1 1739 1 1 1 1 1702 2 1 1 1 1614 3 1 1 1 1762 4 1 1 1 1649 5 1 1 1 1678 6 1 1 1 1653 7 1 1 1

175 8 1 1 1 1764 9 1 1 1 1759 10 1 1 1 1638 1 1 1 1 1676 2 1 1 1 1616 3 1 1 1 1559 4 1 1 1 1608 5 1 1 1 1611 6 1 1 1 1517 7 1 1 1 1557 8 1 1 1 1604 9 1 1 1 1675 10 1 1 1 1667 1 1 1 1 1638 2 1 1 1 1624 3 1 1 1 1587 4 1 1 1 1652 5 1 1 1 1589 6 1 1 1 1607 7 1 1 1 1582 8 1 1 1 1625 9 1 1 1 1616 10 1 1 1 1888 1 1 1 1 1868 2 1 1 1 1953 3 1 1 1 1779 4 1 1 1 1924 5 1 1 1 1861 6 1 1 1 1824 7 1 1 1 1809 8 1 1 1 1864 9 1 1 1 193 10 1 1 1 1939 1 1 1 1 1815 2 1 1 1 1902 3 1 1 1 1871 4 1 1 1 2124 5 1 1 1

1787 6 1 1 1 1844 7 1 1 1 1799 8 1 1 1 1857 9 1 1 1 1942 10 1 1 1 1597 1 1 1 1 1533 2 1 1 1 1462 3 1 1 1 1493 4 1 1 1 1538 5 1 1 1 1492 6 1 1 1 1528 7 1 1 1 1548 8 1 1 1 1513 9 1 1 1 1562 10 1 1 1 159 1 1 1 1 1404 2 1 1 1 1533 3 1 1 1 1519 4 1 1 1 1586 5 1 1 1 1516 6 1 1 1 1538 7 1 1 1 1581 8 1 1 1 1514 9 1 1 1 1431 1 1 1 2 1407 2 1 1 2 1487 3 1 1 2 145 4 1 1 2 1394 5 1 1 2 1396 6 1 1 2 1414 7 1 1 2 1401 8 1 1 2 1412 9 1 1 2 1435 10 1 1 2 1411 1 1 1 2 1395 2 1 1 2 1398 3 1 1 2 1439 4 1 1 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

59

1422 5 1 1 2 1393 6 1 1 2 139 7 1 1 2 1382 8 1 1 2 146 9 1 1 2 141 10 1 1 2 1632 1 1 2 2 1531 1 1 2 2 1524 1 1 2 2 1454 1 1 2 2 1465 1 1 2 2 1512 1 1 2 2 1418 1 1 2 2 1496 1 1 2 2 1588 1 1 2 2 1574 1 1 2 2 1557 1 1 2 2 15 1 1 2 2 1672 1 1 2 2 1551 1 1 2 2 1573 1 1 2 2 1516 1 1 2 2 1612 1 1 2 2 1598 1 1 2 2 1575 1 1 2 2 1658 1 1 2 2 1547 1 1 2 2 1529 1 1 2 2 1525 1 1 2 2 154 1 1 2 2 1458 1 1 2 2 154 1 1 2 2 1491 1 1 2 2 148 1 1 2 2 156 1 1 2 2 1581 1 1 2 2 1513 1 1 2 2 1508 1 1 2 2 1476 1 1 2 2 1634 1 1 2 2 1568 1 1 2 2 152 1 1 2 2 1516 1 1 2 2 1535 1 1 2 2 1514 1 1 2 2 1568 1 1 2 2 1611 1 1 2 2 1528 1 1 2 2 1511 1 1 2 2 1549 1 1 2 2 1533 1 1 2 2 1427 1 1 2 2 1504 1 1 2 2 1491 1 1 2 2 1455 1 1 2 2

1501 1 1 2 2 1637 1 1 2 2 157 1 1 2 2 1633 1 1 2 2 1515 1 1 2 2 1804 1 1 2 2 1538 1 1 2 2 1617 1 1 2 2 1628 1 1 2 2 1672 1 1 2 2 1596 1 1 2 2 1586 1 1 2 2 1578 1 1 2 2 1532 1 1 2 2 1616 2 1 2 2 1615 2 1 2 2 1582 2 1 2 2 1627 2 1 2 2 1607 2 1 2 2 1511 2 1 2 2 1479 2 1 2 2 1514 2 1 2 2 1474 2 1 2 2 1594 2 1 2 2 1571 2 1 2 2 1626 2 1 2 2 149 2 1 2 2 1617 2 1 2 2 1634 2 1 2 2 1613 2 1 2 2 1529 2 1 2 2 1571 2 1 2 2 1652 2 1 2 2 1647 2 1 2 2 1537 2 1 2 2 164 2 1 2 2 1589 2 1 2 2 1152 2 1 2 2 1605 2 1 2 2 1633 2 1 2 2 1483 2 1 2 2 1557 2 1 2 2 1615 2 1 2 2 1438 2 1 2 2 1713 2 1 2 2 1611 2 1 2 2 1586 2 1 2 2 1516 2 1 2 2 1614 2 1 2 2 1627 2 1 2 2 1567 2 1 2 2 1495 2 1 2 2 1535 2 1 2 2 1545 2 1 2 2 1607 2 1 2 2

1513 2 1 2 2 1582 2 1 2 2 1601 2 1 2 2 1555 2 1 2 2 1607 2 1 2 2 1586 2 1 2 2 1524 2 1 2 2 1656 2 1 2 2 154 2 1 2 2 1454 2 1 2 2 1669 2 1 2 2 1631 2 1 2 2 1608 2 1 2 2 1497 2 1 2 2 1639 2 1 2 2 1575 2 1 2 2 1567 2 1 2 2 1528 2 1 2 2 1511 2 1 2 2 1627 2 1 2 2 1612 2 1 2 2 1536 2 1 2 2 1568 1 1 3 2 152 1 1 3 2 1486 1 1 3 2 1441 1 1 3 2 1706 1 1 3 2 144 1 1 3 2 15 1 1 3 2 1537 1 1 3 2 1387 1 1 3 2 1463 1 1 3 2 157 1 1 3 2 1429 1 1 3 2 1667 1 1 3 2 1632 1 1 3 2 1589 1 1 3 2 1474 2 1 3 2 1543 2 1 3 2 1555 2 1 3 2 1599 2 1 3 2 1494 2 1 3 2 1624 2 1 3 2 1577 2 1 3 2 1593 2 1 3 2 1479 2 1 3 2 1517 2 1 3 2 1538 2 1 3 2 159 2 1 3 2 1479 2 1 3 2 1479 2 1 3 2 1522 2 1 3 2 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

60

Table C-2 Example OpenBUGS Certification Summary for Component A

mean sd MC_error val25pc median val975pc start sample b 2584 02948 0005376 2001 2588 3158 100001 100000 mu[1] 1417 02706 0005366 1364 1416 147 100001 100000 mu[2] 1559 008368 4776E-4 1543 1559 1576 100001 100000 mu[3] 1533 01504 0001419 1504 1533 1563 100001 100000

Certification Data for Component B list(Ntot=278) lamb[] asy[] cmp[] typ[] pip[] 1869 1 2 1 1 185 2 2 1 1 1709 3 2 1 1 1856 4 2 1 1 1848 5 2 1 1 1828 6 2 1 1 1861 7 2 1 1 1825 8 2 1 1 1871 9 2 1 1 1776 10 2 1 1 196 1 2 1 1 1781 2 2 1 1 1707 3 2 1 1 1787 4 2 1 1 1829 5 2 1 1 1826 6 2 1 1 1858 7 2 1 1 1869 8 2 1 1 1773 9 2 1 1 1902 10 2 1 1 2174 1 2 1 1 2202 2 2 1 1 2172 3 2 1 1 223 5 2 1 1 2094 6 2 1 1 1949 7 2 1 1 1971 8 2 1 1 2123 9 2 1 1 2107 10 2 1 1 2231 1 2 1 1 2292 2 2 1 1 2216 3 2 1 1 2078 4 2 1 1 2192 5 2 1 1 2034 6 2 1 1 211 7 2 1 1 2123 8 2 1 1 2066 9 2 1 1 2046 10 2 1 1 181 1 2 1 1 1865 2 2 1 1 1806 3 2 1 1 1878 4 2 1 1 1824 5 2 1 1 1768 6 2 1 1 1811 7 2 1 1 1763 8 2 1 1 1865 9 2 1 1 1815 10 2 1 1

1806 1 2 1 1 1801 2 2 1 1 1786 3 2 1 1 1768 4 2 1 1 179 5 2 1 1 167 6 2 1 1 1749 7 2 1 1 1741 8 2 1 1 1781 9 2 1 1 1866 10 2 1 1 1784 1 2 1 1 1735 2 2 1 1 1778 3 2 1 1 1835 4 2 1 1 1794 5 2 1 1 17 6 2 1 1 1799 7 2 1 1 1765 8 2 1 1 1754 9 2 1 1 1792 10 2 1 1 1821 1 2 1 1 1813 2 2 1 1 1813 3 2 1 1 1749 4 2 1 1 1848 5 2 1 1 1711 6 2 1 1 1779 7 2 1 1 1769 8 2 1 1 1725 9 2 1 1 1814 10 2 1 1 1718 1 2 1 1 1753 2 2 1 1 1773 3 2 1 1 1777 4 2 1 1 1803 5 2 1 1 1718 6 2 1 1 1802 7 2 1 1 1834 8 2 1 1 1731 9 2 1 1 1842 10 2 1 1 1768 1 2 1 1 1698 2 2 1 1 1792 3 2 1 1 174 4 2 1 1 1795 5 2 1 1 1693 6 2 1 1 1743 7 2 1 1 1745 8 2 1 1 1714 9 2 1 1 1791 1 2 1 2

1746 2 2 1 2 1803 3 2 1 2 1767 4 2 1 2 1745 5 2 1 2 1824 6 2 1 2 1752 7 2 1 2 1813 8 2 1 2 1648 9 2 1 2 1818 10 2 1 2 177 1 2 1 2 182 2 2 1 2 1803 3 2 1 2 1739 4 2 1 2 1773 5 2 1 2 1729 6 2 1 2 1744 7 2 1 2 1754 8 2 1 2 1671 9 2 1 2 1741 10 2 1 2 179 1 2 2 2 1712 1 2 2 2 1693 1 2 2 2 1754 1 2 2 2 1757 1 2 2 2 1697 1 2 2 2 1722 1 2 2 2 1884 1 2 2 2 1713 1 2 2 2 1759 1 2 2 2 1749 1 2 2 2 1748 1 2 2 2 1632 1 2 2 2 1752 1 2 2 2 1871 1 2 2 2 1688 1 2 2 2 1789 1 2 2 2 1826 1 2 2 2 1751 1 2 2 2 1738 1 2 2 2 1742 1 2 2 2 1761 1 2 2 2 1909 1 2 2 2 1719 1 2 2 2 1755 1 2 2 2 1745 1 2 2 2 1687 1 2 2 2 1709 1 2 2 2 1625 1 2 2 2 1802 1 2 2 2 1734 1 2 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

61

1809 1 2 2 2 1782 1 2 2 2 1712 1 2 2 2 1668 1 2 2 2 1535 1 2 2 2 1835 1 2 2 2 1725 1 2 2 2 1769 1 2 2 2 178 1 2 2 2 1759 1 2 2 2 1678 1 2 2 2 1836 1 2 2 2 1743 1 2 2 2 1885 1 2 2 2 1704 1 2 2 2 1749 1 2 2 2 1739 1 2 2 2 1735 1 2 2 2 1714 1 2 2 2 1703 1 2 2 2 1772 1 2 2 2 1727 1 2 2 2 1738 1 2 2 2 174 1 2 2 2 1718 1 2 2 2 1701 1 2 2 2 1673 1 2 2 2 1887 1 2 2 2 1743 1 2 2 2 1825 1 2 2 2 1849 1 2 2 2 188 1 2 2 2 1768 1 2 2 2 1713 1 2 2 2 166 1 2 2 2 1775 2 2 2 2 1824 2 2 2 2 167 2 2 2 2 1683 2 2 2 2 1693 2 2 2 2 169 2 2 2 2 1864 2 2 2 2 1709 2 2 2 2 1764 2 2 2 2

1741 2 2 2 2 1797 2 2 2 2 1626 2 2 2 2 1724 2 2 2 2 177 2 2 2 2 1744 2 2 2 2 1716 2 2 2 2 1805 2 2 2 2 1827 2 2 2 2 1752 2 2 2 2 1746 2 2 2 2 1666 2 2 2 2 187 2 2 2 2 1718 2 2 2 2 1483 2 2 2 2 1632 2 2 2 2 1759 2 2 2 2 1723 2 2 2 2 1684 2 2 2 2 1887 2 2 2 2 1756 2 2 2 2 1816 2 2 2 2 1792 2 2 2 2 1614 2 2 2 2 1566 2 2 2 2 1741 2 2 2 2 1752 2 2 2 2 1697 2 2 2 2 1762 2 2 2 2 1801 2 2 2 2 1766 2 2 2 2 1747 2 2 2 2 1744 2 2 2 2 1692 2 2 2 2 186 2 2 2 2 1756 2 2 2 2 1896 2 2 2 2 1814 2 2 2 2 1703 2 2 2 2 1722 2 2 2 2 1688 2 2 2 2 1821 2 2 2 2 1758 2 2 2 2 1787 2 2 2 2

1784 2 2 2 2 179 2 2 2 2 1681 2 2 2 2 1685 2 2 2 2 1798 2 2 2 2 171 2 2 2 2 1638 2 2 2 2 1931 2 2 2 2 1786 2 2 2 2 1768 2 2 2 2 1764 2 2 2 2 1819 2 2 2 2 1666 1 2 3 2 1626 1 2 3 2 188 1 2 3 2 1653 1 2 3 2 1733 1 2 3 2 1551 1 2 3 2 1884 1 2 3 2 1718 1 2 3 2 1549 1 2 3 2 1723 1 2 3 2 194 1 2 3 2 164 1 2 3 2 1565 1 2 3 2 1787 1 2 3 2 1655 2 2 3 2 167 2 2 3 2 1673 2 2 3 2 1705 2 2 3 2 1755 2 2 3 2 1625 2 2 3 2 1708 2 2 3 2 1705 2 2 3 2 1862 2 2 3 2 1686 2 2 3 2 1723 2 2 3 2 1909 2 2 3 2 1859 2 2 3 2 1897 2 2 3 2 1972 2 2 3 2 END

Table C-3 Example OpenBUGS Certification Summary for Component B

mean sd MC_error val25pc median val975pc start sample b 09337 03387 0007504 02695 09371 1592 100001 100000 mu[1] 1764 03101 0007483 1704 1763 1825 100001 100000 mu[2] 1748 008014 4719E-4 1733 1748 1764 100001 100000 mu[3] 1735 02338 0003093 1688 1736 1781 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

62

Certification data for Component C list(Ntot=280) lamb[] asy[] cmp[] typ[] pip[] 163 1 3 1 1 1596 2 3 1 1 1574 3 3 1 1 159 4 3 1 1 161 5 3 1 1 1589 6 3 1 1 1569 7 3 1 1 1565 8 3 1 1 1564 9 3 1 1 1548 10 3 1 1 1706 1 3 1 1 1576 2 3 1 1 1538 3 3 1 1 1555 4 3 1 1 1607 5 3 1 1 1548 6 3 1 1 1572 7 3 1 1 1681 8 3 1 1 1492 9 3 1 1 1655 10 3 1 1 1499 1 3 1 1 1576 2 3 1 1 1545 3 3 1 1 1584 4 3 1 1 1577 5 3 1 1 1599 6 3 1 1 1587 7 3 1 1 152 8 3 1 1 1479 9 3 1 1 1512 10 3 1 1 1497 1 3 1 1 1611 2 3 1 1 156 3 3 1 1 1546 4 3 1 1 1529 5 3 1 1 1478 6 3 1 1 1504 7 3 1 1 1551 8 3 1 1 1438 9 3 1 1 1494 10 3 1 1 1452 1 3 1 1 1347 2 3 1 1 1395 3 3 1 1 1442 4 3 1 1 1448 5 3 1 1 1463 6 3 1 1 1402 7 3 1 1 1474 8 3 1 1 1466 9 3 1 1 1439 10 3 1 1 1424 1 3 1 1 1397 2 3 1 1 135 3 3 1 1 1459 4 3 1 1 1602 5 3 1 1 1435 6 3 1 1 1433 7 3 1 1 1481 8 3 1 1 1482 9 3 1 1

1568 10 3 1 1 147 1 3 1 1 1411 2 3 1 1 1512 3 3 1 1 1452 4 3 1 1 1567 5 3 1 1 148 6 3 1 1 1442 7 3 1 1 143 8 3 1 1 1413 9 3 1 1 1486 10 3 1 1 1444 1 3 1 1 1427 2 3 1 1 1439 3 3 1 1 1468 4 3 1 1 1524 5 3 1 1 1424 6 3 1 1 1459 7 3 1 1 1467 8 3 1 1 1411 9 3 1 1 148 10 3 1 1 1485 1 3 1 1 1422 2 3 1 1 1438 3 3 1 1 1439 4 3 1 1 1515 5 3 1 1 1409 6 3 1 1 1469 7 3 1 1 1453 8 3 1 1 1465 9 3 1 1 1363 10 3 1 1 1458 1 3 1 1 1459 2 3 1 1 1465 3 3 1 1 1468 4 3 1 1 1457 5 3 1 1 1412 6 3 1 1 1474 7 3 1 1 1428 8 3 1 1 1425 9 3 1 1 1393 10 3 1 1 1357 1 3 1 2 1447 2 3 1 2 1448 3 3 1 2 1409 4 3 1 2 1389 5 3 1 2 1415 6 3 1 2 1431 7 3 1 2 1381 8 3 1 2 1403 9 3 1 2 1435 10 3 1 2 1364 1 3 1 2 1361 2 3 1 2 1404 3 3 1 2 1411 4 3 1 2 1429 5 3 1 2 1417 6 3 1 2 138 7 3 1 2 1336 8 3 1 2 1387 9 3 1 2

1372 10 3 1 2 1493 1 3 2 2 1516 1 3 2 2 1451 1 3 2 2 144 1 3 2 2 1523 1 3 2 2 1437 1 3 2 2 1485 1 3 2 2 1433 1 3 2 2 1463 1 3 2 2 1534 1 3 2 2 1533 1 3 2 2 145 1 3 2 2 1537 1 3 2 2 1515 1 3 2 2 1511 1 3 2 2 1449 1 3 2 2 1388 1 3 2 2 1475 1 3 2 2 1468 1 3 2 2 1549 1 3 2 2 153 1 3 2 2 1442 1 3 2 2 1385 1 3 2 2 1513 1 3 2 2 1528 1 3 2 2 1486 1 3 2 2 1472 1 3 2 2 1493 1 3 2 2 1477 1 3 2 2 1405 1 3 2 2 1499 1 3 2 2 1458 1 3 2 2 1514 1 3 2 2 1525 1 3 2 2 1527 1 3 2 2 149 1 3 2 2 1568 1 3 2 2 1452 1 3 2 2 1421 1 3 2 2 1455 1 3 2 2 1525 1 3 2 2 1531 1 3 2 2 144 1 3 2 2 1487 1 3 2 2 151 1 3 2 2 1577 1 3 2 2 1506 1 3 2 2 152 1 3 2 2 1483 1 3 2 2 1522 1 3 2 2 1463 1 3 2 2 1478 1 3 2 2 1505 1 3 2 2 1454 1 3 2 2 1555 1 3 2 2 1497 1 3 2 2 1515 1 3 2 2 1494 1 3 2 2 1594 1 3 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

63

1471 1 3 2 2 1394 1 3 2 2 1445 1 3 2 2 1446 1 3 2 2 1499 1 3 2 2 1484 1 3 2 2 1436 1 3 2 2 1565 2 3 2 2 149 2 3 2 2 1485 2 3 2 2 1529 2 3 2 2 1513 2 3 2 2 1492 2 3 2 2 1475 2 3 2 2 1497 2 3 2 2 1503 2 3 2 2 1448 2 3 2 2 1488 2 3 2 2 1496 2 3 2 2 1499 2 3 2 2 1469 2 3 2 2 1515 2 3 2 2 1543 2 3 2 2 1452 2 3 2 2 1458 2 3 2 2 1435 2 3 2 2 1499 2 3 2 2 1499 2 3 2 2 1589 2 3 2 2 1562 2 3 2 2 1547 2 3 2 2 1505 2 3 2 2 1508 2 3 2 2 152 2 3 2 2

1481 2 3 2 2 1547 2 3 2 2 1501 2 3 2 2 1491 2 3 2 2 1465 2 3 2 2 147 2 3 2 2 1483 2 3 2 2 1479 2 3 2 2 148 2 3 2 2 1543 2 3 2 2 1448 2 3 2 2 1531 2 3 2 2 1478 2 3 2 2 1465 2 3 2 2 1454 2 3 2 2 1591 2 3 2 2 1466 2 3 2 2 157 2 3 2 2 1471 2 3 2 2 1506 2 3 2 2 1491 2 3 2 2 1461 2 3 2 2 1485 2 3 2 2 1495 2 3 2 2 1516 2 3 2 2 1474 2 3 2 2 1483 2 3 2 2 1519 2 3 2 2 1475 2 3 2 2 1466 2 3 2 2 1565 2 3 2 2 1491 2 3 2 2 1435 2 3 2 2 1459 2 3 2 2

1421 2 3 2 2 1468 2 3 2 2 1556 2 3 2 2 1424 1 3 3 2 1427 1 3 3 2 1381 1 3 3 2 1394 1 3 3 2 1473 1 3 3 2 133 1 3 3 2 1446 1 3 3 2 1424 1 3 3 2 133 1 3 3 2 1452 1 3 3 2 1512 1 3 3 2 1314 1 3 3 2 1525 1 3 3 2 1475 1 3 3 2 1542 1 3 3 2 1461 2 3 3 2 1443 2 3 3 2 1547 2 3 3 2 1394 2 3 3 2 1364 2 3 3 2 1395 2 3 3 2 1522 2 3 3 2 1483 2 3 3 2 1452 2 3 3 2 1482 2 3 3 2 1439 2 3 3 2 1458 2 3 3 2 155 2 3 3 2 1502 2 3 3 2 152 2 3 3 2 END

Table C-4 Example OpenBUGS Certification Summary for Component C

mean sd MC_error val25pc median val975pc start sample b 09626 01701 0002428 06285 09633 1293 100001 100000 mu[1] 1399 01566 0002413 1369 1399 143 100001 100000 mu[2] 1491 005701 2386E-4 148 1491 1503 100001 100000 mu[3] 1449 0132 0001181 1423 1449 1475 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

64

34 Results Table C-5 lists the results in CopiesnL for the three components

Table C-5 Parameter Values for Components A B and C

Posterior Distribution Summary Statistics Component Parameter Mean SD 25 50 975

A

β 260 029 203 260 318 μ1 1415 027 1362 1415 1467 μ2 1559 010 1539 1559 158 μ3 1533 016 1501 1533 1565

B

β 095 034 029 095 162 μ1 1762 031 1701 1762 1822 μ2 1748 010 1728 1748 1769 μ3 1735 024 1687 1735 1783

C

β 097 017 064 097 130 μ1 1399 016 1368 1399 1429 μ2 1491 008 1476 1491 1507 μ3 1449 014 1421 1449 1477

Since in each case the three data set types represent independent measurements the three estimates can be combined to obtain a consensus value In this way a between data set type variability is included in the final standard uncertainty Table C-6 lists the results in Copies per Nanoliter obtained using the NIST Consensus Builder (NICOB)C3

Table C-6 Certified Values for Components A B and C

Component Mean u(Mean) Uk=2(Mean) A 151 07 14 B 175 03 06 C 145 05 10

4 Mito-to-nuclear DNA ratios (mtDNAnDNA)

41 Data Measurements were also obtained on the same samples using three mitochondrial detection assays mtND1 mtBatz and mtKav and a single nDNA assay (genomic) to compute the ratios mtDNAnDNA For the complete mtDNAnDNA data see

Table 16 mtDNAnDNA Ratios

42 Model For each component the data was analyzed using the following random effects model to account for between assay variability

119903119903119894119894~119873119873120572120572119894119894 1205901205901198861198861198861198861198861198861198861198861198861198862

where 119894119894 denotes assay (for A and B 1 = mtND1 2 = mtBatz 3 = mtKav for C 1 = mtBatz 2 = mtKav) and 120572120572119894119894~119873119873micro120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119886119886119886119886119886119886119886119886119886119886

2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

65

43 OpenBUGS code inits and data Two versions of the same model are required one for the results from the three mtDNA assays available for components A and B mtND1 mtBatz and mtKav and the other for the results from the two mtDNA assays available for component C mtBatz and mtKav mtDNAnDNA code for components A and B ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 13)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for components A and B list(sigt=c(111)) Data for Component A list(Ndata=102) set[] rep[] asy[] rat[] 1 1 1 1931 1 2 1 187 1 1 2 1845 1 2 2 1764 2 1 1 181 2 2 1 1906 2 1 2 1785 2 2 2 1837 3 1 1 1743 3 2 1 1761 3 1 2 1829 3 2 2 1847 4 1 1 1716 4 2 1 180 4 1 2 174 4 2 2 1813 5 1 1 1923 5 2 1 1835 5 1 2 1935 5 2 2 1851 6 1 1 1757 6 2 1 1724 6 3 1 1768 6 1 2 1743 6 2 2 1671 6 3 2 1766 6 1 3 1692 6 2 3 1662 6 3 3 1767 7 1 1 1924 7 2 1 1944 7 3 1 1893 7 1 2 1786

7 2 2 1752 7 3 2 1743 7 1 3 153 7 2 3 1542 7 3 3 1523 7 1 2 158 7 2 2 1557 7 3 2 1592 7 1 3 2103 7 2 3 210 7 3 3 2049 7 1 1 1749 7 2 1 1769 7 3 1 1777 7 1 2 1965 7 2 2 1889 7 3 2 1947 7 1 3 1813 7 2 3 1864 7 3 3 1864 8 1 1 178 8 2 1 1776 8 3 1 1706 8 1 2 1804 8 2 2 1751 8 3 2 1738 8 1 3 1769 8 2 3 1786 8 3 3 1761 9 1 2 1584 9 2 2 1544 10 1 1 1947 10 2 1 1928 10 3 1 1922 10 1 1 1829

10 2 1 1895 10 3 1 1851 10 1 1 1674 10 2 1 1617 11 1 2 1667 11 2 2 1813 11 3 2 1708 11 4 2 1598 11 1 2 1566 11 2 2 1687 11 3 2 1663 11 4 2 1589 11 1 2 1643 11 2 2 1668 11 3 2 1604 11 4 2 1573 12 1 2 1719 12 2 2 1716 12 3 2 167 12 4 2 1666 12 1 2 1581 12 2 2 1573 12 3 2 1601 12 4 2 1612 12 1 2 1546 12 2 2 1542 12 3 2 1538 12 4 2 1587 13 1 1 1617 13 2 1 1599 13 1 1 1622 13 2 1 1579 13 1 1 1575 13 2 1 1528 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

66

Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A

mean sd MC_error val25pc median val975pc start sample mu 1730 1443 004419 1703 1730 1759 100001 100000

data for Component B list(Ndata=87) set[] rep[] asy[] rat[] 1 1 1 2461 1 2 1 2401 1 1 2 2555 1 2 2 2608 2 1 1 2229 2 2 1 2200 2 1 2 2108 2 2 2 2192 3 1 1 2127 3 2 1 2198 3 1 2 2455 3 2 2 2506 4 1 1 2211 4 2 1 2199 4 1 2 2380 4 2 2 2249 5 1 1 2105 5 2 1 1943 5 1 2 2136 5 2 2 2119 6 1 1 1990 6 2 1 1921 6 3 1 1982 6 1 2 1990 6 2 2 1924 6 3 2 1958 6 1 3 1903 6 2 3 1831

6 3 3 1919 7 1 1 2304 7 2 1 2341 7 3 1 2332 7 1 2 2048 7 2 2 2078 7 3 2 2067 7 1 3 2399 7 2 3 2479 7 3 3 2493 7 1 1 2065 7 2 1 2166 7 3 1 2164 7 1 2 2314 7 2 2 2280 7 3 2 2258 7 1 3 1867 7 2 3 1987 7 3 3 1971 8 1 1 2046 8 2 1 2057 8 3 1 2024 8 1 2 1998 8 2 2 2043 8 3 2 1952 8 1 3 2077 8 2 3 2117 8 3 3 2041 9 1 2 1890 9 2 2 1886

10 1 2 1928 10 2 2 1916 10 3 2 1915 10 1 2 1947 10 2 2 1946 10 3 2 1961 10 4 2 1909 10 1 2 1953 10 2 2 1952 10 3 2 1941 10 4 2 1881 11 1 2 1865 11 2 2 1863 11 3 2 1828 11 4 2 1828 11 1 2 1944 11 2 2 1964 11 3 2 1947 11 4 2 1895 11 1 2 1908 11 2 2 1909 11 3 2 1940 11 4 2 1839 12 1 1 1688 12 2 1 1783 12 1 1 1823 12 2 1 1631 12 1 1 1920 12 2 1 1887 END

Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B

mean sd MC_error val25pc median val975pc start sample mu 2045 2246 0111 1999 2045 2088 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

67

mtDNAnDNA code for component C ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 12)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for component C list(sigt=c(11)) mtDNAnDNA data for component C list(Ndata=66) set[] rep[] asy[] rat[] 1 1 1 3201 1 2 1 3299 2 1 1 3347 2 2 1 3287 3 1 1 3217 3 2 1 3384 4 1 1 3200 4 2 1 3363 5 1 1 3294 5 2 1 3013 6 1 1 2805 6 2 1 2825 6 3 1 2829 6 1 2 2709 6 2 2 2738 6 3 2 2727 7 1 1 2797 7 2 1 2642 7 3 1 2792 7 1 2 2804 7 2 2 2789

7 3 2 2842 7 1 1 2715 7 2 1 2606 7 3 1 2687 7 1 2 2799 7 2 2 2628 7 3 2 2559 7 1 1 2639 7 2 1 2535 7 3 1 2657 7 1 2 2953 7 2 2 2757 7 3 2 2802 8 1 1 2927 8 2 1 2776 8 3 1 2762 8 1 2 3044 8 2 2 2830 8 3 2 2825 9 1 1 2501 9 2 1 2522 10 1 1 2832 10 2 1 2840

10 3 1 2788 10 4 1 2725 10 1 1 2835 10 2 1 2790 10 3 1 2736 10 4 1 2679 10 1 1 2803 10 2 1 2773 10 3 1 2757 10 4 1 2752 11 1 1 2599 11 2 1 2508 11 3 1 2551 11 4 1 2609 11 1 1 2617 11 2 1 2483 11 3 1 2626 11 4 1 2602 11 1 1 2665 11 2 1 2535 11 3 1 2570 11 4 1 2650 END

Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C

mean sd MC_error val25pc median val975pc start sample mu 2774 2367 01176 2729 2774 2821 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

68

Table C-10 Results for mtDNAnDNA Posterior Distribution Summary Statistics

Component Mean Std Dev 25 50 975 A 1730 14 1703 1730 1758 B 2045 23 1994 2047 2087 C 2774 23 2729 2774 2821

5 ReferencesC1 Lunn DJ Spiegelhalter D Thomas A Best N Statistics in Medicine 2009283049ndash3082C2 Polson NG Scott JG Bayesian Analysis 20127887ndash902C3 NIST Consensus Builder Software and userrsquos guide freely available at

httpsconsensusnistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

69

Appendix D Performance in Commercial qPCR Chemistries The performance of the SRM 2372a materials in commercial qPCR chemistries was assessed using manufacturerrsquos recommended protocols Each panel in Figures D-1 to D-4 displays the average cycle threshold (Ct) as a function of DNA concentration for the 5-point dilutions of SRM 2372a components In all cases the external calibrant was SRM 2372 component A Error bars represent approximate 95 confidence intervals Table D-1 lists the estimated [DNA] for the three SRM 2372a components for all commercial assays evaluated This Table also lists the Component BComponent A (BA) and Component CComponent A (CA) ratios calculated from the [DNA] estimates

Figure D-1 Results for the Innogenomics Commercial qPCR assays

20

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantLong(IQ_h)

ABCStd

[A] = (384 plusmn 26) ngμL[B] = (441 plusmn 24) ngμL[C] = (516 plusmn 24) ngμL

Rsup2 = 0999 se(Ct) = 01016

17

18

19

20

21

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantShort(IQ_d)

ABCStd

[A] = (435 plusmn 28) ngμL[B] = (543 plusmn 12) ngμL[C] = (510 plusmn 28) ngμL

Rsup2 = 0999 se(Ct) = 007

24

25

26

27

28

29

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYLong(IQHY_h)

ABCStd

[A] = (387 plusmn 24) ngμL[B] = (448 plusmn 21) ngμL[C] = (559 plusmn 31) ngμL

Rsup2 = 0998 se(Ct) = 01314

16

18

20

22

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYShort(IQHY_d)

ABCStd

[A] = (353 plusmn 23) ngμL[B] = (267 plusmn 15) ngμL[C] = (390 plusmn 18) ngμL

Rsup2 = 1000 se(Ct) = 00420

22

24

26

28

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYY(IQHY_y)

ABCStd

[A] = (572 plusmn 13) ngμL[B] = 0 ngμL[C] = (137 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

70

Figure D-2 Results for the Thermo Fisher Commercial qPCR assays

21

22

23

24

25

26

27

28

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HumanHuman(QFH_h)

ABCStd

[A] = (464 plusmn 13) ngμL[B] = (529 plusmn 19) ngμL[C] = (472 plusmn 22) ngμL

Rsup2 = 1000 se(Ct) = 004

22

23

24

25

26

27

28

29

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoHuman(QFD_h)

ABCStd

[A] = (701 plusmn 53) ngμL[B] = (561 plusmn 11) ngμL[C] = (472 plusmn 11) ngμL

Rsup2 = 0994 se(Ct) = 01922

24

26

28

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoMale(QFD_y)

ABCStd

[A] = (706 plusmn 49) ngμL[B] = 0 ngμL[C] = (122 plusmn 02) ngμL

Rsup2 = 0999 se(Ct) = 007

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPLarge(QFHP_h)

ABCStd

[A] = (340 plusmn 18) ngμL[B] = (432 plusmn 05) ngμL[C] = (342 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 00720

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPSmall(QFHP_d)

ABCStd

[A] = (434 plusmn 15) ngμL[B] = (514 plusmn 10) ngμL[C] = (466 plusmn 07) ngμL

Rsup2 = 1000 se(Ct) = 005

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioLarge(QFT_h)

ABCStd

[A] = (362 plusmn 23) ngμL[B] = (430 plusmn 06) ngμL[C] = (378 plusmn 07) ngμL

Rsup2 = 0999 se(Ct) = 00920

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioSmall(QFT_d)

ABCStd

[A] = (468 plusmn 08) ngμL[B] = (505 plusmn 08) ngμL[C] = (523 plusmn 14) ngμL

Rsup2 = 1000 se(Ct) = 00318

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioY(QFT_y)

ABCStd

[A] = (372 plusmn 11) ngμL[B] = 0 ngμL[C] = (110 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 003

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

71

Figure D-3 Results for the Qiagen Commercial qPCR assays

Figure D-4 Results for the Promega Commercial qPCR assays

19

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresAutosomal(QPH_h)

ABCStd

[A] = (554 plusmn 21) ngμL[B] = (479 plusmn 05) ngμL[C] = (381 plusmn 09) ngμL

Rsup2 = 1000 se(Ct) = 00418

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresY(QPH_y)

ABCStd

[A] = (358 plusmn 10) ngμL[B] = 0 ngμL[C] = (118 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 007

16

17

18

19

20

21

22

23

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProHuman(QPP_h)

ABCStd

[A] = (525 plusmn 49) ngμL[B] = (480 plusmn 42) ngμL[C] = (467 plusmn 45) ngμL

Rsup2 = 0999 se(Ct) = 01019

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProDegradation(QPP_d)

ABCStd

[A] = (567 plusmn 18) ngμL[B] = (558 plusmn 08) ngμL[C] = (484 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 00410

15

20

25

30

35

-50 -40 -30 -20 -10 00 10

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProY(QPP_y)

ABCStd

[A] = (538 plusmn 27) ngμL[B] = (00 plusmn 00) ngμL[C] = (162 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 003

16

18

20

22

24

-05 00 05 10 15 20

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorAutosomal(P_h)

ABCStd

[A] = (759 plusmn 64) ngμL[B] = (775 plusmn 24) ngμL[C] = (1056 plusmn 17) ngμL

Rsup2 = 0995 se(Ct) = 01818

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorY(P_y)

ABCStd

[A] = (596 plusmn 51) ngμL[B] = 0 ngμL[C] = (117 plusmn 01) ngμL

Rsup2 = 0994 se(Ct) = 020

18

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantAutosomal(PQ_h)

ABCStd

[A] = (436 plusmn 13) ngμL[B] = (606 plusmn 20) ngμL[C] = (504 plusmn 10) ngμL

Rsup2 = 1000 se(Ct) = 00218

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantDegradation(PQ_d)

ABCStd

[A] = (420 plusmn 18) ngμL[B] = (529 plusmn 16) ngμL[C] = (530 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 00516

18

20

22

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantY(PQ_y)

ABCStd

[A] = (410 plusmn 20) ngμL[B] = 0 ngμL[C] = (139 plusmn 05) ngμL

Rsup2 = 0999 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

72

Table D-1 Summary [DNA] Results for the Commercial qPCR Assays Estimated [DNA] ngμL Ratios

A B C BA CA Assaya xb u(x)c xb u(x)c xb u(x)c yd u(y)e yd u(y)e

IQd 435 28 543 12 510 28 125 008 117 010 IQh 384 26 441 24 516 24 115 010 134 011

IQHYd 353 23 267 15 390 18 076 006 111 009 IQHYh 387 24 448 21 559 31 116 009 144 012

Ph 759 64 775 24 1056 16 102 009 139 012 PQd 420 18 529 16 530 06 126 007 126 006 PQh 436 13 606 20 504 10 139 006 116 004

QFDh 701 53 561 11 472 11 080 006 067 005 QFHh 464 13 529 19 472 22 114 005 102 006

QFHPd 434 15 514 10 466 07 118 005 107 004 QFHPh 340 18 432 05 342 03 127 007 101 005

QFTd 468 08 505 08 523 14 108 003 112 004 QFTh 362 23 430 06 378 07 118 008 104 007 QPHh 554 21 479 05 380 09 086 003 069 003 QPPd 567 18 558 08 484 11 098 003 085 003 QPPh 525 49 480 42 467 45 092 012 089 012

IQHYy 572 13 NDf 137 06

NRg

024 001 Py 596 51 NDf 117 01 020 002

PQy 410 20 NDf 139 05 034 002 QFDy 706 49 NDf 122 02 017 001 QFTy 372 11 NDf 110 06 030 002 QPHy 358 10 NDf 118 03 033 001 QPPy 538 27 0002 0001 162 11 030 003

a Assay codes are listed in Table 8 b Mean of dilution-adjusted results from five dilutions of the SRM 2372a component c Standard uncertainty of the mean estimated from the stdanrd deviation of the five dilutions d Ratio of the estimated [DNA] of component B or C relative to that of component A e Standard uncertainty of the ratio estimated from the standard uncertainties of the individual components f Not detected Component B was prepared using tissue from only female donors it is not expected to

respond to male-specific assays g Not relevant

  • Abstract
  • Keywords
  • Technical Information Contact for this SRM
  • Table of Contents
  • Table of Tables
  • Table of Figures
  • Purpose and Description
  • Warning SRM 2372a is a Human Source Material
  • Storage and Use
  • History and Background
    • Figure 1 Sales History of SRM 2372
      • Experimental Methods
        • Polymerase Chain Reaction Assays
          • Table 1 NIST-Developed Human Nuclear DNA Assays
          • Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays
          • Table 2 NIST-Optimized Human Mitochondrial DNA Assays
          • Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays
            • Droplet Digital Polymerase Chain Reaction Measurements
              • Nuclear DNA (nDNA)
                • Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays
                  • Mitochondrial DNA (mtDNA)
                    • Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays
                    • Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays
                        • Sample Preparation
                        • Absorbance Measurements
                        • SRM Unit Production
                        • Component Homogeneity
                          • Figure 4 Plate Layout Design for Homogeneity Measurements
                            • Component Stability
                              • Figure 5 Plate Layout Design for Stability Measurements
                                • Component Certification Measurements
                                  • Table 6 Tubes used for Certification Measurements
                                  • Figure 6 Basic Plate Layout Design for Certification Measurements
                                    • Mitochondrial DNA Measurements
                                      • Table 7 Run Parameters for Mitochondrial DNA
                                        • Performance in Commercial qPCR Chemistries
                                          • Table 8 Commercial qPCR Chemistries Tested with SRM 2372a
                                          • Figure 7 Plate Layout Design for Performance Assessments
                                              • Measurement Results
                                                • Absorbance Spectrophotometry
                                                  • Table 9 Absorbance at Selected Wavelengths
                                                  • Figure 8 Absorbance Spectra of SRM 2372a Components A B and C
                                                    • Droplet Volume
                                                    • Certification Measurements
                                                      • Homogeneity Results
                                                        • Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component
                                                        • Table 11 Within- and Between-Tube Relative Standard Deviations
                                                          • Stability Results
                                                            • Figure 9 ddPCR Stability Measurements as Function of Time
                                                            • Table 12 Stability Data as λacute Copy Number per Nanoliter of Component
                                                              • Ten-Assay Certification Results
                                                                • Figure 10 Comparison of Assay Performance
                                                                • Table 13 Certification Data as λacute Copy Number per Nanoliter of Component
                                                                  • Estimation of Copy Number Concentration
                                                                    • Table 14 Measured Copy Number Concentrations
                                                                      • Conversion of Copy Number to Mass Concentration
                                                                        • Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution
                                                                            • Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)
                                                                              • Figure 11 mtDNAnDNA for Components A B and C
                                                                              • Table 16 mtDNAnDNA Ratios
                                                                              • Table 17 mtDNAnDNA Summary Results
                                                                                • Performance in Commercial qPCR Chemistries
                                                                                  • Figure 12 qPCR Autosomal Quantitation Performance Summary
                                                                                  • Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary
                                                                                      • Qualification of ddPCR as Fit-For-Purpose
                                                                                        • Figure 14 Absorbance Spectra of SRM 2372 Components A and B
                                                                                        • Digital Polymerase Chain Reaction Assays
                                                                                        • Heat Treatments (Melting of the DNA into Single Strands)
                                                                                        • Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a
                                                                                          • Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a
                                                                                            • Quantitative Evaluation of ssDNA Content
                                                                                              • ddPCR Measurement of ssDNA Proportion
                                                                                              • Proof of Principle
                                                                                                • Table 19 Comparison of Snap-Cooling to Slow-Cooling
                                                                                                  • Time at Denaturation Temperature
                                                                                                    • Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures
                                                                                                      • Temperature
                                                                                                        • Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a
                                                                                                          • DNA Concentration in the ddPCR Solution
                                                                                                            • Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution
                                                                                                              • Equivalence of Assays
                                                                                                                • Table 23 Comparison of Assays
                                                                                                                    • ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials
                                                                                                                      • Table 24 Evaluation of Limiting Ratio
                                                                                                                          • Conclusions and Recommendations
                                                                                                                            • Recommended Certification Values
                                                                                                                              • Table 25 Recommended Values and 95 Uncertainties for Certification
                                                                                                                                • Use of SRM 2372a for Value-Assigning In-House Reference Materials
                                                                                                                                  • Certificate of Analysis
                                                                                                                                  • References
                                                                                                                                  • Appendix A DNA Extraction Method
                                                                                                                                    • Procedure
                                                                                                                                    • Reagents
                                                                                                                                    • Reference
                                                                                                                                      • Appendix B Spectrophotometric Measurement Protocol
                                                                                                                                        • Spectrophotometer Calibration
                                                                                                                                          • Establish Baseline
                                                                                                                                          • Enable Verification of Wavelength Calibration
                                                                                                                                          • Enable Verification of Absorbance Calibration
                                                                                                                                            • For ldquonativerdquo dsDNA Samples
                                                                                                                                              • Verify Optical Balance of Cuvettes
                                                                                                                                                • For NaOH denatured ssDNA Samples
                                                                                                                                                  • Prepare a 04 molL NaOH Stock Solution
                                                                                                                                                  • Prepare the Sample Solutions
                                                                                                                                                  • Measure Sample Solutions
                                                                                                                                                    • After All Sample Measurements are Complete
                                                                                                                                                      • Enable Verification of Spectrophotometer Stability
                                                                                                                                                      • Clean Up
                                                                                                                                                        • Evaluation
                                                                                                                                                        • Reference
                                                                                                                                                          • Appendix C SRM 2372a Statisticianrsquos report
                                                                                                                                                            • 1 Introduction
                                                                                                                                                            • 11 OpenBUGS Analysis
                                                                                                                                                            • 12 OpenBUGS Exemplar Output
                                                                                                                                                            • 2 Homogeneity assessment
                                                                                                                                                            • 21 Data
                                                                                                                                                            • 22 Model
                                                                                                                                                            • 23 OpenBUGS code and data
                                                                                                                                                              • Table C-1 Example OpenBUGS Homogeneity Summary
                                                                                                                                                                • 24 Results
                                                                                                                                                                  • Figure C-1 Homogeneity Assessment
                                                                                                                                                                    • 3 Certification measurements
                                                                                                                                                                    • 31 Data
                                                                                                                                                                    • 32 Model
                                                                                                                                                                    • 33 OpenBUGS code inits and data
                                                                                                                                                                      • Table C-2 Example OpenBUGS Certification Summary for Component A
                                                                                                                                                                      • Table C-3 Example OpenBUGS Certification Summary for Component B
                                                                                                                                                                      • Table C-4 Example OpenBUGS Certification Summary for Component C
                                                                                                                                                                        • 34 Results
                                                                                                                                                                          • Table C-5 Parameter Values for Components A B and C
                                                                                                                                                                          • Table C-6 Certified Values for Components A B and C
                                                                                                                                                                            • 4 Mito-to-nuclear DNA ratios (mtDNAnDNA)
                                                                                                                                                                            • 41 Data
                                                                                                                                                                            • 42 Model
                                                                                                                                                                            • 43 OpenBUGS code inits and data
                                                                                                                                                                              • Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A
                                                                                                                                                                              • Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B
                                                                                                                                                                              • Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C
                                                                                                                                                                              • Table C-10 Results for mtDNAnDNA
                                                                                                                                                                                • 5 References
                                                                                                                                                                                  • Appendix D Performance in Commercial qPCR Chemistries
                                                                                                                                                                                    • Table D-1 Summary [DNA] Results for the Commercial qPCR Assays
Page 6: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

iii

Appendix A DNA Extraction Method 46 Procedure 46 Reagents 46 Reference 47

Appendix B Spectrophotometric Measurement Protocol 48 Spectrophotometer Calibration 48

Establish Baseline 48 Enable Verification of Wavelength Calibration 48 Enable Verification of Absorbance Calibration 48

For ldquonativerdquo dsDNA Samples 48 Verify Optical Balance of Cuvettes 48

For NaOH denatured ssDNA Samples 49 Prepare a 04 molL NaOH Stock Solution 49 Prepare the Sample Solutions 49 Measure Sample Solutions 50

After All Sample Measurements are Complete 50 Enable Verification of Spectrophotometer Stability 50 Clean Up 50

Evaluation 50 Appendix C SRM 2372a Statisticianrsquos report 51

1 Introduction 51 11 OpenBUGS Analysis 51 12 OpenBUGS Exemplar Output 51 2 Homogeneity assessment 52 21 Data 52 22 Model 52 23 OpenBUGS code and data 52 24 Results 55 3 Certification measurements 57 31 Data 57 32 Model 57 33 OpenBUGS code inits and data 58 34 Results 64 4 Mito-to-nuclear DNA ratios (mtDNAnDNA) 64 41 Data 64 42 Model 64 43 OpenBUGS code inits and data 65 5 References 68

Appendix D Performance in Commercial qPCR Chemistries 69

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

iv

Table of Tables Table 1 NIST-Developed Human Nuclear DNA Assays 5 Table 2 NIST-Optimized Human Mitochondrial DNA Assays 6 Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays 8 Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays 8 Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays 8 Table 6 Tubes used for Certification Measurements 12 Table 7 Run Parameters for Mitochondrial DNA 13 Table 8 Commercial qPCR Chemistries Tested with SRM 2372a 14 Table 9 Absorbance at Selected Wavelengths 15 Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component 17 Table 11 Within- and Between-Tube Relative Standard Deviations 18 Table 12 Stability Data as λacute Copy Number per Nanoliter of Component 19 Table 13 Certification Data as λacute Copy Number per Nanoliter of Component 21 Table 14 Measured Copy Number Concentrations 22 Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution 23 Table 16 mtDNAnDNA Ratios 24 Table 17 mtDNAnDNA Summary Results 25 Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a 29 Table 19 Comparison of Snap-Cooling to Slow-Cooling 32 Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures 34 Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 36 Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution 37 Table 23 Comparison of Assays 39 Table 24 Evaluation of Limiting Ratio 41 Table 25 Recommended Values and 95 Uncertainties for Certification 43 Table C-1 Example OpenBUGS Homogeneity Summary 55 Table C-2 Example OpenBUGS Certification Summary for Component A 60 Table C-3 Example OpenBUGS Certification Summary for Component B 61 Table C-4 Example OpenBUGS Certification Summary for Component C 63 Table C-5 Parameter Values for Components A B and C 64 Table C-6 Certified Values for Components A B and C 64 Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A 66 Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B 66 Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C 67 Table C-10 Results for mtDNAnDNA 68 Table D-1 Summary [DNA] Results for the Commercial qPCR Assays 72

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

i

Table of Figures Figure 1 Sales History of SRM 2372 3 Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays 6 Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays 7 Figure 4 Plate Layout Design for Homogeneity Measurements 11 Figure 5 Plate Layout Design for Stability Measurements 11 Figure 6 Basic Plate Layout Design for Certification Measurements 12 Figure 7 Plate Layout Design for Performance Assessments 14 Figure 8 Absorbance Spectra of SRM 2372a Components A B and C 16 Figure 9 ddPCR Stability Measurements as Function of Time 18 Figure 10 Comparison of Assay Performance 20 Figure 11 mtDNAnDNA for Components A B and C 24 Figure 12 qPCR Autosomal Quantitation Performance Summary 26 Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary 26 Figure 14 Absorbance Spectra of SRM 2372 Components A and B 27 Figure 15 Comparison of Snap-Cooling to Slow-Cooling 32 Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures 33 Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 35 Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution 37 Figure 19 Comparison of Assays 38 Figure 20 Evaluation of Limiting Ratio 40 Figure 21 Degrees of Equivalence 42 Figure C-1 Homogeneity Assessment 56 Figure D-1 Results for the Innogenomics Commercial qPCR assays 69 Figure D-2 Results for the Thermo Fisher Commercial qPCR assays 70 Figure D-3 Results for the Qiagen Commercial qPCR assays 70 Figure D-4 Results for the Promega Commercial qPCR assays 71

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

1

Purpose and Description Standard Reference Material (SRM) 2372a is intended for use in the value assignment of human genomic deoxyribonucleic acid (DNA) forensic quantitation materials A unit of SRM 2372a consists of three well-characterized human genomic DNA materials solubilized in 10 mmolL 2-amino-2-(hydroxymethyl)-13 propanediol hydrochloride (Tris HCl) and 01 mmolL ethylenediaminetetraacetic acid disodium salt (disodium EDTA) pH 80 buffer (TE-4) The three component genomic DNA materials labeled A B and C are respectively derived from a single male donor a single female donor and 13 mixture of a male and a female donor A unit of the SRM consists of one 05 mL tube of each component each tube containing approximately 55 microL of DNA solution Each of these tubes is labeled and is sealed with a color-coded screw cap NIST is guided by and adheres to the ethical principles set forth in the Belmont Report1 SRM 2372a was developed after an appropriate human subject research determination by the NIST Human Subjects Protections Office

Warning SRM 2372a is a Human Source Material The three SRM 2372a components are human genomic DNA extracts prepared at NIST from buffy coat white blood cells from single-source anonymous donors All buffy coats were purchased from Interstate Blood Bank (Memphis TN) The supplier reported that each donor unit used in the preparation of this product was tested by FDA-licensed tests and found to be negative for human immunodeficiency virus (HIV-1 RNA and Anti-HIV 12) hepatitis B (HBV DNA and HBsAg) hepatitis C (HCV RNA and Anti-HCV) West Nile virus (WNV RNA) and syphilis Additionally each donor was tested according to FDA guidelines for Trypanosoma cruzi (Chagas) and found to be negativenon-reactive However no known test method can offer complete assurance that infectious agents are absent from this material Accordingly this human blood-based product should be handled at the Biosafety Level 12 SRM 2372a components and derived solutions should be disposed of in accordance with local state and federal regulations

Storage and Use Until required for use SRM 2372a should be stored in the dark between 2 degC and 8 degC The SRM 2372a component tubes should be mixed briefly and centrifuged (without opening the tube cap) prior to removing sample aliquots for analysis For the certified values to be applicable materials should be withdrawn immediately after opening the tubes and processed without delay Certified values do not apply to any material remaining in recapped tubes The certification only applies to the initial use and the same results are not guaranteed if the remaining material is used later

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

2

History and Background SRM 2372a is the second member of the SRM 2372 Human DNA Quantitation Standard series The supply of the initial material SRM 2372 was exhausted in May 2017 SRM 2372a is derived from commercially obtained buffy coat cells from different donors than those used in SRM 2372 A series of interlaboratory studies coordinated by what was then the Biotechnology Division of the National Institute of Standards and Technology (NIST) during the 1990s demonstrated that the DNA quantitation technologies then in use by the forensic human identification community adequately met the needs of the DNA typing tools then available34 However our 2004 study demonstrated that the increasingly sensitive and informative polymerase chain reaction (PCR)-based multiplex typing tools required more reliable determinations of DNA input quantity for best performance5 Quantitative PCR (qPCR) methods for DNA quantitation require calibration to an external standard of defined DNA concentration ([DNA]) expressed in units of nanograms of DNA per microliter of extract solution The true [DNA] of a commercial secondary standard may vary from its nominal concentration depending on supplier and lot number To enable providers of secondary standards to more reliably evaluate their materials and forensic testing laboratories to validate their use of commercial quantitation methods and supplies in 2007 we delivered the human DNA calibration standard SRM 2372 Human DNA Quantitation Standard6 The quantity of DNA in each of the three components of SRM 2372 was originally assigned based on the absorbance of double-stranded DNA (dsDNA) at 260 nm7 However by 2012 the absorbance properties of the materials had increased beyond the certified uncertainty limits due to slow but progressive changes in DNA tertiary structure Sales of the SRM were restricted until the DNA quantities were reassigned based on measurements of single-stranded DNA (ssDNA) in early 20138-10 Figure 1 displays the sales history of this relatively high-selling SRM While SRM 2372 was fit for the practical purpose of harmonizing DNA quantitation measurements to an accessible and reproducible reference material the conversion of both dsDNA and ssDNA absorbance to mass [DNA] is through conventional proportionality constants that are not metrologically traceable to the modern reference system for measurements the International System of Units (SI) Responding to requests from the forensic and clinical communities for calibration standards that provide traceability to the SI we and others have established droplet digital polymerase chain reaction (ddPCR) as a potentially primary measurement process that can provide traceable results11-18 The copy numbers of human nuclear DNA (nDNA) in the reaction mixture per partition symbolized as λ were determined by ddPCR using ten PCR assays optimized for use with our ddPCR platform These assays target loci on eight chromosomes with three assays on widely separated locations on one chromosome (see Table 1)

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

3

Figure 1 Sales History of SRM 2372

Each small black dot represents the sale of one unit of SRM 2372 The red lines summarize the average rate of sales as originally certified (2007 to 2012) and after re-certification (2013 to 2017) The magenta labels state the average yearly sales rate during these two periods

The nDNA copy number concentration human haploid genome equivalents (HHGE) per nanoliter component solution and symbolized here as λacute is calculated as

copy number [DNA] = λ = λ(VF) HHGEnL

where V is the average droplet volume and F is the total dilution factor the volume fraction of the component solution in the ddPCR reaction mixture The manufacturer of the ddPCR platform used a Bio-Rad QX200 Droplet Digital PCR System (Hercules CA) system does not provide metrologically traceable droplet volume information This lack of information prevented metrologically traceable conversion of the number per partition measurements to the desired number per reaction volume units In consequence NIST developed a measurement method that provides metrologically traceable droplet volumes19 We show elsewhere that the mass [DNA] nanograms of human DNA per microliter component can be traceably estimated as

mass [DNA] = 3031timesλacute ngmicroL

where the constant is the product of the number of nucleotide basepairs (bp) in the human reference genome the average bp molecular weight Avogadrorsquos number of bp per mole and

2372

2372(Recertif ied)

050

010

00

1713 unitsyear

1416 unitsyear

2007 2009 2011 2013 2015 2017

Uni

ts S

old

Calendar Year

Human DNA Quantitation StandardSRM 2372

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

4

unit scaling factors18 The three components of SRM 2372a are certified for both copy number and mass [DNA] Mitochondrial DNA (mtDNA) analysis within the forensic community is useful when the nDNA may be degraded or is in low-to-undetectable quantity The challenges of mtDNA testing include contamination of the standard due to routine laboratory use degradation of plasmids or synthetic oligomers or improper commutability of the standard relative to casework samples Commonly cell line materials are used as standards for qPCR with human genomic DNA which may lead to an incorrect mtDNA-to-nDNA ratio (mtDNAnDNA) for unknown samples Non-certified values of mtDNAnDNA for the three components are provided to bridge the current gap between well characterized mitochondrial DNA quantification assays and the availability of commercial standards With the understanding that the forensic DNA community primarily uses qPCR as a method for DNA quantification the performance of the three SRM 2372a materials was evaluated in eleven commercial qPCR chemistries currently used employed within the forensic DNA community

Experimental Methods Production and value-assignment of any certified reference material (CRM) is a complex process In addition to acquiring processing and packaging the materials included in the CRM the materials must be sufficiently homogenous across the units stable over time within its packaging and well-characterized to deliver results that are fit-for-purpose within the intended measurement community20 The process has been particularly complicated for SRM 2372a because we needed to determine the metrological properties of the required ddPCR measurement systems before we could trust the results they provide The following sections describe the methods and processes used to produce and characterize the SRM 2372a materials including sample preparation SRM unit production PCR assays ddPCR measurement processes spectrophotometric absorbance measurements homogeneity testing stability testing ddPCR nDNA and mtDNAnDNA measurements and performance in commercial qPCR assessments Many of these processes were mutually dependent therefore the sections are presented in a ldquodefinitionalrdquo order that sequentially introduces terminology Polymerase Chain Reaction Assays

Ten PCR assays were developed for characterization of SRM 2372a Each assay was confirmed to target one locus per HHGE using the National Center for Biotechnology Informationrsquos BLASTblastn system21 Primers were purchased from Eurofins Operon (Huntsville AL) TaqMan probes labeled with 6-carboxyfluorescein (FAM) were purchased from Thermo Fisher (Waltham MA) FAM-labeled Blackhole Quencher and Blackhole Quencher+ probes were purchased from LGC Biosearch Technologies (Novato CA) Table 1 details the ten nDNA assays used to certify copy number and mass [DNA] of the SRM 2372a components FAM-labeled TaqMan probes were purchased from Thermo Fisher

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

5

and used in monoplex reactions VIC (a proprietary dye)-labeled probes were purchased from the same source and were used when nDNA and mtDNA were analyzed in the same reaction mixture Figure 2 displays exemplar droplet patterns for the assays All assays were developed at NIST and have been optimized for ddPCR18 Table 2 details three mitochondrial DNA (mtDNA) assays used to estimate mtDNAnDNA in the SRM 2372a components Figure 3 displays exemplar droplet patterns for the assays These assays were optimized for ddPCR from real-time quantitative PCR (qPCR) assays from the literature All primers probes were diluted with TE-4 buffer purchased from Affymetrix-USB Products (Affymetrix Inc Cleveland OH) to a nominal 5 micromol working solution for all single-plex assays (individual nDNA and mtDNA assays) For duplex assays (combined nDNA and mtDNA) primer and probe working solutions were diluted to a nominal 10 micromol

Table 1 NIST-Developed Human Nuclear DNA Assays

Assay Target

Chromosome Band Accession Primers and Probe a

Amplicon Length bp

NEIF Gene EIF5B

Chr 2 p111-q111 NC_00000212

F gccaaacttcagccttctcttc R ctctggcaacatttcacactaca PB+ tcatgcagttgtcagaagctg

67

2PR4 Gene RPS27A

Chr 2 p16 NC_00000212

F cgggtttgggttcaggtctt R tgctacaatgaaaacattcagaagtct PB tttgtctaccacttgcaaagctggccttt

97

POTP STR TPOX

Chr 2 p253 NC_00000212

F ccaccttcctctgcttcacttt R acatgggtttttgcctttgg PT caccaactgaaatatg

60

NR4Q Gene DCK

Chr 4 q133-q211 NC_00000412

F tggtgggaatgttcttcagatga R tcgactgagacaggcatatgtt PB+ tgtatgagaaacctgaacgatggt

83

D5 STR D5S2500

Chr 5 q112 NC_00000510

F ttcatacaggcaagcaatgcat R cttaaagggtaaatgtttgcagtaatagat PT ataatatcagggtaaacaggg

75

ND6 STR D6S474

Chr 6 q21-22 NC_00000612

F gcatggctgagtctaaagttcaaag R gcagcctcagggttctcaa PB+ cccagaaccaaggaagatggt

82

D9 STR D9S2157

Chr 9 q342 NC_00000912

F ggctttgctgggtactgctt R ggaccacagcacatcagtcact PT cagggcacatgaat

60

HBB1 Gene HBB

Chr 11 p155 NC_00001110

F gctgagggtttgaagtccaactc R ggtctaagtgatgacagccgtacct PT agccagtgccagaagagccaagga

76

ND14 STR D14S1434

Chr 14 q3213C NC_0000149

F tccaccactgggttctatagttc R ggctgggaagtcccacaatc PB+ tcagactgaatcacaccatcag

109

22C3 Gene PMM1

Chr 22 q132 NC_00002210

F cccctaagaggtctgttgtgttg R aggtctggtggcttctccaat PB caaatcacctgaggtcaaggccagaaca

78

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

6

a F Forward primer R Reverse primer PB Blackhole quencher probe (FAM labeled) PB+ Blackhole Plus quencher probe (FAM labeled) PT TaqMan MGB probe (FAM labeled)

Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays

Table 2 NIST-Optimized Human Mitochondrial DNA Assays

Assay Location Primers and Probe a Amplicon Length bp Ref

mtND1 34853504 F ccctaaaacccgccacatct

69 22 35333553 R gagcgatggtgagagctaaggt 35063522 PT ccatcaccctctacatc

mtBatz 84468473 F aatattaaacacaaactaccacctacct

79 23 85038524 R tggttctcagggtttgttataa 84758491 PT cctcaccaaagcccata

mtKav 1318813308 F ggcatcaaccaaccacaccta

105 24 1336913392 R attgttaaggttgtggatgatgga 1331013325 PT cattcctgcacatctg

a F Forward primer R Reverse primer PT TaqMan MGB probe (FAM labeled)

2PR4

POTP

NEI

F

R4Q

5

D5

ND6

D9

HBB1

ND1

4

22C3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

7

Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays

Droplet Digital Polymerase Chain Reaction Measurements

Nuclear DNA (nDNA) All nDNA ddPCR measurements used in this study followed the same measurement protocol

bull Droplets were generated on the Auto Droplet Generator using the Bio-Rad ldquoddPCR ldquoSupermix for Probes (No dUTP)rdquo (Bio-Rad cat 186-3024 Control 64079080) and ldquoDroplet Generation Oil for Probesrdquo (Bio-Rad cat 186-4110 Control 64046051) Note desoxyuridine 5-triphosphatase (dUTP) is an enzyme sometimes included in PCR reaction mixtures to minimize PCR carryover contamination

bull Table 3 details the composition of the ddPCR mastermix setup for a given assay and sample where ldquomastermix refers to reaction mixture of manufacturerrsquos proprietary PCR reagent ldquoSupermixrdquo forward and reverse PCR primers for a given assay probe for that assay the sample DNA and sometimes additional water to produce the desired value dilution factor F This setup was used for all assays and samples

bull Non-Template Controls (NTCs) for each assay were included in every analysis bull Droplets were thermal cycled on a ProFlex thermal cycler (Applied Biosystems) for

95 degC for 10 min followed by 60 cycles of 94 degC for 05 min and 61 degC for 1 min then 98 degC for 10 min before a 4 degC hold until the plate was removed

bull Droplets were read on the QX200 Droplet Reader and analysis was performed using the QuantaSoft Analysis Software version 1740917

mtND1 mtBatz mtKav

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

8

bull The numbers of positive and negative droplets at the end of 60 cycles were determined and were exported into a spreadsheet for further analysis Assay-specific intensity thresholds were determined by visual inspection for each assay for each measurement session

Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays

Concentration nDNA Mastermix Microliter

per Reaction 2X ldquoSupermix for Probes (no dUTPs)rdquo 125

5 μmolL Forward Primer 1875 5 μmolL Reverse Primer 1875 5 μmolL Probe (FAM) 125

PCR Grade Water 50 14 Diluted Sample 25 Total Volume 250

Mitochondrial DNA (mtDNA) All the mtDNA measurements used in this study followed the nDNA protocol described above but with different mastermix setups The duplex mtDNA measurements used the setup described in Table 4 The monoplex mtDNA measurements used the setup described in Table 5

Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 125

10 μmolL Forward Primer 09375 10 μmolL Reverse Primer 09375 10 μmolL Probe (VIC) 0625

PCR Grade Water 95 nDNA Mastermix from Table 3 05 Total Volume 250

Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 120

10 μmolL Forward Primer 09 10 μmolL Reverse Primer 09 10 μmolL Probe (FAM) 06

PCR Grade Water 72 Diluted Sample (See Table 7) 24

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

9

Total Volume 240 Sample Preparation

The buffy coat white blood cells used to prepare SRM 2372a materials were purchased from Interstate Blood Bank These materials were from four single-source anonymous individuals supplied to us labeled as Female1 Female2 Male1 and Male2 In our laboratories each donor buffy coat bag was aliquoted (5 mL per aliquot) into sterile 50 mL conical tubes and stored at 4 degC prior to extraction The modified salt-out manual extraction protocol described in Appendix A was performed for each aliquot with rehydration of the DNA in TE-4 buffer From 150 mL to 250 mL of solubilized DNA for each component was accumulated through the extraction of the 5 mL aliquots The DNA was stored in perfluoroalkoxy polymer (PFA) containers that had been bleach sterilized thoroughly rinsed with deionized water rinsed with ethanol and air-dried The minimum volume for production of 2000 units was 120 mL Preliminary estimates of the mass [DNA] of the re-solubilized components were obtained using a Nanodrop microvolume spectrophotometer Dilutions of the component materials were made until the absorbance at 260 nm was approximately 1 corresponding to a nominal mass [DNA] of 50 ngμL of dsDNA7 After addition of the TE-4 buffer the containers were placed on an orbital shaker and gently rotated for several hours prior to refrigerated storage at 4 ordmC The four single-source components were initially screened with the nDNA ddPCR measurement process using four nDNA assays 2PR4 NR4Q ND6 and 22C3 Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate The gravimetric 1rarr4 dilution was prepared by weighing an empty PFA vial and recording the weight then pipetting 30 microL TE-4 into the vial and recording the weight then pipetting 10 microL DNA and recording the final weight The dilution factor was calculated in a spreadsheet These screening measurements verified that the individual component materials were fit-for-purpose The four single-source components were evaluated with the nDNA ddPCR measurement process using the ten nDNA assays Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate Quantitative information was derived from the average of the ten nDNA assays to prepare the component C mixture Thirty milliliters of sample Male2 (estimated at 452 ngμL mass [DNA]) were combined with 80 mL of sample Female2 (estimated at 507 ngμL mass [DNA]) and 10 mL of TE-4 buffer to create a 1 male nDNA to 3 female nDNA producing a 14 malehuman mixture After addition of the TE-4 buffer the component C container was gently rotated on the orbital shaker for several hours prior to refrigerated storage at 4 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

10

Absorbance Measurements

In February 2017 absorbance spectra of the recently prepared fully diluted nominally dsDNA and ssDNA solutions of the three components were acquired with a BioCary 100 spectrophotometer using the measurement protocol described in Appendix B In December 2017 additional spectra were acquired for the remaining bulk solution of components A and B and for SRM 2372 components A and B These solutions had been stored at 4 degC in well-sealed PFA containers There was no remaining bulk component C solution SRM Unit Production

In the morning of the day a component was transferred into the sterile component tubes (Sarstedt Biosphere SC micro tube-05 mL Newton NC) used to package the SRM the container holding the diluted component was removed from the refrigerator and placed on a slow orbital shaker for 2 h The tubes were opened and placed under the laminar flow hood in rows of 5 within 80-hole tube racks for a total of 30 tubes per rack Tubes were placed in every second row of the racks to facilitate filling with an Eppendorf Repeater Xstream pipette (Eppendorf North America Inc Hauppauge NY) fitted with a 1-mL positive displacement tip set to dispense 55 microL per tube The component container was removed from the shaker a magnetic stir-bar was added the container was placed on a magnetic stir plate in the laminar flow hood and the material was stirred gently Filling proceeded until 2200 vials were filled with pipet tip replacement after filling every 200 tubes The filling process consisted of two individuals manually filling each vial within a biosafety cabinet (one physically pipetting each sample the other verifying that each tube was filled) and four individuals tightly closing the lids to each tube on a sterilized and covered bench When all tubes in a rack were filled the rack was moved out of the hood the tubes checked for proper filling their lids closed and the tubes transferred to 100-unit pre-labeled storage boxes As the 100-unit boxes were filled they were transferred to the labeling room where SRM labels were applied by hand and colored inserts were added to the lids red for component A white for component B and blue for component C After all tubes for a given component were labeled the 22 boxes were placed in a locked refrigerator Filling of each of the components took 60 min to 90 min All units were transferred to the refrigerator within 4 hours of beginning the filling process All units of the three materials were allowed to equilibrate for a minimum of two weeks prior to beginning homogeneity analysis Component Homogeneity

The ddPCR measurement protocol described above was used to assess the nDNA homogeneity of the three components One tube from the approximate center of each of the 22 storage boxes was assayed using the NEIF and ND6 assays (see Table 1) A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Three technical replicates were obtained for every tube A total

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

11

of six plates were used two plates per component one with the NEIF assay and one with the ND6 assay Figure 4 details the layout of the samples

Figure 4 Plate Layout Design for Homogeneity Measurements

Component Stability

Stability of the materials was evaluated at 4 degC room temperature (RT asymp 22 degC) and 37 degC Tubes from the homogeneity study were divided among the three temperatures for the stability study Additionally two never-opened tubes (one from box 8 and one from box 12 of each component) were held at each temperature Two tubes at each temperature for each component were opened sampled and resealed 4 times over the course of 8 weeks Measurements were made at week 1 week 3 week 5 and week 8 The measurement on week 8 was from an unopened tube (from box 8) for each of the components at all temperatures The data from the homogeneity study was used as timepoint 0 in the stability study anchored to 4 degC by the storage recommendations A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used for both the NEIF and ND6 assays All temperatures components and both assays fit onto one plate for timepoints 1 to 5 Twelve NTCs placed in rows G and H were included for each assay Figure 5 details the sample layout

Figure 5 Plate Layout Design for Stability Measurements

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

ND6 ND6 ND6

NEIFNEIFNEIFComponent A

Component A

Component B

Component B

Component C

Component C

1 2 3 4 5 6 7 8 9 10 11 12A A-7 A-7 B-7 B-7 C-7 C-7 A-7 A-7 B-7 B-7 C-7 C-7B A-10 A-10 B-10 B-10 C-10 C-10 A-10 A-10 B-10 B-10 C-10 C-10C A-9 A-9 B-9 B-9 C-9 C-9 A-9 A-9 B-9 B-9 C-9 C-9D A-12 A-12 B-12 B-12 C-12 C-12 A-12 A-12 B-12 B-12 C-12 C-12E A-8 A-8 B-8 B-8 C-8 C-8 A-8 A-8 B-8 B-8 C-8 C-8F A-11 A-11 B-11 B-11 C-11 C-11 A-11 A-11 B-11 B-11 C-11 C-11G NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTCH NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

NEIF

4C

RT

37C

ND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

12

Component Certification Measurements

After establishing homogeneity and stability measurements intended to be used to certify nDNA values and inform mtDNA values were started using randomly selected SRM tubes from the boxes listed in Table 7 The ten nDNA assays described in Table 1 and two of the three mtDNA assays described in Table 2 were run simultaneously on the same sample plate A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used in each of the six measurement campaigns

Table 6 Tubes used for Certification Measurements

Component Plate A B C

dd20170711 Box 1 Box 1 Box 1 dd20170712a Box 10 tube a Box 10 tube a Box 10 tube a dd20170712b Box 10 tube b Box 10 tube b Box 10 tube b dd20170713 Box 20 Box 20 Box 20 dd20170714 Box 5 Box 5 Box 5 dd20170801 Box 5 Box 5 Box 5

Every sample was analyzed in duplicate with every assay in all six plates Component A was always located in rows B and C component B in rows D and E component C in rows F and G and NTCs in rows A and H Figure 6 displays the basic layout of the ten nDNA and two mtDNA assays In addition to the plate column assignment scheme shown the three assay groupings were also ordered as Group 2 Group 3 Group 1 and Group 3 Group 1 Group 2 to address potential positional bias

Group 1 Group 2 Group 3 2PR4 POTP NEIF NR4Q D5 ND6 D9 HBB1 ND14 22C3 mtND1 mtBatz 1 2 3 4 5 6 7 8 9 10 11 12 A NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Comp A B Comp A C Comp B D Comp B E Comp C F Comp C G

H NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Figure 6 Basic Plate Layout Design for Certification Measurements

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

13

Mitochondrial DNA Measurements

A ldquoduplexed reactionrdquo was used to determine mtDNAnDNA where 1 μL of one of the nDNA mastermix was added to 49 μL of the mtDNA mastermix resulting in a 1rarr50 volumetric dilution of the nDNA The nDNA was targeted with a FAM-labeled probe while the mtDNA was targeted with a VIC labeled probe The primer and probe concentrations were increased from 5 micromol to 10 micromol in the duplexed reactions Additional testing of the mtDNA assays were performed in monoplex with FAM-labeled probes to determine the additional sources of bias between sampling assay setup and dilution factors Table 7 outlines the different methods examined to determine mtDNAnDNA for the three components

Table 7 Run Parameters for Mitochondrial DNA Droplet

Generation nDNA mtDNA

Plate Component Assay Dilution Reps Assay Dilution Reps dd20170824 A manualDG HBB1 1rarr4 3 mtND1 1rarr200 9 dd20170825 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170829 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170831 AB autoDG HBB1 1rarr3 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr6 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr12 2 mtND1 1rarr10rarr100 3 dd20170907 ABC manualDG HBB1 1rarr3 2 mtBatz 1rarr10rarr100 3 dd20170913 AB manualDG NEIF 1rarr4 3 mtND1 mtBatz mtKav 1rarr10rarr100 3 dd20170913 C manualDG NEIF 1rarr4 3 mtBatz mtKav 1rarr10rarr100 3 Performance in Commercial qPCR Chemistries

Table 8 lists the 11 qPCR chemistries that were commercially available in late 2017 All three of the SRM 2372a components were evaluated with these chemistries using component A of SRM 2372 as the reference standard For each component of SRM 2372a four tubes were pulled from box 22 and combined into one tube A 1rarr5 dilution of all samples was prepared followed by three serial 1rarr2 dilutions A five-point serial dilution of each SRM 2372a component was amplified in triplicate for each chemistry This consisted of the neat material the 1rarr5 dilution then the three serial 1rarr2 dilutions A six-point serial dilution of SRM 2372 Component A was also created for the standard curve for the commercial qPCR chemistries This consisted of the neat material followed by a 1rarr5 dilution then followed by four 1rarr2 dilutions serially The six-point serial dilution series of SRM 2372 Component A was also amplified in triplicate These commercial chemistries were amplified in individual plates on an Applied Biosystems 7500 HID Real-Time PCR Instrument (Foster City CA) following the manufacturersrsquo recommended protocols for reaction mix setup and cycling conditions Figure 7 details the plate layout for the qPCR plates Data was analyzed with the HID Real-Time PCR Analysis Software v12 and further analyzed in a spreadsheet

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

14

Table 8 Commercial qPCR Chemistries Tested with SRM 2372a

Manufacturer Commercial Kit Target Codea

Innogenomics

InnoQuant Long IQh Short IQd

InnoQuantHY Long IQHYh Short IQHYd Y IQHYy

Thermo Fisher Quantifiler Duo Human QFDh Male QFDy

Quantifiler HP Large QFHPh Small QFHPd

Quantifiler Human Human QFHh

Quantifiler Trio Large QFTh Small QFTd Y QFTy

Qiagen

Quantiplex Autosomal QPh

Quantiplex Hyres Autosomal QPHh Y QPHy

Quantiplex Pro Human QPPh Degradation QPPd Y QPPy

Promega

Plexor Autosomal Ph Y Py

PowerQuant Autosomal PQh Degradation PQd Y PQy

a Assays are coded with a short acronym derived from the kit name followed by the following codes for the type of target ldquohrdquo autosomal non-degradation described as ldquoAutosomalrdquo ldquoHumanrdquo or ldquoLongrdquo ldquodrdquo autosomal degradation described as ldquoDegradationrdquo or ldquoShortrdquo ldquoyrdquo Y-chromosome described as ldquoMalerdquo or ldquoYrdquo

1 2 3 4 5 6 7 8 9 10 11 12

A

B 2372-A_57 2372-A_57 2372-A_57 2372a-neat_A

2372a-neat_A

2372a-neat_A

2372a-neat_B

2372a-neat_B

2372a-neat_B

2372a-neat_C

2372a-neat_C

2372a-neat_C

C 2372-A_11

4 2372-A_11

4 2372-A_11

4 2372a-15_A

2372a-15_A

2372a-15_A

2372a-15_B

2372a-15_B

2372a-15_B

2372a-15_C

2372a-15_C

2372a-15_C

D 2372-A_5

7 2372-A_5

7 2372-A_5

7 2372a-110_A

2372a-110_A

2372a-110_A

2372a-110_B

2372a-110_B

2372a-110_B

2372a-110_C

2372a-110_C

2372a-110_C

E 2372-A_2

85 2372-A_2

85 2372-A_2

85 2372a-120_A

2372a-120_A

2372a-120_A

2372a-120_B

2372a-120_B

2372a-120_B

2372a-120_C

2372a-120_C

2372a-120_C

F 2372-A_1

425 2372-A_1

425 2372-A_1

425 2372a-140_A

2372a-140_A

2372a-140_A

2372a-140_B

2372a-140_B

2372a-140_B

2372a-140_C

2372a-140_C

2372a-140_C

G 2372-A_0

7125 2372-A_0

7125 2372-A_0

7125

H NTC NTC NTC

Figure 7 Plate Layout Design for Performance Assessments

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

15

Measurement Results After preparing the SRM 2372a bulk solutions their approximate 50 ngmicroL mass [DNA] were confirmed spectrophotometrically The bulk solutions were then packaged into individual tubes The materials contained in the tubes were characterized for homogeneity stability nDNA copy number [DNA] and mtDNAnDNA ratio The following sections detail the results of these measurements Absorbance Spectrophotometry

By widely accepted convention an aqueous DNA solution having an absorbance of 10 at 260 nm corresponds to a mass [DNA] of 50 ngmicroL for dsDNA and (37 to 40) ngmicroL for ssDNA7 Absorbance at (230 270 280 and 330) nm are traditionally used in the assessment of DNA quality725 Due to minor baseline shifts during acquisition of some spectra it was necessary to bias-adjusted all spectra to have absorbance of 00 at 330 nm Table 9 lists the resulting bias-adjusted values for selected wavelengths and the conventional mass [DNA] conversion of the absorbance at 260 nm The close agreement between the mass concentration values for the nominally dsDNA and the converted-to-ssDNA spectra confirm that the bulk solutions were correctly prepared to be asymp 50 ngμL dsDNA The slight increase in absorbance over the 10 months between acquiring the spectra suggests that like SRM 2372 the tertiary structure of the dsDNA in the SRM 2372a materials may be changing slowly with time Figure 8 displays the spectra

Table 9 Absorbance at Selected Wavelengths

Wavelength nm Form Date Component 230 260 270 280 ngμL dsDNA 2142017 A 0394 0931 0758 0488 466

B 0480 1112 0905 0583 556 C 0451 1006 0815 0524 503

ssDNA 2142017 A 0390 0614 0540 0316 454 B 0457 0723 0634 0372 535 C 0411 0641 0563 0330 474

dsDNA 12142017 A 0402 0956 0775 0503 478 B 0484 1128 0914 0589 564

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

16

Figure 8 Absorbance Spectra of SRM 2372a Components A B and C

dsDNA is denatured to ssDNA by diluting with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to DNA mass concentration

Droplet Volume

Droplet volumes for the two lots of ldquoSupermixrdquo used in the studies described in this report were determined using NISTrsquos Special Test Method 11050S-D as described in Dagata et al19 The lot used for most of the ddPCR analyses produced (07349 plusmn 00085) nL droplets where the ldquoplusmnrdquo value is the combined standard uncertainty The lot used for analyses performed after October 11 2017 produced (07376 plusmn 00085) nL droplets NIST measurements have demonstrated that droplet volumes are not influenced by the presence or absence of DNA in the sample solution Measurements repeated over time on an earlier lot of the Supermix product used in these studies suggest that droplet volumes remain constant over the shelf-life of the lot However droplet volumes for different lots of this product as well as for different Supermix products have been observed to differ by several percent We intend to document these measurements in a future report Other researchers have demonstrated that non-constant droplet volume distributions may contribute to ddPCR measurement imprecision and bias2627 However these effects become significant only with distributions wider than observed using Special Test Method 11050S-D19 and at copydroplet values larger than the λ le 05 copydroplet typically used in our measurements Certification Measurements

Homogeneity Results Table 10 lists the λacute nDNA copy number values adjusted for both droplet volume and sample dilution from the homogeneity measurements The data was analyzed using the three-level Gaussian hierarchical ANOVA model described in Appendix C Table 11 lists the values of the within- and between-tube variance components expressed as percent relative standard deviations At the 95 confidence level all components are homogeneous

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372a-A2141721417-NaOH121417

220 260 300 340

Wavelength nm

2372a-B2141721417-NaOH121517

220 260 300 340

Wavelength nm

2372a-C21417

21417-NaOH

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

17

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 NEIF A-1-01 1632 1529 1511 B-1-01 1790 1909 1885 C-1-01 1493 1385 1510 ND6 1616 1640 1582 NAe 1870 1860 1565 1562 1466

NEIF A-1-02 1531 1525 1549 B-1-02 1712 1719 1704 C-1-02 1516 1513 1577 ND6 1615 1589 1601 1775 1718 1756 1490 1547 1570

NEIF A-1-03 1524 1540 1533 B-1-03 1693 1755 1749 C-1-03 1451 1528 1506 ND6 1582 1152 1555 1824 1483 1896 1485 1505 1471

NEIF A-1-04 1454 1458 1427 B-1-04 1754 1745 1739 C-1-04 1440 1486 1520 ND6 1627 1605 1607 1670 1632 1814 1529 1508 1506

NEIF A-1-05 1465 1540 1504 B-1-05 1757 1687 1735 C-1-05 1523 1472 1483 ND6 1607 1633 1586 1683 1759 1703 1513 1520 1491

NEIF A-1-06 1512 1491 1491 B-1-06 1697 1709 1714 C-1-06 1437 1493 1522 ND6 1511 1483 1524 1693 1723 1722 1492 1481 1461

NEIF A-1-07 1418 1480 1455 B-1-07 1722 1625 1703 C-1-07 1485 1477 1463 ND6 1479 1557 1656 1690 1684 1688 1475 1547 1485

NEIF A-1-08 1496 1560 1501 B-1-08 1884 1802 1772 C-1-08 1433 1405 1478 ND6 1514 1615 1540 1864 1887 1821 1497 1501 1495

NEIF A-1-09 1588 1581 1637 B-1-09 1713 1734 1727 C-1-09 1463 1499 1505 ND6 1474 1438 1454 1709 1756 1758 1503 1491 1516

NEIF A-1-10 1574 1513 1570 B-1-10 1759 1809 1738 C-1-10 1534 1458 1454 ND6 1594 1713 1669 1764 1816 1787 1448 1465 1474

NEIF A-1-11 1557 1508 1633 B-1-11 1749 1782 1740 C-1-11 1533 1514 1555 ND6 1571 1611 1631 1741 1792 1784 1488 1470 1483

NEIF A-1-12 1500 1476 1515 B-1-12 1748 1712 1718 C-1-12 1450 1525 1497 ND6 1626 1586 1608 1797 1614 1790 1496 1483 1519

NEIF A-1-13 1672 1634 1804 B-1-13 1632 1668 1701 C-1-13 1537 1527 1515 ND6 1490 1516 1497 1626 1566 1681 1499 1479 1475

NEIF A-1-14 1551 1568 1538 B-1-14 1752 1535 1673 C-1-14 1515 1490 1494 ND6 1617 1614 1639 1724 1741 1685 1469 1480 1466

NEIF A-1-15b 1793 1722 1733 B-1-15 1871 1835 1887 C-1-15 1511 1568 1594 ND6 1818 1753 1762 1770 1752 1798 1515 1543 1565

NEIF A-1-16 1573 1520 1617 B-1-16 1688 1725 1743 C-1-16 1449 1452 1471 ND6 1634 1627 1575 1744 1697 1710 1543 1448 1491

NEIF A-1-17 1516 1516 1628 B-1-17 1789 1769 1825 C-1-17 1388 1421 1394 ND6 1613 1567 1567 1716 1762 1638 1452 NAe 1435

NEIF A-1-18 1612 1535 1672 B-1-18 1826 1780 1849 C-1-18 1475 1455 1445 ND6 1529 1495 1528 1805 1801 1931 1458 1531 1459

NEIF A-1-19 1598 1514 1596 B-1-19 1751 1759 1880 C-1-19 1468 1525 1446 ND6 1571 1535 1511 1827 1766 1786 1435 1478 1421

NEIF A-1-20 1575 1568 1586 B-1-20 1738 1678 1768 C-1-20 1549 1531 1499 ND6 1652 1545 1627 1752 1747 1768 1499 1465 1468

NEIF A-1-21 1658 1611 1578 B-1-21 1742 1836 1713 C-1-21 1530 1440 1484 ND6 1647 1607 1612 1746 1744 1764 1499 1454 NAe

NEIF A-1-22 1547 1528 1532 B-1-22 1761 1743 1660 C-1-22 1442 1487 1436 ND6 1537 1513 1536 1666 1692 1819 1589 1591 1556

NEIF A-1-15c 1456 1448 1523 ND6 1525 1402 1436

NEIF A-2-15d 1679 1599 1674 ND6 1581 1550 1584

a Tube coded as Component-Tube-Box b These results were identified as a probable error in the preparation or volumetric delivery of the sample

solution into the microfluidic device and were not used in the homogeneity evaluations c Repeat assessment of the solution in sample A-1-15 d Results for a second tube from box 15 analyzed at the same time as the repeat assessment of A-1-15 e Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

18

Table 11 Within- and Between-Tube Relative Standard Deviations

Within-Tube Component NEIF ND6 Between-Tube

A 28 40 33 B 27 38 25 C 23 18 19

Stability Results Table 12 lists the λacute volume- and dilution-adjusted nDNA copy number values from the stability measurements Figure 9 summarizes these results The results for the tubes held at 4 degC room temperature (RT asymp 22 degC) and 37 degC are very similar for all three components The absence of between-temperature differences identifies the between-week differences as attributable to variation in the measurement processes rather than sample instability These results indicate that the SRM 2372a genomic copy number is thermally stable from 4 degC to 37 degC over an extended period

Figure 9 ddPCR Stability Measurements as Function of Time

Each dot-and-bar symbol represents the mean λacute and the approximate 95 expanded uncertainty for all results for a given component at a given storage temperature The thick black horizontal lines represent an approximate 95 confidence interval on the λacute values estimated from the homogeneity data

13

14

15

16

17

18

19

0 1 2 3 4 5 6 7 8

λacute H

aplo

id n

DNA

Copi

es μL

Weeks

A

0 1 2 3 4 5 6 7 8

Weeks

B

4 degC 22 degC 37 degC

0 1 2 3 4 5 6 7 8

Weeks

C

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

19

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Week Temp Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3

1 4 degC NEIF A-10 1568 1441 B-10 1666 1653 C-10 1424 1394 ND6 A-10 1474 1494 B-10 1655 1705 C-10 1461 1394

NEIF A-07 1500 1463 B-07 1718 1640 C-07 1446 1452 ND6 A-07 1479 1479 B-07 1708 1686 C-07 1522 1482

RT (22 degC) NEIF A-12 1674 1608 B-12 1805 1777 C-12 1506 1495 ND6 A-12 1522 1507 B-12 1759 1788 C-12 1478 1461 NEIF A-09 1635 1539 B-09 1791 1793 C-09 1411 NAb ND6 A-09 1543 1519 B-09 1661 1682 C-09 1482 1496

37 degC NEIF A-11 1614 1637 B-11 1748 1744 C-11 1515 1464 ND6 A-11 1494 1502 B-11 1740 1697 C-11 1495 1535 NEIF A-08 1592 1614 B-08 1691 1725 C-08 1528 1500 ND6 A-08 1581 1595 B-08 1727 1699 C-08 1458 1483 3 4 degC NEIF A-10 1520 1706 B-10 NAb 1733 C-10 1427 1473 ND6 A-10 1543 1624 B-10 1670 1755 C-10 1443 1364 NEIF A-07 1537 1570 B-07 1549 1565 C-07 1424 1512 ND6 A-07 1517 1479 B-07 1705 1723 C-07 1483 1439

RT (22 degC) NEIF A-12 1521 1607 B-12 1817 1842 C-12 1566 1471 ND6 A-12 1644 1597 B-12 1845 1786 C-12 1504 1511 NEIF A-09 1432 1626 B-09 1825 1856 C-09 1392 1387 ND6 A-09 1539 1528 B-09 1734 1737 C-09 1535 1465

37 degC NEIF A-11 1651 1685 B-11 1759 1779 C-11 1437 1440 ND6 A-11 1680 1594 B-11 1806 1806 C-11 1471 1509 NEIF A-08 1580 1613 B-08 1740 1791 C-08 1582 1590 ND6 A-08 1592 1524 B-08 1769 1694 C-08 1468 1520 5 4 degC NEIF A-10 1486 1440 B-10 1626 1551 C-10 1381 1330 ND6 A-10 1555 1577 B-10 1673 1625 C-10 1547 1395 NEIF A-07 1387 1429 B-07 1723 1787 C-07 1330 1314 ND6 A-07 1538 1522 B-07 1862 1909 C-07 1452 1458

RT (22 degC) NEIF A-12 1601 1631 B-12 1718 1752 C-12 1546 1541 ND6 A-12 1699 1677 B-12 1772 1721 C-12 1490 1471 NEIF A-09 1381 1484 B-09 1675 1694 C-09 1371 1376 ND6 A-09 1625 1615 B-09 1904 1815 C-09 1468 1472

37 degC NEIF A-11 1456 1517 B-11 1720 1704 C-11 1505 1493 ND6 A-11 1650 1658 B-11 1148 1152 C-11 1527 1600 NEIF A-08 1590 1666 B-08 1745 1723 C-08 1469 1540 ND6 A-08 1235 1244 B-08 1809 1749 C-08 1508 1529 8 4 degC NEIF A-08a 1667 1632 1589 B-08a 1880 1884 1940 C-08a 1525 1475 1542 ND6 A-08a 1599 1593 1590 B-08a 1859 1897 1972 C-08a 1550 1502 1520

RT (22 degC) NEIF A-08b 1693 1666 1634 B-08b 1769 1884 1771 C-08b 1595 1557 1545 ND6 A-08b 1644 1620 1647 B-08b 1862 1738 1811 C-08b 1611 1585 1535

37 degC NEIF A-08c 1620 1576 1608 B-08c 1934 1952 1956 C-08c 1595 1600 1569 ND6 A-08c 1576 1553 1532 B-08c 1889 1844 1889 C-08c 1569 1514 1581

a Tube coded as Component-Tube-Box b Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

20

Ten-Assay Certification Results The original intention was to certify the copy number [DNA] of the three components with results from five independently prepared sets of samples analyzed over four days Each sample of the three components was from a new tube from different boxes and was evaluated with 10 nDNA assays The assay positions were alternated across these five ldquo10-assay certificationrdquo plates The λ results for these five plates are systematically somewhat larger than those for the homogeneity and stability data On review we realized that a different workspace and its suite of pipettes had been used to prepare the 1rarr10 dilution of the 1rarr4 diluted components into the reaction mastermix The ldquohighrdquo workstation was used for stability and homogeneity measurements while the ldquolowrdquo workstation was used for certification measurements Gravimetric testing of the pipettes established that the volumes delivered by these alternate pipettes was sufficiently biased to account for the observed λ differences A sixth ldquocertificationrdquo plate was prepared using the original (and now re-verified) pipettes The λ results or this plate agreed well with the homogeneity and stability data Table 13 lists the λacute volume- and dilution-adjusted nDNA copy number values for all six of the ten-assay datasets Figure 10 displays the extent of agreement among the ten assays While there are small-but-consistent differences between the assays there is only a 4 difference between the assays providing smallest and largest λ

Figure 10 Comparison of Assay Performance

The blue dot-and-bars represent the mean and its combined standard uncertainty for the A B and C components for each assay relative to the over-all grand mean for each component The black dot-and-bars represent combination of three component results for each assay The solid horizontal line represents the median difference of the assay means from the grand mean the dotted horizontal lines bound an approximate 95 confidence region about the median

2PR4POTPNEIFR4Q5 D5ND6 D9 HBB1ND14 22C3-10

-8

-6

-4

-2

0

2

4

6

8

10

Rela

tive

Diffe

renc

e fr

om M

ean

Assays

Component MeansAssay MeansMedianU95

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

21

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

7112017 2PR4 1667 1698 1869 1960 1630 1706 POTP 1655 1685 1850 1781 1596 1576 NEIF 1647 1665 1709 1707 1574 1538 R4Q5 1594 1658 1856 1787 1590 1555 D5 1655 1683 1848 1829 1610 1607 ND6 1629 1647 1828 1826 1589 1548 D9 1655 1652 1861 1858 1569 1572 HBB1 1644 1611 1825 1869 1565 1681 ND14 1645 1707 1871 1773 1564 1492 22C3 1545 1603 1776 1902 1548 1655

7122017 2PR4 1668 1739 2174 2231 1499 1497 POTP 1587 1702 2202 2292 1576 1611 NEIF 1678 1614 2172 2216 1545 1560 R4Q5 1628 1762 NA 2078 1584 1546 D5 1738 1649 2230 2192 1577 1529 ND6 1667 1678 2094 2034 1599 1478 D9 1695 1653 1949 2110 1587 1504 HBB1 1757 1750 1971 2123 1520 1551 ND14 1771 1764 2123 2066 1479 1438 22C3 1762 1759 2107 2046 1512 1494

7122017 2PR4 1638 1667 1810 1806 1452 1424 POTP 1676 1638 1865 1801 1347 1397 NEIF 1616 1624 1806 1786 1395 1350 R4Q5 1559 1587 1878 1768 1442 1459 D5 1608 1652 1824 1790 1448 1602 ND6 1611 1589 1768 1670 1463 1435 D9 1517 1607 1811 1749 1402 1433 HBB1 1557 1582 1763 1741 1474 1481 ND14 1604 1625 1865 1781 1466 1482 22C3 1675 1616 1815 1866 1439 1568

7132017 2PR4 1888 1939 1784 1821 1470 1444 POTP 1868 1815 1735 1813 1411 1427 NEIF 1953 1902 1778 1813 1512 1439 R4Q5 1779 1871 1835 1749 1452 1468 D5 1924 2124 1794 1848 1567 1524 ND6 1861 1787 1700 1711 1480 1424 D9 1824 1844 1799 1779 1442 1459 HBB1 1809 1799 1765 1769 1430 1467 ND14 1864 1857 1754 1725 1413 1411 22C3 1930 1942 1792 1814 1486 1480

7142017 2PR4 1597 1590 1718 1768 1485 1458 POTP 1533 1404 1753 1698 1422 1459 NEIF 1462 1533 1773 1792 1438 1465 R4Q5 1493 1519 1777 1740 1439 1468 D5 1538 1586 1803 1795 1515 1457 ND6 1492 1516 1718 1693 1409 1412 D9 1528 1538 1802 1743 1469 1474 HBB1 1548 1581 1834 1745 1453 1428 ND14 1513 1514 1731 1714 1465 1425 22C3 1562 NA 1842 NA 1363 1393

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

22

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

101217 2PR4 1431 1411 1791 1770 1357 1364 POTP 1407 1395 1746 1820 1447 1361 NEIF 1487 1398 1803 1803 1448 1404 R4Q5 1450 1439 1767 1739 1409 1411 D5 1394 1422 1745 1773 1389 1429 ND6 1396 1393 1824 1729 1415 1417 D9 1414 1390 1752 1744 1431 1380 HBB1 1401 1382 1813 1754 1381 1336 ND14 1412 1460 1648 1671 1403 1387 22C3 1435 1410 1818 1741 1435 1372

Estimation of Copy Number Concentration It was decided that all three sets of results (homogeneity stability and certification) would be used to derive the certified values with the λ of the first five ldquoten-assayrdquo plates adjusted by a data-driven estimate of the volumetric bias This bias was incorporated into the statistical model providing better estimates of the λ means while capturing all of the variance components (homogeneity stability between-assay and between data sets) as random effects using Bayesian MCMC programmed in OpenBUGS28 Table 14 lists the estimated values and their uncertainties See Appendix C for details

Table 14 Measured Copy Number Concentrations

Genomic Copies per Nanoliter of Solution Component Mean a u(Mean) b U95(Mean) c CV95

d A 151 07 14 93 B 175 03 06 34 C 145 05 10 69

a Mean the experimentally determined most probable estimate of the true value b u(Mean) the standard uncertainty of the Mean c Mean plusmnU95(Mean) the interval which is believed with about 95 confidence to contain the true value d CV95 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the counting unit one29 2) the validity of the Poisson endpoint transformation for digital PCR endpoint assays when applied to samples providing le 05 copies per droplet 3) the determination that the copies are dispersed as dsDNA and 4) calibrated mean droplet volume measurements made at NIST during sample dilution and mastermix preparation Conversion of Copy Number to Mass Concentration As described in Reference 18 metrologically valid conversion from copies per nanoliter to nanograms per microliter requires quantitative estimates for the number of nucleotide bases per reference human genome and the average nucleotide mass Using the values estimated in that report and asserting that each target copy corresponds to one HHGE the required transformation is

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

23

[DNA] ngmicroL

= [DNA] HHGE

nLtimes

3301 ngHHGE

The combined standard uncertainty is

119906119906 [DNA] ng

microL = 1199061199062 [DNA] HHGE

nL + 0531100

times[DNA] HHGE

nL 2

Both the measured and the conversion factor uncertainties are associated with ldquolargerdquo degrees of freedom Therefor the approximate 95 confidence uncertainty is estimated

11988011988095 [DNA] ng

microL = 2 times 119906119906

[DNA] ngmicroL

Table 15 lists the transformation of the Table 14 copy number values These values agree well with the conventional 260 nm absorbance values averaging (4 plusmn 5) larger

Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution

ngmicroL Component Value a u(Measurement)b u(Conversion)c U95(Value)d CV95

e A 498 23 03 47 93 B 578 10 03 21 36 C 479 17 03 33 70

a Estimated true value b Standard uncertainty component from the ddPCR measurements c Standard uncertainty estimated from the relative uncertainty of the conversion factor d Value plusmnU95(Value) is believed with about 95 confidence to contain the true value e 100U95(Value)Value the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the measured estimates of copies per nanoliter and 2) determinations made by us from internationally recognized literature resources of the average number of nucleotide bases per HHGE and the mean molecular weights of the bases18 Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)

While not intended to provide certified values mtDNAnDNA was measured as part of 13 experiments over 5 months Measurements were made using dilution-adjusted λ for the mtND1 mtBatz andor mtKav mitochondrial assays referenced against the dilution-adjusted λ of either the NEIF or HBB1 nDNA assays By chance both DNAs in the component C mixture have a binding site mutation that silences the mtND1 probe Figure 11 graphically summarizes the results by assay and component Table 16 lists the data Table 17 summarizes results from a random effects model that accounts for between-assay variability See Appendix C for details

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

24

Due to the limited number of mtDNA assays used in these measurements these results are metrologically traceable only to the assays used

Figure 11 mtDNAnDNA for Components A B and C

Each dot-and-bar symbol represents a mean mtDNAnDNA and its approximate 95 expanded uncertainty The dotted horizontal lines represent approximate 95 confidence intervals on the ratios

Table 16 mtDNAnDNA Ratios

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

20170711a 1 mtND1 1931 1870 2461 2401 1 mtBatz 1845 1764 2555 2608 3201 3299

20170711b 1 mtND1 1810 1906 2229 2200 1 mtBatz 1785 1837 2108 2192 3347 3287

20170712 1 mtND1 1743 1761 2127 2198 1 mtBatz 1829 1847 2455 2506 3217 3384

20170713 1 mtND1 1716 1800 2211 2199 1 mtBatz 1740 1813 2380 2249 3200 3363

20170714 1 mtND1 1923 1835 2105 1943 1 mtBatz 1935 1851 2136 2119 3294 3013

20170814 1 mtND1 1780 1776 1706 2046 2057 2024 1 mtBatz 1804 1751 1738 1998 2043 1952 2927 2776 2762 1 mtKav 1769 1786 1761 2077 2117 2041 3044 2830 2825

20170824 1 mtND1 1947 1928 1922 2 mtND1 1829 1895 1851 3 mtND1 1674 1617

20170825 1 mtBatz 1667 1813 1708 1598 1928 1916 1915 2832 2840 2788 2725 2 mtBatz 1566 1687 1663 1589 1947 1946 1961 1909 2835 2790 2736 2679 3 mtBatz 1643 1668 1604 1573 1953 1952 1941 1881 2803 2773 2757 2752

20170829 1 mtBatz 1719 1716 1670 1666 1865 1863 1828 1828 2599 2508 2551 2609 2 mtBatz 1581 1573 1601 1612 1944 1964 1947 1895 2617 2483 2626 2602 3 mtBatz 1546 1542 1538 1587 1908 1909 1940 1839 2665 2535 2570 2650

20170831 1 mtND1 1617 1599 1688 1783

ND1

ND1

ND1Ka

v

Kav

Kav

Batz

Batz

Batz

A

B

C

160

210

260

mtD

NAn

DNA

Ratio

Assay

Bind

ing s

ite m

utat

ion

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

25

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

2 mtND1 1622 1579 1823 1631 3 mtND1 1575 1528 1920 1887

20170907 1 mtBatz 1584 1544 1890 1886 2501 2522 20170913 1 mtND1 1757 1724 1768 1990 1921 1982

1 mtBatz 1743 1671 1766 1990 1924 1958 2805 2825 2829 1 mtKav 1692 1662 1767 1903 1831 1919 2709 2738 2727

20171016 1 mtND1 1924 1944 1893 2304 2341 2332 1 mtBatz 1786 1752 1743 2048 2078 2067 2797 2642 2792 1 mtKav 1530 1542 1523 2399 2479 2493 2804 2789 2842 2 mtBatz 1580 1557 1592 2715 2606 2687 2 mtKav 2103 2100 2049 2799 2628 2559 3 mtND1 1749 1769 1777 2065 2166 2164 3 mtBatz 1965 1889 1947 2314 2280 2258 2639 2535 2657 3 mtKav 1813 1864 1864 1867 1987 1971 2953 2757 2802

Table 17 mtDNAnDNA Summary Results

mtDNAnDNA Component Value a u(Value)b U95(Value)c CV95

d A 1737 16 38 17 B 2058 22 44 21 C 2791 23 47 17

a Estimated true value b Standard uncertainty c Value plusmnU95(Value) is believed with about 95 confidence to contain the true value d 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) Performance in Commercial qPCR Chemistries

Traditional qPCR assays only quantify relative to a reference standard Thus qPCR assays do not provide absolute [DNA] for the individual materials but they can provide [DNA] of one component relative to another Figure 12 summarizes the BA and CA ratios for the commercial autosomal (Human) targets from the eleven assays evaluated (see Table 8) The Y-chromosome (Male) assays cannot be summarized in this manner as the (all female) component B does not include a Y chromosome Figure 13 summaries the mass [DNA] results for the (all male) component A and (male-female mixture) component C Since this material was prepared as a nominal 13 malefemale mixture the expected malehuman (total) ratio is 14 sjc As displayed in these figures there is intrinsic variability between the SRM 2372a components and between the commercial qPCR chemistries examined This reflects the difference in the targets and amplification chemistries for each of the commercial kits Many of the commercial chemistries on the market to date contain multi-copy targets for both autosomal and Y-chromosomal quantification to increase the sensitivity of the assays

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

26

Figure 12 qPCR Autosomal Quantitation Performance Summary

Open circles denote the BA CA ratios the 17 commercial qPCR assays evaluated each labeled with the code listed in Table 8 The solid black circle represents ratios as estimated by ddPCR The red lines bound the approximate 95 confidence intervals on the ddPCR-estimated ratios

Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary

Open squares denote the ngμL values for SRM 2372a Components A (male) and C (mixture of male and female) for the seven commercial Y-chromosome qPCR assays evaluated each labeled with the code listed in Table 8 The dotted red lines represent malehuman ratios of 13 14 and 15

Appendix D provides a graphical analysis of all the data generated in the performance assessments for each of the targets amplified in the eleven commercial qPCR chemistries and the [DNA] values for all of the assays

dd

IQ_h

IQ_d

IQHY_h

IQHY_d

QFD_h

QFH_h QFHP_h

QFHP_dQFT_h

QFT_d

QPH_h

QPP_hQPP_d

P_h

PQ_h

PQ_d

06

08

10

12

14

07 09 11 13

CA

BA

IQHY_y

QFD_y

QFT_y

QPH_y

QPP_y

P_y

PQ_y

1413

15

10

12

14

16

18

30 40 50 60 70

[DNA

] C n

gμL

[DNA]A nguL

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

27

Qualification of ddPCR as Fit-For-Purpose We have documented the requirements for establishing metrological traceability of ddPCR human genomic assay results expressed as copies per nanoliter of sample in a series of three peer-reviewed papers151617 Through the experimental evidence they are based upon these papers demonstrate that ddPCR can accurately enumerate the number of independently migrating genomic copies in an aqueous extract Our most recent paper establishes the metrological traceability of these ddPCR results when transformed into the nanogram per microliter units used by the forensic and clinical communities18 However our previous studies have not addressed a fundamental requirement for correctly interpreting digital PCR measurements In broad simplification end-point assay systems count the number of independently partitioning entities regardless of the number of copies the entities contain With assays that target just one site per HHGE entities that migrate as ssDNA will yield counts that are twice the value had they migrated as dsDNA30 Should some originally dsDNA be converted over time to ssDNA ddPCR measurements of the number of independently migrating entities would increase although the number of HHGE would be unchanged Measurements of mass [DNA] provided by absorbance spectrophotometry and ddPCR reflect different properties absorbance at 260 nm is sensitive to the spatial configuration of the nucleotides as well as their number while ddPCR counts independently migrating entities The SRM 2372a materials have absorbance spectra shown in Figure 8 consistent with being predominantly dsDNA The SRM 2372 materials now have the absorbance spectra shown in Figure 14 which are consistent with them being entirely ssDNA

Figure 14 Absorbance Spectra of SRM 2372 Components A and B dsDNA is denatured to ssDNA by diluting the source material with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to mass [DNA]

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372-A200720122012-NaOH121417

0

4

6

0

4

6

220 260 300 340

Wavelength nm

2372-B200720122012-NaOH121517

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

28

To ensure the validity of ddPCR-based value assignments it is thus necessary to evaluate the nature of the partitioning entities Should a solution of what is thought to be dsDNA contain an appreciable fraction of ssDNA then the ddPCR results would require adjustment31 To avoid further delay in making SRM 2372a available to the forensic community the following sections detail a series of experiments that have enabled us to conclude with high confidence that the SRM 2372a components have a dsDNA conformation and that for this material ldquoentitiesrdquo can be considered dsDNA ldquocopiesrdquo The remaining stock of the SRM 2372 components also have a dsDNA conformation however the strands in the old materials are much less robustly connected than they are in the new materials Digital Polymerase Chain Reaction Assays

Three nDNA assays were used HBB1 NEIF and ND6 The HBB1 and NEIF assays reproducibly yield very similar results for DNA extracts from many different donors The HBB1 target is at the tip of the short (p) arm of chromosome 11 The NEIF target is in the centromere region of chromosome 2 The ND6 assay was selected for use because it typically yields the lowest results of the ten nDNA assays used to certify the SRM 2372a components (see Figure 10 below) The ND6 target is about halfway down the long (q) arm of chromosome 6 Heat Treatments (Melting of the DNA into Single Strands)

In all experiments the non-heated control and the heat-treated diluted solutions were stored in the same type of PCR reaction tube The starting point for the heat treatment was to determine if there was a difference between snap-cooling on ice and slower cooling at 25 degC Two tubes of the diluted material were placed in a GeneAmp 9700 thermal cycler (Applied Biosystems) and cycled up to 98 degC and held for 15 min After 15 min one tube was removed and placed on ice while the remaining tube was cooled at 1 degCs in the thermal cycler to 25 degC then placed on ice until prepared for ddPCR Additional heat treatments included cycling samples on a GeneAmp 9700 thermal cycler

a) from 98 degC for 5 min then cooling to 25 degC for 2 min During the 25 degC hold a randomly selected tube was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 45 min at 98 degC

b) from 94 degC for 1 min then cooling to 25 degC for 1 min During the 1 min hold at 25 degC a random sample was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 9 min at 94 degC

c) from 90 degC for 1 min then cooling to 25 degC for 1 min for a total of 20 cycles Samples were pulled at 25 degC after 1 min 3 min 5 min 10 min and 20 min and placed on ice

After the above experiments heat treatments were performed using a ProFlex thermal cycler This thermal cycler has six temperature zones that can be le 5 degC apart This enables heating

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

29

tubes at the same time but at different temperatures After treatment tubes were cooled at 5 degCs to 4 degC in the thermal cycler While designed for use with PCR plates a MicroAmp 96-well tray for the VeriFlex block enabled its use with single PCR reaction tubes Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a

As a first step in evaluating whether the DNA in the SRM materials disperse as dsDNA or ssDNA a direct comparison of the five materials was made using the HBB1 and NEIF assays Table 18 reports summary statistics for the ddPCR measurements reported as λ copiesdroplet and the estimated mass [DNA] values from the ddPCR results18 from the conventional dsDNA relationship7 and from the conventional ssDNA relationship after denaturing with 04 molL NaOH89 While the mass [DNA] values differ somewhat between the three methods there is no evidence for the SRM materials segregating exclusively or even mostly as ssDNA The extent of the hypochromic interactions among the nucleotides does not accurately reflect the extent to which the two strands of dsDNA have separated

Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a

λ copies per droplet λ copies per droplet ngmicroL Sample Assay Nrep

a Mean CVb ncmbc Mean CVb ddPCRd dsDNAe ssDNAf

2372-A HBB1 4 03645 19 8 03644 20 566g 524 57 2372-A NEIF 4 03644 23 2372-B HBB1 4 03513 23 8 03487 35 580h 536 61 2372-B NEIF 4 03461 47 2372a-A HBB1 4 02603 27 8 02649 30 474 465 454 2372a-A NEIF 4 02694 24 2372a-B HBB1 4 02908 34 8 02891 37 517 556 534 2372a-B NEIF 4 02873 45 2372a-C HBB1 4 02757 28 8 02711 28 485 502 474 2372a-C NEIF 4 02665 16

a Number of technical replicates for the given assay b CV = 100(standard deviation)mean c Combined number of technical replicates ignoring potential differences between the assays d Estimated from the ddPCR λ measurements listed in this Table18 e Estimated from the conventional relationship between the mass [DNA] of dsDNA and absorbance at

260 nm7 using absorbance spectra acquired shortly after solutions were produced f Estimated from the relationship between the mass [DNA] of ssDNA and absorbance at 260 nm after 11

dilution with 04 molL NaOH89 g Values adjusted to measured (132 plusmn19) volume loss of archived units due to evaporation h Values adjusted to measured (71 plusmn25) volume loss of archived due to evaporation

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

30

Quantitative Evaluation of ssDNA Content

Absorbance spectrophotometry is routinely used to quantitatively assess DNA tertiary structure3233 On storage ldquoat physiological pH the intrastrand repulsion between the negatively charged phosphate groups forces the double helix into more rigid rodlike conformationrdquo34 suggesting a mechanism for the observed slow changes in the SRM 2372 materials As demonstrated in Table 18 absorbance at 260 nm does not necessarily correlate with ddPCR behavior Yet ldquofurthermore the repulsion between phosphate groups on opposite strands tends to separate the complementary strandsrdquo34 Therefore a method that is independent of hypochromism is needed to adequately evaluate the proportion (if any) of ldquotruerdquo (completely disjoint) ssDNA in the SRM materials Circular dichroism spectroscopy is sensitive to the conformational structures of DNA in solution35 and different DNA configurations migrate differentially in agarose gel electrophoresis7 Both techniques have been used to qualitatively discriminate ssDNA from dsDNA in ldquopurerdquo single-sequence materials but neither technique appears to be able to quantify small proportions of ssDNA in a complex multi-chromosome genomic material Wilson and Ellison have proposed a real-time chamber digital PCR (cdPCR) technique that can estimate excess proportions of ssDNA copies from the crossing-threshold distribution31 However their method is insufficiently sensitive for low ssDNA proportions It is well known that heating DNA beyond the melting temperature separates mammalian dsDNA into ssDNA36 albeit we have shown using cdPCR that boiling dsDNA for some minutes also renders some target sites non-amplifiable15 In the following section we document a ddPCR-based approach for estimating the proportion if any of ssDNA in a sample that involves less destructive heating While the method is not yet optimized the results from these preliminary studies demonstrate that the SRM 2372 and 2372a components disperse into the ddPCR droplets as dsDNA ddPCR Measurement of ssDNA Proportion Given two otherwise identical aliquots of a given DNA material one subjected to some treatment and the other untreated the ratio between the average number of copies per droplet of treated material and that of its untreated sibling λtλu quantitatively estimates the change in the number of independently dispersing entities Should the treatment reduce the number of accessible amplifiable targets (AAT) λtλu will be less than 1 Reduction in AAT may result from damage to the target itself andor conformational changes to the fragment containing the target that render the target inaccessible or non-amplifiable16 Assuming no denaturation of dsDNA to ssDNA either because of an ineffective treatment or the material having started as ldquopurerdquo ssDNA if the treatment does not itself damage targets then λtλu will equal 1 Should treatment convert at least some dsDNA to ssDNA without damaging targets the number of AAT will increase and λtλu will be greater than 1 Assuming complete conversion of ldquopurerdquo dsDNA to ldquopurerdquo ssDNA without other damage then λtλu will equal 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

31

Values greater than 1 but less than 2 indicate some conversion of dsDNA to ssDNA but also some combination of (a) incomplete denaturation of dsDNA to ssDNA (b) partial renaturation of ssDNA to dsDNA after treatment but before ddPCR analysis (c) damage to targets during the process or ndash of greatest interest ndash (d) the original sample material was an impure mixture of dsDNA and ssDNA The following discussions graphically summarize sets of λtλu followed by tabular listings of the relevant experimental conditions ddPCR measurements and the numerical λtλu values The information in these tables includes

Plate Index Experiment identifier indicates date performed in YYYYMMDD format Material Sample material SRM 2372-A 2372-B 2372a-A 2372a-B or 2372a-C Assay ddPCR assay HBB1 NEIF or ND6 degC Denaturation temperature in degC Min Length of time the treated material was held at denaturation temperature in

minutes Cool Method used to cool treated material from denaturation temperature Snap

Slow R1 or R5 where ldquoSnaprdquo indicates cooling on ice ldquoSlowrdquo cooling at room temperature to 25 degC ldquoR1rdquo cooling in a thermal cycler to 25 degC at 1 degCs and ldquoR5rdquo cooling in a thermal cycler to 4 degC at 5 degCs

F Dilution factor expressed as (volume Material) per (final volume of diluted solution) 12 14 or 18

n Number of technical replicates Mean Mean of technical replicates CV Coefficient of Variation expressed as a percentage 100 Standard Deviation

Mean

λtλu the copies per droplet result of the treated solution λt divided by the copies per droplet result of its untreated sibling λu

u(λtλu) Standard uncertainty of λtλu 119906119906(120582120582t 120582120582ufrasl ) = 120582120582t 120582120582ufrasl CV(120582120582t)100

2

+ CV(120582120582u)100

2

Proof of Principle Figure 15 summarizes two replicate sets of measurements performed to assess whether the SRM 2372 and 2372a materials respond similarly whether the denatured materials renature if allowed to cool slowly and whether results are (reasonably) repeatable Table 19 details the experimental conditions The λtλu values for the four materials are similar all much greater than 1 but not approaching 2 Most of the ratios are on or close to the ldquoslow cooling is the same as snap coolingrdquo line suggesting that renaturation is not a significant concern While the replicate results for the SRM 2372-A and 2372a-A materials do not overlap as closely as those for SRM 2372-B and 2372a-B similar treatment conditions produce similar results

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

32

Figure 15 Comparison of Snap-Cooling to Slow-Cooling

Symbols mark the means of replicate measurements made using the same dilutions on successive days with bars representing plusmn one standard combined uncertainty Results are labeled by material ldquoArdquo and ldquoBrdquo for SRM 2372 components and ldquoaArdquo ldquoaBrdquo and ldquoaCrdquo for SRM 2372a components The diagonal black line represents equality between the snap- and slow-cooling treatments

Table 19 Comparison of Snap-Cooling to Slow-Cooling

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171218 2372-A HBB1 14 4 03830 29

dd20171218 2372-A HBB1 98 15 Slow 14 2 05644 18 1474 0051 dd20171218 2372-A HBB1 98 15 Snap 14 4 05852 15 1528 0051

dd20171218 2372a-A HBB1 14 4 03029 18

dd20171218 2372a-A HBB1 98 15 Slow 14 4 04430 21 1463 0040 dd20171218 2372a-A HBB1 98 15 Snap 14 4 04454 33 1471 0055

dd20171218 2372-B HBB1 14 4 03594 31

dd20171218 2372-B HBB1 98 15 Slow 14 4 05806 18 1615 0058 dd20171218 2372-B HBB1 98 15 Snap 14 4 05825 13 1620 0055

dd20171218 2372a-B HBB1 14 4 03383 15

dd20171218 2372a-B HBB1 98 15 Slow 14 4 05043 38 1491 0061 dd20171218 2372a-B HBB1 98 15 Snap 14 4 05254 20 1553 0039

dd20171219 2372-A HBB1 14 4 03357 23

dd20171219 2372-A HBB1 98 15 Slow 14 4 05340 16 1591 0046 dd20171219 2372-A HBB1 98 15 Snap 14 4 05340 12 1591 0042

A

aA

B

aB

A

aA

B

aB

14

15

16

17

14 15 16 17

λ tλ

u Sn

ap C

oole

d

λtλu Slow Cooled

Rep 1Rep 2Snap = Slow

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

33

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171219 2372a-A HBB1 14 4 02735 24

dd20171219 2372a-A HBB1 98 15 Slow 14 4 04151 03 1518 0037 dd20171219 2372a-A HBB1 98 15 Snap 14 4 03918 21 1433 0045 dd20171219 2372-B HBB1 14 4 03346 22

dd20171219 2372-B HBB1 98 15 Slow 14 4 05476 17 1637 0045 dd20171219 2372-B HBB1 98 15 Snap 14 4 05480 29 1638 0059 dd20171219 2372a-B HBB1 14 4 03158 19

dd20171219 2372a-B HBB1 98 15 Slow 14 4 04798 16 1519 0038 dd20171219 2372a-B HBB1 98 15 Snap 14 4 04856 17 1538 0039

Time at Denaturation Temperature Figure 16 summarizes λtλu for the SRM 2372-B material as a function of denaturation temperature and the cumulative time the solution is held at that temperature Table 20 details the experimental conditions

Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures

The symbols represent the ratio of the mean result of three technical replicates of the untreated and each of the treated solutions the bars represent combined standard uncertainties The lines denote regression fits to the function λtλu = atimese-bt where t is the cumulative time Values and the standard uncertainties of the coefficients are listed in the legend

12

14

16

18

20

0 5 10 15 20

λ t λ

u

Cummulative Treatment Time min

degC a u (a ) b u (b )90 1804 0039 00198 0000494 1788 0029 00099 0002998 1855 0019 00067 00019

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

34

For all three of the denaturation temperatures studied the relationship between λtλu and cumulative time t can be modeled as

λtλu = ae-bt

where a is the extrapolated value of λtλu at t = 0 and b is the rate of change in AAT per minute As expected the rate of loss due to the treatment increases with increasing treatment temperature and cumulative time at that temperature37 Given the much longer treatment time for the 98 degC series the agreement among the three t = 0 parameter (a) values is remarkable This suggests that the first 1 min of treatment efficiently separates dsDNA into ssDNA The pooled relative standard uncertainty of these regression parameter is 17 somewhat smaller than the 25 pooled relative standard uncertainties of the individual λtλu However the plusmnstandard uncertainty interval of the 1-min λtλu results for the 94 degC and 90 degC both overlap the extrapolated t = 0 value suggesting that reliable results can be obtained using a single treatment of short duration

Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171228 2372-B HBB1 14 3 03650 17 dd20171228 2372-B HBB1 98 5 R1 14 3 06157 12 1687 0035 dd20171228 2372-B HBB1 98 10 R1 14 3 05493 08 1505 0028 dd20171228 2372-B HBB1 98 15 R1 14 3 05019 13 1375 0029 dd20171228 2372-B HBB1 98 20 R1 14 3 04583 21 1256 0034 dd20171228 2372-B HBB1 98 25 R1 14 3 04154 32 1138 0041 dd20171228 2372-B HBB1 98 30 R1 14 3 03784 14 1037 0023 dd20171228 2372-B HBB1 98 35 R1 14 3 03408 10 0934 0019 dd20171228 2372-B HBB1 98 40 R1 14 3 02980 01 0816 0014 dd20171228 2372-B HBB1 98 40 R1 14 3 02815 08 0771 0015

dd20171229 2372-B HBB1 14 3 03708 27

dd20171229 2372-B HBB1 94 1 R1 14 3 06598 26 1779 0056 dd20171229 2372-B HBB1 94 2 R1 14 3 06553 32 1767 0063 dd20171229 2372-B HBB1 94 3 R1 14 3 06301 20 1699 0045 dd20171229 2372-B HBB1 94 4 R1 14 3 06252 14 1686 0037 dd20171229 2372-B HBB1 94 5 R1 14 3 06577 14 1773 0039 dd20171229 2372-B HBB1 94 6 R1 14 3 06156 07 1660 0030 dd20171229 2372-B HBB1 94 7 R1 14 3 06118 10 1650 0033 dd20171229 2372-B HBB1 94 8 R1 14 3 06256 12 1687 0035 dd20171229 2372-B HBB1 94 9 R1 14 3 05999 10 1618 0032

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

35

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180104 2372-B HBB1 14 4 03713 25 dd20180104 2372-B HBB1 90 1 R1 14 3 06916 22 1863 0052 dd20180104 2372-B HBB1 90 3 R1 14 3 06465 06 1741 0032 dd20180104 2372-B HBB1 90 5 R1 14 3 06431 20 1732 0045 dd20180104 2372-B HBB1 90 10 R1 14 3 06014 25 1620 0049 dd20180104 2372-B HBB1 90 15 R1 14 3 06024 10 1623 0032 dd20180104 2372-B HBB1 90 20 R1 14 3 06003 14 1617 0035

Temperature Figure 17 summarizes the effect of denaturation temperature using a single treatment of 05 min on the 2372-B and 2372a-B materials Table 21 details the experimental conditions

Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

The symbols represent the ratio of the mean result of four technical replicates of the untreated and three replicates of the treated solutions the bars represent combined standard uncertainties In all cases the duration of the treatment was 05 min The lines denote regression fits to the linear function λtλu = a +bT where Tis the denaturation temperature

The SRM 2372 and 2372a materials respond very differently to changes in denaturation temperature The older spectrophotometrically ssDNA material suffers somewhat greater loss in λtλu as temperature increases The new spectrophotometrically dsDNA material suffers much greater loss in λtλu as temperature decreases

14

15

16

17

18

19

20

82 86 90 94 98

λ tλ

u

Temperature degC

2372-B

2372a-B

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

36

Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180109 2372-B HBB1 14 4 03577 25 dd20180109 2372-B HBB1 100 05 R5 14 3 06314 16 1765 0051

dd20180104a 2372-B HBB1 14 4 03577 05

dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032 dd20180104a 2372-B HBB1 94 05 R5 14 3 06459 25 1806 0046 dd20180104a 2372-B HBB1 90 05 R5 14 2 06591 08 1842 0017 dd20180104a 2372-B HBB1 88 05 R5 14 3 06589 05 1842 0013 dd20180104a 2372-B HBB1 84 05 R5 14 3 06729 10 1881 0022 dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

We hypothesize that in the older SRM 2372 material the between-strand bond is relatively fragile and can be completely severed at relatively low temperature while avoiding much temperature-dependent AAT loss In contrast the between-strand bond in the SRM 2372a material requires higher temperature to completely disassociate the strands This suggests that there is not a sample-independent ldquooptimumrdquo combination of denaturation temperature and treatment duration However 98 degC for 05 min appears to provide comparable results for these two materials DNA Concentration in the ddPCR Solution Figure 18 summarizes the effect of [DNA] in the ddPCR solution with SRM 2372-B at a denaturation temperature of 98 degC and treatment duration of 05 min Table 22 details the experimental conditions

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

37

Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution

The solid black symbols represent the ratio of the mean result of four technical replicates of the untreated and treated solutions with the λtλu values shown on the left-hand vertical axis the bars represent combined standard uncertainties The solid blue circles represent λu and the open red circles λt with values shown on the right-hand vertical axis at this graphical scale the standard deviations of the technical replicates are covered by the symbol The diagonal blue and red lines represent the log-linear relationships between just the 18 and 14 λ values

The primary effect of [DNA] appears to be the loss of linearity in λt with copiesdroplet values greater than about 1 copiesdroplet There may be some advantage to use somewhat greater dilutions than the routine 14 however the small increase in λtλu may be offset by the increased variability with small λu This suggests the use of dilutions that give λu values of (01 to 04) copiesdroplet

Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180117 2372-B HBB1 12 4 07127 11 dd20180117 2372-B HBB1 98 05 R5 12 4 12331 14 1730 0031

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 3 06151 15 1843 0040

dd20180117 2372-B HBB1 18 4 01549 18 dd20180117 2372-B HBB1 98 05 R5 18 4 02897 20 1870 0051

12 14 18

170

175

180

185

190

02

04

06

08

101214

λ T

arge

t D

ropl

et

λ tλ

u

Dilution Factor

λ₊λᵤλᵤλ₊

λtλu

λu

λt

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

38

Equivalence of Assays Figure 19 summarizes the agreement between the HBB1 NEIF and ND6 assays at a denaturation temperature of 98 degC and treatment duration of 05 min Table 23 details the experimental conditions

Figure 19 Comparison of Assays

The symbols compared the λtλu results from the HBB1 and either the NEIF or ND6 assays Each λtλu value is the ratio of the mean result of four technical replicates of an untreated and three or four replicates of the treated solutions The bars represent combined standard uncertainties The diagonal line represents equality between the assays

There is little or no difference between the λtλu results from the HBB1 NEIF and ND6 assays suggesting that the chromosomal location of the assay target has little influence Since the NEIFHBB1 comparison includes results for both the SRM 2372 and 2372a materials this equivalence is not material-specific

14

16

18

20

14 16 18 20

λ tλ

u NE

IF o

r ND6

λtλu HBB1

NEIFHBB1

ND6HBB1

2nd Assay = HBB1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

39

Table 23 Comparison of Assays

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu)

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045 dd20180105 2372a-B ND6 94 05 R5 14 3 05526 36 1711 0074 dd20180105 2372a-B ND6 90 05 R5 14 3 05326 18 1649 0051 dd20180105 2372a-B ND6 88 05 R5 14 3 05077 22 1572 0052 dd20180105 2372a-B ND6 86 05 R5 14 3 05038 23 1560 0053 dd20180105 2372a-B ND6 84 05 R5 14 3 04906 10 1519 0041

dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

40

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials

Figure 20 summarizes the currently available λtλu results obtained by denaturing at 98 degC for 05 min rapidly cooling in the thermal cycler with dilutions that provide asymp 03 copiesdroplet The figure also displays the consensus result and its 95 confidence region estimated using the Hierarchical Bayes method provided in the NIST Consensus Builder (NICOB) system3839 Table 24 details the experimental conditions

Figure 20 Evaluation of Limiting Ratio

The symbols represent the ratio of the mean result of three or four technical replicates of the untreated and treated solutions the bars represent combined standard uncertainties The thick horizontal line represents a consensus value covering all displayed results the dotted horizontal lines bracket the 95 confidence interval about the consensus value

The Hierarchical Bayes consensus result for these values is 183 with a 95 confidence interval of plusmn003 All but three of the 16 λtλu values are within one standard uncertainty of this consensus value with most of the λtλu means lying within the 95 confidence interval This suggests that there are no statistically significant differences between the SRM 2372 and 2372a materials with regard to ssDNA proportion

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

17

18

19

20

λ tλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

41

Table 24 Evaluation of Limiting Ratio

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180104a 2372-B HBB1 14 4 03577 05 dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032

dd20180108 2372-B HBB1 14 3 03463 09 dd20180108 2372-B HBB1 98 05 R5 14 3 06562 14 1895 0031

dd20180109 2372-B HBB1 14 3 03577 25 dd20180109 2372-B HBB1 98 05 R5 14 3 06495 21 1816 0059

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 4 06151 15 1843 0040

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

42

The agreement among the measured λtλu is more directly visualized using estimated unilateral ldquoDegrees of Equivalencerdquo d Each d is the absolute difference between the measured and consensus values

119889119889 = |120582120582119905119905 120582120582119906119906frasl minus consensus|

and is associated with a 95 confidence interval U95(d) that combines the uncertainty estimates of the measured and consensus values The U95(d) calculation details are specific to the consensus estimation method39 Values of d plusmn U95(d) that include 0 indicate that the λtλu result does not differ from the majority at about a 95 level of confidence Figure 21 displays the d for the λtλu results for the Hierarchical Bayes consensus value39 By this figure of merit all the λtλu can be considered as members of a single population

Figure 21 Degrees of Equivalence

The symbols represent the Degree of Equivalence d for the λtλu results relative to the Hierarchical Bayes consensus value The bars represent 95 level of confidence uncertainty intervals d plusmn U95(d) The horizontal line marks the d = 0 ldquono differencerdquo value Results with d plusmn U95(d) that include this line are not significantly different from the majority

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

-024

-018

-012

-006

000

006

012

018

024

Degr

ees o

f Equ

ival

ence

λtλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

43

Conclusions and Recommendations With high confidence all tubes of the SRM 2372a solution have the same nDNA copy number content within measurement repeatability Additionally within measurement repeatability all tubes of the SRM 2372a have the same mtDNAnDNA With high confidence the nDNA copy number content of the SRM 2372a solution is thermally stable from 4 degC to 37 degC over an extended period The solution should not be shipped or stored below 4 degC With high confidence the ddPCR measurements used to establish the entity and mass concentrations and the mtDNAnDNA ratios are adequately precise and unbiased With high confidence the DNA in the SRM 2372a solutions is now dsDNA and can be expected to remain dsDNA for at least ten years when stored at 4 degC Recommended Certification Values

Since the component materials are believed to be homogeneous are of very similar composition and were prepared and processed similarly the observed differences in the measured entity concentration uncertainties are likely artifacts of the measurement process To limit over-interpretation of these differences we recommend that the copy per nanoliter values of the three components be assigned same 95 relative uncertainty interval (CV95) based on the largest of the measured CV95 values (93 ) rounded up to 10 Since the ngμL transformation does not appreciably inflate this relative uncertainty we recommend that the mass concentration values also be assigned CV95 of 10 Similarly we recommend that the mtDNAnDNA ratios be rounded to integer values all three components be assigned the same CV95 and that this uncertainty be based on the largest (21 ) of the measured CV95 rounded up to 25 Table 25 lists the values that we recommend be used in the Certificate of Analysis

Table 25 Recommended Values and 95 Uncertainties for Certification

Component Units Value U95(Value) a A Copies per nanoliter 151 15 B Copies per nanoliter 175 18 C Copies per nanoliter 145 15 A ngμL 498 50

B ngμL 578 58 C ngμL 479 48 A mtDNAnDNA 174 4

B mtDNAnDNA 206 5 C mtDNAnDNA 279 7

a The interval Value plusmn U95(Value) is believed with about 95 confidence to contain the true value of the measurand

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

44

Use of SRM 2372a for Value-Assigning In-House Reference Materials

The within-component and between-assay variability observed for the commercial qPCR assays is not unexpected since different qPCR methods exploit qualitatively different molecular targets Given the intrinsic variability of the human genome the prospect of developing a human DNA source material that responds equally to all current and future qPCR assays is improbable All three components should be used to elucidate potential bias when using these materials to value assign in-house DNA reference and control materials for qPCR assays

Certificate of Analysis In accordance with ISO Guide 31 2000 a NIST SRM certificate is a document containing the name description and intended purpose of the material the logo of the US Department of Commerce the name of NIST as a certifying body instructions for proper use and storage of the material certified property value(s) with associated uncertainty(ies) method(s) used to obtain property values the period of validity if appropriate and any other technical information deemed necessary for its proper use A Certificate is issued for an SRM certified for one or more specific physical or engineering performance properties and may contain NIST reference information or both values in addition to certified values A Certificate of Analysis is issued for an SRM certified for one or more specific chemical properties Note ISO Guide 31 is updated periodically check with ISO for the latest version [httpswwwnistgovsrmsrm-definitions] For the most current version of the Certificate of Analysis for NIST SRM 2372a Human DNA Quantification Standard please visit httpswww-snistgovsrmorsview_detailcfmsrm=2372a

References

1 National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research Department of Health Education and Welfare (1978) Belmont Report Ethical Principles and Guidelines for the Protection of Human Subjects of Research DHEW Publication No (OS) 78-0012 httpvideocastnihgovpdfohrp_belmont_reportpdf

2 CDCNIH Biosafety in Microbiological and Biomedical Laboratories 5th ed HHS publication No (CDC) 21-1112 Chosewood LC Wilson DE Eds US Government Printing Office Washington DC (2009) httpwwwcdcgovbiosafetypublicationsbmbl5indexhtm

3 Duewer DL Kline MC Redman JW Newall PJ Reeder DJ NIST Mixed Stain Studies 1 and 2 Interlaboratory Comparison of DNA Quantification Practice and Short Tandem Repeat Multiplex Performance with Multiple-Source Samples J Forensic Sci 200146(5)1199-1210 PMID 11569565

4 Kline MC Duewer DL Redman JW Butler JM NIST Mixed Stain Study 3 DNA quantitation accuracy and its influence on short tandem repeat multiplex signal intensity Anal Chem 200375(10)2463-2469 httpsdoiorg101021ac026410i

5 Kline MC Duewer DL Redman JW Butler JM Results from the NIST 2004 DNA Quantitation Study J Forensic Sci 200550(3)571-578 PMID 15932088

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

45

6 Kline MC Duewer DL Travis JC Smith MV Redman JW Vallone PM Decker AE Butler JM Production and Certification of Standard Reference Material 2372 Human DNA Quantitation Standard Anal Bioanal Chem 2009394(4)1183-1192 httpsdoiorg101007s00216-009-2782-0

7 Sambrook J and Russell DW (2001) Molecular Cloning a Laboratory Manual Cold Spring Harbor Laboratory Press Cold Spring Harbor New York

8 EU-JRC Protocol NK603 ndash Community Reference Laboratory Event-specific method for the quantification of maize line NK603 using real-time PCR Section 33 Spectrophotometric Measurement of DNA Concentration httpgmo-crljrceceuropaeusummariesNK603-WEB-ProtocolValidationpdf

9 ISO 215712005(E) Foodstuffs ndash Methods of analysis for the detection of genetically modified organisms and derived products ndash Nucleic acid extraction Annex B Methods for quantitation of extracted DNA pp 34-36 International Organization for Standardization Geneva (2005)

10 Certificate of Analysis Standard Reference Material 2372 Human DNA Quantitation Standard (2013) Gaithersburg MD USA Standard Reference Materials Program National Institute of Standards and Technology httpswww-snistgovsrmorsview_certcfmsrm=2372

11 Vogelstein B Kinzler KW Digital PCR Proc Natl Acad Sci USA 1999969236-9241 12 Hindson BJ Ness KD Masquelier DA Belgrader P Heredia NJ Makarewicz AJ Bright IJ Lucero

MY Hiddessen AL Legler TC Kitano TK Hodel MR Petersen JF Wyatt PW Steenblock ER Shah PH Bousse LJ Troup CB Mellen JC Wittmann DK Erndt NG Cauley TH Koehler RT So AP Dube S Rose KA Montesclaros L Wang S Stumbo DP Hodges SP Romine S Milanovich FP White HE Regan JF Karlin-Neumann GA Hindson CM Saxonov S Colston BW High-Throughput Droplet Digital PCR System for Absolute Quantitation of DNA Copy Number Anal Chem 2011838604-8610 httpsdoiorg101021ac202028g

13 Pinheiro LB Coleman VA Hindson CM Herrmann J Hindson BJ Bhat S Emslie KR Evaluation of a Droplet Digital Polymerase Chain Reaction Format for DNA Copy Number Quantification AnalChem 2012841003-1011 httpsdoiorg101021ac202578x

14 Corbisier P Pinheiro L Mazoua S Kortekaas AM Chung PY Gerganova T Roebben G Emons H Emslie K DNA copy number concentration measured by digital and droplet digital quantitative PCR using certified reference materials Anal Bioanal Chem 2015407(7)1831-1840 httpsdoiorg101007s00216-015-8458-z

15 Duewer DL Kline MC Romsos EL Real-time cdPCR opens a window into events occurring in the first few PCR amplification cycles Anal Bioanal Chem 2015407(30)9061-9069 httpsdoiorg101007s00216-015-9073-8

16 Kline MC Romsos EL Duewer DL Evaluating Digital PCR for the Quantification of Human Genomic DNA Accessible Amplifiable Targets Anal Chem 201688(4)2132-2139 httpsdoiorg101021acsanalchem5b03692

17 Kline MC Duewer DL Evaluating Droplet Digital Polymerase Chain Reaction for the Quantification of Human Genomic DNA Lifting the Traceability Fog Anal Chem 201789(8)4648-4654 httpsdoiorg101021acsanalchem7b00240

18 Duewer DL Kline MC Romsos EL Toman B Evaluating Droplet Digital PCR for the quantification of human genomic DNA Conversion from Copies per Nanoliter to Nanogram Genomic DNA per Microliter Anal Bioanal Chem 2018in press httpsdoiorg101007s00216-018-0982-1

19 Dagata JA Farkas N Kramer JA Method for measuring the volume of nominally 100 μm diameter spherical water in oil emulsion droplets NIST Special Publication 260-184 2016 httpsdoiorg106028NISTSP260-184

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

20 ISO 170342016 General requirements for the competence of reference material producers International Organization for Standardization Geneva (2016)

21 NCBI Standard Nucleotide BLAST httpsblastncbinlmnihgovBlastcgiPAGE_TYPE=BlastSearch

22 Timken MD Swango KL Orrego C Buoncristiani MR A Duplex Real-Time qPCR Assay for the Quantification of Human Nuclear and Mitochondrial DNA in Forensic Samples Implications for Quantifying DNA in Degraded Samples J Forensic Sci 200550(5)1044-60 PMID 16225209

23 Walker JA Hedges DJ Perodeau BP Landry KE Stoilova N Laborde ME Shewale J Sinha SK Batzer MA Multiplex polymerase chain reaction for simultaneous quantitation of human nuclear mitochondrial and male Y-chromosome DNA application in human identification Anal Biochem 200533789ndash97 httpsdoiorg101016jab200409036

24 Kavlick MF Lawrence HS Merritt RT Fisher C Isenberg A Robertson JM Budowle B Quantification of Human Mitochondrial DNA Using Synthesized DNA Standards J Forensic Sci 201156(6) 1457ndash1463 httpsdoiorg101111j1556-4029201101871x

25 Stulnig TM Amberger A Exposing contaminating phenol in nucleic acid preparations BioTechniques 199416(3) 403-404

26 Majumdar N Wessel T Marks J Digital PCR Modeling for Maximal Sensitivity Dynamic Range and Measurement Precision PLoS ONE 201510(3)e0118833 httpsdoi101371journalpone0118833

27 Tellinghuisen J Partition Volume Variability in Digital Polymerase Chain Reaction Methods Polydispersity Causes Bias but Can Improve Precision Anal Chem 2016881283-12187 httpsdoiorg101021acsanalchem6b03139

28 Lunn DJ Spiegelhalter D Thomas A Best N (2009) Statistics in Medicine 283049ndash3082 29 De Biegravevre P Dybkaer R Fajgelj A Hibbert DB Metrological traceability of measurement results in

chemistry Concepts and implementation Pure Appl Chem 201183(10)1873-1935 httpsdoiorg101007s00769-011-0805-y

30 Huggett JF Foy CA Benes V Emslie K Garson JA Haynes R Hellemans J Kubista M Mueller RD Nolan T Pfaffl MW Shipley GL Vandesompele J Wittwer CT Bustin SA The Digital MIQE Guidelines Minimum Information for Publication of Quantitative Digital PCR Experiments Clin Chem 201359(6)892-902 httpsdoiorg101373clinchem2013206375

31 Wilson PJ Ellison SLR Extending digital PCR analysis by modelling quantification cycle data BMC Bioinfo 201617421 httpsdoiorg101186s12859-016-1275-3

32 Wang X Son A Effects of pretreatment on the denaturation and fragmentation of genomic DNA for DNA hybridization Environ Sci Processes Impacts 2013152204 httpsdoiorg101039c3em00457k

33 Nwokeoji AO Kilby PM Portwood DE Dickman MJ Accurate Quantification of Nucleic Acids Using Hypochromicity Measurements in Conjunction with UV Spectrophotometry Anal Chem 201789(24)13567-13574 https101021acsanalchem7b04000

34 Farkas DH Kaul KL Wiedbrauk DL Kiechle FL Specimen collection and storage for diagnostic molecular pathology investigation Arch Pathol Lab Med 1996120(6)591-6

35 Kypr J Kejnovskaacute I Renčiuk D Vorličkovaacute M Circular dichroism and conformational polymorphism of DNA Nucleic Acids Res 200937(6)1713-1725 httpsdoiorg101093nargkp026

36 Strachan T Read AP (1999) Human Molecular Genetics 2nd Ed John Wiley amp Sons NY

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

37 Bhat S McLaughlin JL Emslie KR Effect of sustained elevated temperature prior to amplification on template copy number estimation using digital polymerase chain reaction Analyst 2011136724ndash32 httpsdoiorg101039c0an00484g

38 NIST Consensus Builder (NICOB) httpsconsensusnistgov 39 Koepke A Lafarge T Toman B Possolo A NIST Consensus Builder Userrsquos Manual

httpsconsensusnistgovNISTConsensusBuilder-UserManualpdf

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

Appendix A DNA Extraction Method The following method has been adapted from Miller et al (1988)A1 Procedure

1 Transfer approximately 5 mL of buffy coat into a 50 mL sterile polypropylene centrifuge tube and add 30 mL Red Blood Cell Lysis Buffer

2 Mix by gentle inversion and let stand at room temperature for 15 min

3 Centrifuge at approximately 209 rads (2000 rpm) for 20 min

4 Discard supernatant

5 Add 9 mL of Nucleic Lysis Buffer and re-suspend the pelleted cells by agitating with a plastic transfer pipette

6 Add 300 microL of 20 sodium dodecyl sulfate (SDS) and 18 mL of Protease K Solution (freshly prepared 2 mg to 5 mg Protease K per 1 mL Protease K Buffer)

7 Mix by inversion followed by incubation overnight or longer at 37 ordmC to 40 ordmC in a shaking water bath The length of incubation is determined by the time required to solubilize the cells present in the tube

8 Add 4 mL saturated ammonium acetate solution (in deionized water stored at 2 ordmC to 8 ordmC) Shake vigorously for 15 s

9 Centrifuge at approximately 471 rads (4500 rpm) for 10 min

10 Pipette supernatant into a 50 mL sterile polypropylene centrifuge tube and add two volumes of room temperature absolute ethanol Mix by gentle inversion until the DNA aggregates and floats to the top of the tube

11 Spool the aggregated DNA from the current tube into a fresh tube of 80 ethanol using a sterile inoculation loop At this step multiple tubes of precipitated DNA may be pooled to make a large lot of the purified material Mix by gentle inversion

12 Transfer the aggregated DNA to a fresh dry 50 mL polypropylene centrifuge tube Remove as much ethanol as possible by squeezing the DNA against the wall of the tube with a sterile inoculation loop Air-dry the material in a laminar flow hood

13 Resolubilize DNA in TE-4 buffer and store material at 4 ordmC Reagents

Red Blood Cell Lysis Buffer 144 mmolL ammonium chloride 1 mmolL sodium bicarbonate in deionized (DI) water Store at 2 ordmC to 8 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

Nucleic Lysis Buffer 10 mmolL Tris-HCl 400 mmolL sodium chloride 20 mmolL disodium ethylenediaminetetraacetic acid in DI Water adjust pH to 82 Store at 2 ordmC to 8 ordmC

Protease K Buffer

20 mmolL disodium ethylenediaminetetraacetic acid 10 (mass per volume) Sodium Dodecyl Sulfate (SDS) in DI water Store at 2 ordmC to 8 ordmC

Protease K

Sigma-Aldrich cat P6556 Tritirachium album lyophilized powder Store at ndash20 ordmC

Ammonium Acetate

192 molL ammonium acetate (saturated) in DI water Store at 2 ordmC to 8 ordmC

Absolute Ethanol (200 proof)

Store at room temperature Reference

A1 Miller SA Dykes DD Polesky HF A simple salting out procedure for extracting DNA from human nucleated cells Nucleic Acids Res 198816(3)1215 httpsdoiorg101093nar1631215

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

48

Appendix B Spectrophotometric Measurement Protocol The following is for use with a scanning recording spectrophotometer The procedures include the spectrophotometer calibration measurement of the ldquonativerdquo dsDNA and measurement of the NaOH denatured ssDNA The procedure for evaluating solutions after treatment with NaOH is adapted from references89 Spectrophotometer Calibration

Establish Baseline Set the spectrophotometer to acquire data from 220 nm to 345 nm in steps of 01 nm Zero the instrument with nothing (air) in the sample and reference cells Acquire an air-air baseline of the instrument with nothing (air) in the sample and reference cells If the trace has significant structure or excessive noise take corrective action If the trace is acceptable proceed with data collection Enable Verification of Wavelength Calibration SRM 2034 is a solution of holmium oxide in 10 perchloric acid contained in a sealed quartz cuvette delivering an intrinsic UVVis wavelength standard for absorption spectrophotometryB1 Evaluation of the location of peak maxima in the spectra of this solution enables verification of the spectrophotometers wavelength calibration Place SRM 2034 in the sample cell Create and name a file to hold the verification spectra Perform three replicate SRM 2034-to-air scans saving each scan to the file Remove SRM 2034 from the spectrophotometer and store appropriately Enable Verification of Absorbance Calibration SRM 2031 consists of three metal-on-fused silica filters held in cuvette-sized filter holders plus an empty filter holder The three filters are certified for transmittance and transmittance density (-log10(transmittance) effectively absorbance) at several wavelengths with nominal wavelength-independent transmittances of 10 30 and 90 Evaluation of the absorbance of each of these three filters at the certified 250 nm 280 nm and 340 nm wavelengths enables verification of the spectrophotometers absorbance calibration Place the empty SRM 2031 filter holder in the reference cell Remove the protective covers from the 10 transmittance filter and place in the sample cell Perform three replicate scans saving each scan to file Remove the filter from the spectrophotometer replace the protective covers and store appropriately Repeat for the 30 and 90 filters Close the file containing the verification spectra For ldquonativerdquo dsDNA Samples

Verify Optical Balance of Cuvettes The measurements of the DNA-containing samples are made using a 100 microL quartz cuvette in the sample cell compared to an optically matched cuvette in the reference cell The optical equivalence of these two cuvettes can be verified by filling both cuvettes with appropriate volumes of a sample solution matrix (TE-4 buffer) that does not contain DNA Fill two dry

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

49

clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes followed by a methanol rinse in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquonativerdquo dsDNA remove and clean both the reference cuvette and the sample cuvette For NaOH denatured ssDNA Samples

Prepare a 04 molL NaOH Stock Solution To a 15 mL disposable centrifuge tube add 016 g NaOH pellets add deionized (DI) water to the 10 mL line and mix until all the NaOH pellets have dissolved Centrifuge at 524 rads (5000 rpm) for 5 min For every 9 samples to be analyzed aliquot 1 mL of this 04 molL NaOH stock solution into a 15 mL screw top tube Use these tubes to prepare the NaOH blanks and DNA samples In a 15 mL screw cap vial mix 200 microL TE-4 buffer with 200 microL 04 molL sodium hydroxide solution to achieve a final NaOH concentration of 02 molL Vortex 5 s then centrifuge at 1257 rads (12 000 rpm) for 1 min Fill two dry clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Prepare the Sample Solutions For each unit of material to be measured (Tube or vial)

a) Vortex the unit for 5 s b) Centrifuge at 1257 rads (12 000 rpm) for 1 min c) Remove a 100 microL aliquot and place in a new labeled vial d) Add 100 microL of the freshly prepared 04 molL NaOH solution (with the same pipet) e) Vortex 5 s f) Centrifuge at 1257 rads (12 000 rpm) for 1 min

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

50

Measure Sample Solutions Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes rinse with methanol in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the NaOH denatured DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquodenaturedrdquo ssDNA close the file holding the spectra for this series Remove and clean both the reference cuvette and the sample cuvette After All Sample Measurements are Complete

Enable Verification of Spectrophotometer Stability The stability of the spectrophotometer over the course of the data collection can be evaluated by comparing spectra for SRM 2034 and 2031 taken at the start of data collection with spectra taken at the end of data collection Repeat the wavelength and absorbance verification measurements described above collecting the spectra in a suitably named file Clean Up Transfer all relevant data files from the spectrophotometers storage to a transfer device Clean the spectrophotometer work area as necessary and restore the SRMs and cuvettes to their appropriate storage locations Evaluation

Export the files to a spreadsheet Average the replicate spectra Subtract the average absorbance at A320 from the absorbance at A260 resulting in the corrected absorbance at A260 The conventional spectrophotometric estimate of the mass concentration of the dsDNA in the sample is achieved by multiplying the corrected absorbance by 50 The conventional estimate for the NaOH-treated ssDNA is achieved by multiplying the corrected absorbance by 37times2 where the ldquotimes2rdquo adjusts for the 1rarr2 volumetric dilution of the sample Note reliable spectrophotometric measurements require corrected A260 to be greater than 005 Reference

B1 Travis JC Acosta JC Andor G Bastie J Blattner P Chunnilall CJ Crosson SC Duewer DL Early EA Hengstberger F Kim C-S Liedquist L Manoocheri F Mercader F Metzdorf J Mito A Monardh LAG Nevask S Nilsson M Noeumll M Rodriguez AC Ruiacutez A Schirmacher A Smith MV Valencia G van Tonder N Zwinkels J Intrinsic Wavelength Standard Absorption Bands in Holmium Oxide Solution for UVvisible Molecular Absorption Spectrophotometry J Phys Chem Ref Data 200534(1)41-56

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

51

Appendix C SRM 2372a Statisticianrsquos report Extracted from Blaza Toman

November 30 2017 1 Introduction Candidate material for SRM 2372a is composed of three components labeled A B and C All components are human genomic DNA and prepared at NIST from buffy coat white blood cells from single-source anonymous donors Component A and B are single source (malefemale respectively) and Component C is a 1 male to 3 female mixture of two single-source donors Droplet digital PCR (ddPCR) assays were used to obtain measurements of the concentration of copies per microliter of sample Ten NIST-developed human nuclear DNA and three literature-derived human mitochondrial DNA PCR assays were used in this study

11 OpenBUGS Analysis All parameters in the following sections were estimated using Markov Chain Monte Carlo (MCMC) implemented in the OpenBUGS freewareC1 OpenBUGS is a software application for the Bayesian analysis of complex statistical models

12 OpenBUGS Exemplar Output All following sections list the OpenBUGs model (ldquocoderdquo) initialization values (ldquoinitsrdquo) if they are needed and the data used to evaluate the various data sets discussed in the body of this report homogeneity stability nDNA certification and mtDNAnDNA value assignment The provided information in each of the following sections is sufficient to produce estimates but the estimates may not exactly reproduce the exemplar results provided The MCMC paradigm does not produce ldquoanalyticalrdquo values but approaches them asymptotically as the number o model updates increases All exemplar results presented here are based upon the following bull 10-fold ldquothinningrdquo to minimize autocorrelation there were 10 MCMC estimates for every

model update used in parameter estimation bull a ldquoburn inrdquo of 5 000 updates before beginning parameter estimation and bull 10 000 model updates

The following statistics are provided in the tables listing the exemplar results

mean arithmetic mean of the parameterrsquos model update values sd standard deviation of the parameterrsquos model update values MC_error Monte Carlo standard deviation of the mean meanradic(number of model updates) val25pc 25th percentile of the model update values median 50th percentile of the model update values val975pc 975th percentile of the model update values startnumber of ldquoburn inrdquo updates sample number of model update values

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

52

2 Homogeneity assessment 21 Data

Samples of the material of all three components were drawn from 22 different boxes Two assays (NEIF ND6) were used to obtain 3 replicates of the measurement in copies per nanoliter for each sample and component For the complete homogeneity data see

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

22 Model The data were analyzed using a three-level Gaussian hierarchical ANOVA model

119910119910119894119894119894119894119894119894~119873119873120572120572119894119894119894119894119894119894 120590120590119908119908119894119894119905119905ℎ119894119894119894119894 119887119887119887119887119887119887 1198941198941198941198942

where ~N() denotes ldquodistributed as a normal distribution of some mean μ and variancerdquo 119894119894 denotes component (1 for A 2 for B 3 for C) 119895119895 denotes assay (1 for NEIF 2 for ND6) 119896119896 denotes box (1hellip22) and 120572120572119894119894119894119894119894119894~119873119873micro119894119894119894119894120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119887119887119887119887119887119887119894119894

2

All variance components were given a Half Cauchy prior as described in Reference C2 Note The Half Cauchy prior is specified by ldquo~dnorm(0 00016)I(0001)rdquo The ldquoIrdquo notation is for censoring It cuts off any random draws from the Normal distribution that are lower than 0001 An informative prior is needed when there are less than four data per distribution and the half Cauchy is a good general-purpose prior

23 OpenBUGS code and data Homogeneity code ModelBegin for(i in 13)xiNsb[i] ~ dnorm(0 00016)I(0001) chSqNsb[i] ~ dgamma(0505) sigb[i] lt- xiNsb[i]sqrt(chSqNsb[i]) for(i in 13)for(j in 12)mu[ij]~dnorm(010E-5) for(i in 13) muA[i]lt-mean(mu[i])sigbP[i]lt-100(muA[i]sqrt(sigb[i])) for(i in 13)for(j in 12)xiNs[ij] ~ dnorm(0 00016)I(0001) chSqNs[ij]~dgamma(0505) sigt[ij]lt-xiNs[ij]sqrt(chSqNs[ij]) sigtP[ij]lt-100(mu[ij]sqrt(sigt[ij])) for(i in 13)for(j in 12)for(k in 122)alpha[ijk]~dnorm(mu[ij]sigb[i]) for(idx in 1393)lamb[idx]~dnorm(alpha[cmp[idx]asy[idx]box[idx]]sigt[cmp[idx]asy[idx]]) ModelEnd Inits None required

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

53

Homogeneity Data lamb[] asy[] box[] cmp[] rep[] 1632 1 1 1 1 1529 1 1 1 2 1511 1 1 1 3 1616 2 1 1 1 1640 2 1 1 2 1582 2 1 1 3 1531 1 2 1 1 1525 1 2 1 2 1549 1 2 1 3 1615 2 2 1 1 1589 2 2 1 2 1601 2 2 1 3 1524 1 3 1 1 1540 1 3 1 2 1533 1 3 1 3 1582 2 3 1 1 1152 2 3 1 2 1555 2 3 1 3 1454 1 4 1 1 1458 1 4 1 2 1427 1 4 1 3 1627 2 4 1 1 1605 2 4 1 2 1607 2 4 1 3 1465 1 5 1 1 1540 1 5 1 2 1504 1 5 1 3 1607 2 5 1 1 1633 2 5 1 2 1586 2 5 1 3 1512 1 6 1 1 1491 1 6 1 2 1491 1 6 1 3 1511 2 6 1 1 1483 2 6 1 2 1524 2 6 1 3 1418 1 7 1 1 1480 1 7 1 2 1455 1 7 1 3 1479 2 7 1 1 1557 2 7 1 2 1656 2 7 1 3 1496 1 8 1 1 1560 1 8 1 2 1501 1 8 1 3 1514 2 8 1 1 1615 2 8 1 2 1540 2 8 1 3 1588 1 9 1 1 1581 1 9 1 2 1637 1 9 1 3 1474 2 9 1 1 1438 2 9 1 2 1454 2 9 1 3 1574 1 10 1 1 1513 1 10 1 2 1570 1 10 1 3 1594 2 10 1 1 1713 2 10 1 2 1669 2 10 1 3

1557 1 11 1 1 1508 1 11 1 2 1633 1 11 1 3 1571 2 11 1 1 1611 2 11 1 2 1631 2 11 1 3 1500 1 12 1 1 1476 1 12 1 2 1515 1 12 1 3 1626 2 12 1 1 1586 2 12 1 2 1608 2 12 1 3 1672 1 13 1 1 1634 1 13 1 2 1804 1 13 1 3 1490 2 13 1 1 1516 2 13 1 2 1497 2 13 1 3 1551 1 14 1 1 1568 1 14 1 2 1538 1 14 1 3 1617 2 14 1 1 1614 2 14 1 2 1639 2 14 1 3 1456 1 15 1 1 1448 1 15 1 2 1523 1 15 1 3 1525 2 15 1 1 1402 2 15 1 2 1436 2 15 1 3 1573 1 16 1 1 1520 1 16 1 2 1617 1 16 1 3 1634 2 16 1 1 1627 2 16 1 2 1575 2 16 1 3 1516 1 17 1 1 1516 1 17 1 2 1628 1 17 1 3 1613 2 17 1 1 1567 2 17 1 2 1567 2 17 1 3 1612 1 18 1 1 1535 1 18 1 2 1672 1 18 1 3 1529 2 18 1 1 1495 2 18 1 2 1528 2 18 1 3 1598 1 19 1 1 1514 1 19 1 2 1596 1 19 1 3 1571 2 19 1 1 1535 2 19 1 2 1511 2 19 1 3 1575 1 20 1 1 1568 1 20 1 2 1586 1 20 1 3 1652 2 20 1 1 1545 2 20 1 2 1627 2 20 1 3 1658 1 21 1 1

1611 1 21 1 2 1578 1 21 1 3 1647 2 21 1 1 1607 2 21 1 2 1612 2 21 1 3 1547 1 22 1 1 1528 1 22 1 2 1532 1 22 1 3 1537 2 22 1 1 1513 2 22 1 2 1536 2 22 1 3 1790 1 1 2 1 1909 1 1 2 2 1885 1 1 2 3 1870 2 1 2 2 1860 2 1 2 3 1712 1 2 2 1 1719 1 2 2 2 1704 1 2 2 3 1775 2 2 2 1 1718 2 2 2 2 1756 2 2 2 3 1693 1 3 2 1 1755 1 3 2 2 1749 1 3 2 3 1824 2 3 2 1 1483 2 3 2 2 1896 2 3 2 3 1754 1 4 2 1 1745 1 4 2 2 1739 1 4 2 3 1670 2 4 2 1 1632 2 4 2 2 1814 2 4 2 3 1757 1 5 2 1 1687 1 5 2 2 1735 1 5 2 3 1683 2 5 2 1 1759 2 5 2 2 1703 2 5 2 3 1697 1 6 2 1 1709 1 6 2 2 1714 1 6 2 3 1693 2 6 2 1 1723 2 6 2 2 1722 2 6 2 3 1722 1 7 2 1 1625 1 7 2 2 1703 1 7 2 3 1690 2 7 2 1 1684 2 7 2 2 1688 2 7 2 3 1884 1 8 2 1 1802 1 8 2 2 1772 1 8 2 3 1864 2 8 2 1 1887 2 8 2 2 1821 2 8 2 3 1713 1 9 2 1 1734 1 9 2 2 1727 1 9 2 3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

54

1709 2 9 2 1 1756 2 9 2 2 1758 2 9 2 3 1759 1 10 2 1 1809 1 10 2 2 1738 1 10 2 3 1764 2 10 2 1 1816 2 10 2 2 1787 2 10 2 3 1749 1 11 2 1 1782 1 11 2 2 1740 1 11 2 3 1741 2 11 2 1 1792 2 11 2 2 1784 2 11 2 3 1748 1 12 2 1 1712 1 12 2 2 1718 1 12 2 3 1797 2 12 2 1 1614 2 12 2 2 1790 2 12 2 3 1632 1 13 2 1 1668 1 13 2 2 1701 1 13 2 3 1626 2 13 2 1 1566 2 13 2 2 1681 2 13 2 3 1752 1 14 2 1 1535 1 14 2 2 1673 1 14 2 3 1724 2 14 2 1 1741 2 14 2 2 1685 2 14 2 3 1871 1 15 2 1 1835 1 15 2 2 1887 1 15 2 3 1770 2 15 2 1 1752 2 15 2 2 1798 2 15 2 3 1688 1 16 2 1 1725 1 16 2 2 1743 1 16 2 3 1744 2 16 2 1 1697 2 16 2 2 1710 2 16 2 3 1789 1 17 2 1 1769 1 17 2 2 1825 1 17 2 3 1716 2 17 2 1 1762 2 17 2 2 1638 2 17 2 3 1826 1 18 2 1 1780 1 18 2 2 1849 1 18 2 3 1805 2 18 2 1 1801 2 18 2 2 1931 2 18 2 3 1751 1 19 2 1 1759 1 19 2 2 1880 1 19 2 3 1827 2 19 2 1 1766 2 19 2 2

1786 2 19 2 3 1738 1 20 2 1 1678 1 20 2 2 1768 1 20 2 3 1752 2 20 2 1 1747 2 20 2 2 1768 2 20 2 3 1742 1 21 2 1 1836 1 21 2 2 1713 1 21 2 3 1746 2 21 2 1 1744 2 21 2 2 1764 2 21 2 3 1761 1 22 2 1 1743 1 22 2 2 1660 1 22 2 3 1666 2 22 2 1 1692 2 22 2 2 1819 2 22 2 3 1493 1 1 3 1 1385 1 1 3 2 1510 1 1 3 3 1565 2 1 3 1 1562 2 1 3 2 1466 2 1 3 3 1516 1 2 3 1 1513 1 2 3 2 1577 1 2 3 3 1490 2 2 3 1 1547 2 2 3 2 1570 2 2 3 3 1451 1 3 3 1 1528 1 3 3 2 1506 1 3 3 3 1485 2 3 3 1 1505 2 3 3 2 1471 2 3 3 3 1440 1 4 3 1 1486 1 4 3 2 1520 1 4 3 3 1529 2 4 3 1 1508 2 4 3 2 1506 2 4 3 3 1523 1 5 3 1 1472 1 5 3 2 1483 1 5 3 3 1513 2 5 3 1 1520 2 5 3 2 1491 2 5 3 3 1437 1 6 3 1 1493 1 6 3 2 1522 1 6 3 3 1492 2 6 3 1 1481 2 6 3 2 1461 2 6 3 3 1485 1 7 3 1 1477 1 7 3 2 1463 1 7 3 3 1475 2 7 3 1 1547 2 7 3 2 1485 2 7 3 3 1433 1 8 3 1

1405 1 8 3 2 1478 1 8 3 3 1497 2 8 3 1 1501 2 8 3 2 1495 2 8 3 3 1463 1 9 3 1 1499 1 9 3 2 1505 1 9 3 3 1503 2 9 3 1 1491 2 9 3 2 1516 2 9 3 3 1534 1 10 3 1 1458 1 10 3 2 1454 1 10 3 3 1448 2 10 3 1 1465 2 10 3 2 1474 2 10 3 3 1533 1 11 3 1 1514 1 11 3 2 1555 1 11 3 3 1488 2 11 3 1 1470 2 11 3 2 1483 2 11 3 3 1450 1 12 3 1 1525 1 12 3 2 1497 1 12 3 3 1496 2 12 3 1 1483 2 12 3 2 1519 2 12 3 3 1537 1 13 3 1 1527 1 13 3 2 1515 1 13 3 3 1499 2 13 3 1 1479 2 13 3 2 1475 2 13 3 3 1515 1 14 3 1 1490 1 14 3 2 1494 1 14 3 3 1469 2 14 3 1 1480 2 14 3 2 1466 2 14 3 3 1511 1 15 3 1 1568 1 15 3 2 1594 1 15 3 3 1515 2 15 3 1 1543 2 15 3 2 1565 2 15 3 3 1449 1 16 3 1 1452 1 16 3 2 1471 1 16 3 3 1543 2 16 3 1 1448 2 16 3 2 1491 2 16 3 3 1388 1 17 3 1 1421 1 17 3 2 1394 1 17 3 3 1452 2 17 3 1 1435 2 17 3 3 1475 1 18 3 1 1455 1 18 3 2 1445 1 18 3 3 1458 2 18 3 1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

55

1531 2 18 3 2 1459 2 18 3 3 1468 1 19 3 1 1525 1 19 3 2 1446 1 19 3 3 1435 2 19 3 1 1478 2 19 3 2 1421 2 19 3 3 1549 1 20 3 1

1531 1 20 3 2 1499 1 20 3 3 1499 2 20 3 1 1465 2 20 3 2 1468 2 20 3 3 1530 1 21 3 1 1440 1 21 3 2 1484 1 21 3 3 1499 2 21 3 1

1454 2 21 3 2 1442 1 22 3 1 1487 1 22 3 2 1436 1 22 3 3 1589 2 22 3 1 1591 2 22 3 2 1556 2 22 3 3 END

Table C-1 Example OpenBUGS Homogeneity Summary

mean sd MC_error val25pc median val975pc start sample mu[11] 15460 0122 000123 1522 1546 1570 5001 10000 mu[12] 15640 0137 000141 1537 1564 1591 5001 10000 mu[21] 17490 0113 000118 1727 1749 1772 5001 10000 mu[22] 17480 0129 000125 1722 1748 1774 5001 10000 mu[31] 14870 0073 000073 1473 1487 1501 5001 10000 mu[32] 14960 0069 000063 1482 1496 1509 5001 10000 muA[1] 15550 0092 000092 1536 1555 1573 5001 10000 muA[2] 17490 0085 000086 1732 1749 1765 5001 10000 muA[3] 14910 0050 000046 1481 1491 1501 5001 10000 sigbP[1] 3269 04985 000514 2376 3244 4324 5001 10000 sigbP[2] 2548 04385 000461 1733 2533 3472 5001 10000 sigbP[3] 1850 0292 000260 1329 1832 2477 5001 10000

sigtP[11] 2774 0298 000260 2268 2750 3442 5001 10000 sigtP[12] 4036 0429 000438 3290 4005 4986 5001 10000 sigtP[21] 2664 0301 000321 2155 2640 3313 5001 10000 sigtP[22] 3807 0395 000377 3116 3772 4662 5001 10000 sigtP[31] 2313 0250 000209 1880 2291 2873 5001 10000 sigtP[32] 1812 0195 000170 1473 1797 2243 5001 10000

24 Results The material appears to be homogeneous as can be seen in Figure C-1 Based on these results the between-tube variability is of the same order as within-tube variability

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

56

Component A Assay NEIF Component A Assay ND6 box plot gamma[11]

[111][112][113]

[114]

[115][116]

[117]

[118]

[119]

[1110][1111]

[1112]

[1113]

[1114]

[1115]

[1116][1117]

[1118]

[1119][1120]

[1121]

[1122]

114

box plot gamma[12]

[121][122]

[123]

[124][125]

[126]

[127][128]

[129]

[1210]

[1211][1212]

[1213]

[1214]

[1215]

[1216]

[1217]

[1218][1219]

[1220][1221]

[1222]

115

Component B Assay NEIF Component B Assay ND6

box plot gamma[21]

[211]

[212]

[213][214]

[215][216]

[217]

[218]

[219]

[2110][2111]

[2112]

[2113][2114]

[2115]

[2116]

[2117]

[2118]

[2119]

[2120]

[2121]

[2122]

129

box plot gamma[22]

[221]

[222][223]

[224][225][226][227]

[228]

[229]

[2210][2211]

[2212]

[2213]

[2214]

[2215]

[2216][2217]

[2218]

[2219]

[2220][2221][2222]

128

Component C Assay NEIF Component C Assay ND6

box plot gamma[31]

[311]

[312]

[313][314]

[315][316]

[317]

[318]

[319][3110]

[3111]

[3112]

[3113]

[3114]

[3115]

[3116]

[3117]

[3118]

[3119]

[3120]

[3121]

[3122]109

box plot gamma[32]

[321][322]

[323]

[324][325]

[326]

[327][328][329]

[3210]

[3211]

[3212]

[3213][3214]

[3215]

[3216]

[3217]

[3218]

[3219]

[3220][3221]

[3222]

11

Figure C-1 Homogeneity Assessment

Note These results were obtained using the 1-4 gravimetric dilution-adjusted λ but uncorrected for droplet volume

or the 1-10 volumetric dilution That is the vertical axis in each panel should be multiplied by 07349times10 = 7349 This constant difference in scale does not impact the homogeneity analysis Error bars span approximate 95 confidence intervals

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

57

3 Certification measurements 31 Data

Originally the experiment consisted of 5 plates containing the 3 components (A B C) and 12 assays Ten of the assays were for genomic DNA identification (for the certified value) and two of the assays were for mitochondrial detection which is used for informational values (and are not included in this data set) Each of the plates contained samples from a new vial from different boxes (noted in the ldquovial usedrdquo portion the number references the box) The assay positions were alternated across the five plates as well For the plate design and which tubes were used see

Figure 6 Basic Plate Layout Design for Certification Measurements It was later determined that the homogeneity and stability data sets were obtained using a different set of pipettes than the certification data set and since all data was to be combined for certification another plate was obtained using the original pipette For the complete certification data see

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component Stability data was measured at different temperatures over several weeks using assays NEIF and ND6 Only the 4degC data was also used for certification For the complete stability data see

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

32 Model It was decided that all three sets of data should be used to derive the certified values In this way the uncertainty of the certified values would better reflect any potential heterogeneity in the samples as well as any additional unknown factors that may have effected the results Because two different pipettes were used to obtain different sets of data there was a significant bias in some of the measurements Specifically the ldquocertificationrdquo data from the first five plates was obtained using pipette set ldquo1rdquo and the remaining measurements were obtained using pipette set ldquo2rdquo This bias was introduced into the statistical model as part of the mean Possible effects of different assays and type of data (heterogeneity certification and stability) were accounted for in the statistical model in terms of a random effect Specifically the response for each component (A B C) was represented as

119910119910119894119894119894119894119894119894~1198731198731205741205741198941198941198941198941198941198941205901205901198941198942 where ~N() denotes ldquodistributed as a normal distribution of some mean and variancerdquo i = 123 denotes type of measurement heterogeneity certification and stability j = 1hellip10 denotes assay 2PR4 POTP NEIF R4Q5 D5 ND6 D9 HBB1 ND14 22C3 k = 1 2 denotes pipette set 120574120574119894119894119894119894119894119894 = 120572120572119894119894119894119894 + 120573120573119894119894 120572120572119894119894119894119894~119873119873(micro1198941198941205901205901205721205722) 120573120573119894119894 = 120573120573 for data type 1 and 120573120573119894119894 = 0 for data types 2 and 3 For a given 2372a component the variance 1205901205901205721205722 measures the between assay variability and 1205901205901198941198942 measures the repeatability variability for each data type The data was analyzed using Bayesian MCMC programmed in OpenBUGSC2 with non-informative Gaussian priors for the means and Gamma priors for the variance components

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

58

33 OpenBUGS code inits and data Certification code ModelBegin for(i in 13)mu[i]~dnorm(010E-5) sig[i]~dgamma(10E-510E-5) sigalph~dgamma(10E-510E-5) for(i in 110)alpha[1i]~dnorm(mu[1] sigalph) for(i in 12)alpha[2i]~dnorm(mu[2] sigalph) alpha[3i]~dnorm(mu[3]sigalph) b~dnorm(010E-5) beta[1]lt-b beta[2]lt-00 for(i in 1Ntot)mean[i]lt-alpha[typ[i]asy[i]]+beta[pip[i]] for(i in 1Ntot)lamb[i]~dnorm(mean[i]sig[typ[i]]) ModelEnd Certification Inits list(sig=c(111) sigalph=1) Certification Data for Component A list(Ntot=275) lamb[] asy[] cmp[] typ[] pip[] 1667 1 1 1 1 1655 2 1 1 1 1647 3 1 1 1 1594 4 1 1 1 1655 5 1 1 1 1629 6 1 1 1 1655 7 1 1 1 1644 8 1 1 1 1645 9 1 1 1 1545 10 1 1 1 1698 1 1 1 1 1685 2 1 1 1 1665 3 1 1 1 1658 4 1 1 1 1683 5 1 1 1 1647 6 1 1 1 1652 7 1 1 1 1611 8 1 1 1 1707 9 1 1 1 1603 10 1 1 1 1668 1 1 1 1 1587 2 1 1 1 1678 3 1 1 1 1628 4 1 1 1 1738 5 1 1 1 1667 6 1 1 1 1695 7 1 1 1 1757 8 1 1 1 1771 9 1 1 1 1762 10 1 1 1 1739 1 1 1 1 1702 2 1 1 1 1614 3 1 1 1 1762 4 1 1 1 1649 5 1 1 1 1678 6 1 1 1 1653 7 1 1 1

175 8 1 1 1 1764 9 1 1 1 1759 10 1 1 1 1638 1 1 1 1 1676 2 1 1 1 1616 3 1 1 1 1559 4 1 1 1 1608 5 1 1 1 1611 6 1 1 1 1517 7 1 1 1 1557 8 1 1 1 1604 9 1 1 1 1675 10 1 1 1 1667 1 1 1 1 1638 2 1 1 1 1624 3 1 1 1 1587 4 1 1 1 1652 5 1 1 1 1589 6 1 1 1 1607 7 1 1 1 1582 8 1 1 1 1625 9 1 1 1 1616 10 1 1 1 1888 1 1 1 1 1868 2 1 1 1 1953 3 1 1 1 1779 4 1 1 1 1924 5 1 1 1 1861 6 1 1 1 1824 7 1 1 1 1809 8 1 1 1 1864 9 1 1 1 193 10 1 1 1 1939 1 1 1 1 1815 2 1 1 1 1902 3 1 1 1 1871 4 1 1 1 2124 5 1 1 1

1787 6 1 1 1 1844 7 1 1 1 1799 8 1 1 1 1857 9 1 1 1 1942 10 1 1 1 1597 1 1 1 1 1533 2 1 1 1 1462 3 1 1 1 1493 4 1 1 1 1538 5 1 1 1 1492 6 1 1 1 1528 7 1 1 1 1548 8 1 1 1 1513 9 1 1 1 1562 10 1 1 1 159 1 1 1 1 1404 2 1 1 1 1533 3 1 1 1 1519 4 1 1 1 1586 5 1 1 1 1516 6 1 1 1 1538 7 1 1 1 1581 8 1 1 1 1514 9 1 1 1 1431 1 1 1 2 1407 2 1 1 2 1487 3 1 1 2 145 4 1 1 2 1394 5 1 1 2 1396 6 1 1 2 1414 7 1 1 2 1401 8 1 1 2 1412 9 1 1 2 1435 10 1 1 2 1411 1 1 1 2 1395 2 1 1 2 1398 3 1 1 2 1439 4 1 1 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

59

1422 5 1 1 2 1393 6 1 1 2 139 7 1 1 2 1382 8 1 1 2 146 9 1 1 2 141 10 1 1 2 1632 1 1 2 2 1531 1 1 2 2 1524 1 1 2 2 1454 1 1 2 2 1465 1 1 2 2 1512 1 1 2 2 1418 1 1 2 2 1496 1 1 2 2 1588 1 1 2 2 1574 1 1 2 2 1557 1 1 2 2 15 1 1 2 2 1672 1 1 2 2 1551 1 1 2 2 1573 1 1 2 2 1516 1 1 2 2 1612 1 1 2 2 1598 1 1 2 2 1575 1 1 2 2 1658 1 1 2 2 1547 1 1 2 2 1529 1 1 2 2 1525 1 1 2 2 154 1 1 2 2 1458 1 1 2 2 154 1 1 2 2 1491 1 1 2 2 148 1 1 2 2 156 1 1 2 2 1581 1 1 2 2 1513 1 1 2 2 1508 1 1 2 2 1476 1 1 2 2 1634 1 1 2 2 1568 1 1 2 2 152 1 1 2 2 1516 1 1 2 2 1535 1 1 2 2 1514 1 1 2 2 1568 1 1 2 2 1611 1 1 2 2 1528 1 1 2 2 1511 1 1 2 2 1549 1 1 2 2 1533 1 1 2 2 1427 1 1 2 2 1504 1 1 2 2 1491 1 1 2 2 1455 1 1 2 2

1501 1 1 2 2 1637 1 1 2 2 157 1 1 2 2 1633 1 1 2 2 1515 1 1 2 2 1804 1 1 2 2 1538 1 1 2 2 1617 1 1 2 2 1628 1 1 2 2 1672 1 1 2 2 1596 1 1 2 2 1586 1 1 2 2 1578 1 1 2 2 1532 1 1 2 2 1616 2 1 2 2 1615 2 1 2 2 1582 2 1 2 2 1627 2 1 2 2 1607 2 1 2 2 1511 2 1 2 2 1479 2 1 2 2 1514 2 1 2 2 1474 2 1 2 2 1594 2 1 2 2 1571 2 1 2 2 1626 2 1 2 2 149 2 1 2 2 1617 2 1 2 2 1634 2 1 2 2 1613 2 1 2 2 1529 2 1 2 2 1571 2 1 2 2 1652 2 1 2 2 1647 2 1 2 2 1537 2 1 2 2 164 2 1 2 2 1589 2 1 2 2 1152 2 1 2 2 1605 2 1 2 2 1633 2 1 2 2 1483 2 1 2 2 1557 2 1 2 2 1615 2 1 2 2 1438 2 1 2 2 1713 2 1 2 2 1611 2 1 2 2 1586 2 1 2 2 1516 2 1 2 2 1614 2 1 2 2 1627 2 1 2 2 1567 2 1 2 2 1495 2 1 2 2 1535 2 1 2 2 1545 2 1 2 2 1607 2 1 2 2

1513 2 1 2 2 1582 2 1 2 2 1601 2 1 2 2 1555 2 1 2 2 1607 2 1 2 2 1586 2 1 2 2 1524 2 1 2 2 1656 2 1 2 2 154 2 1 2 2 1454 2 1 2 2 1669 2 1 2 2 1631 2 1 2 2 1608 2 1 2 2 1497 2 1 2 2 1639 2 1 2 2 1575 2 1 2 2 1567 2 1 2 2 1528 2 1 2 2 1511 2 1 2 2 1627 2 1 2 2 1612 2 1 2 2 1536 2 1 2 2 1568 1 1 3 2 152 1 1 3 2 1486 1 1 3 2 1441 1 1 3 2 1706 1 1 3 2 144 1 1 3 2 15 1 1 3 2 1537 1 1 3 2 1387 1 1 3 2 1463 1 1 3 2 157 1 1 3 2 1429 1 1 3 2 1667 1 1 3 2 1632 1 1 3 2 1589 1 1 3 2 1474 2 1 3 2 1543 2 1 3 2 1555 2 1 3 2 1599 2 1 3 2 1494 2 1 3 2 1624 2 1 3 2 1577 2 1 3 2 1593 2 1 3 2 1479 2 1 3 2 1517 2 1 3 2 1538 2 1 3 2 159 2 1 3 2 1479 2 1 3 2 1479 2 1 3 2 1522 2 1 3 2 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

60

Table C-2 Example OpenBUGS Certification Summary for Component A

mean sd MC_error val25pc median val975pc start sample b 2584 02948 0005376 2001 2588 3158 100001 100000 mu[1] 1417 02706 0005366 1364 1416 147 100001 100000 mu[2] 1559 008368 4776E-4 1543 1559 1576 100001 100000 mu[3] 1533 01504 0001419 1504 1533 1563 100001 100000

Certification Data for Component B list(Ntot=278) lamb[] asy[] cmp[] typ[] pip[] 1869 1 2 1 1 185 2 2 1 1 1709 3 2 1 1 1856 4 2 1 1 1848 5 2 1 1 1828 6 2 1 1 1861 7 2 1 1 1825 8 2 1 1 1871 9 2 1 1 1776 10 2 1 1 196 1 2 1 1 1781 2 2 1 1 1707 3 2 1 1 1787 4 2 1 1 1829 5 2 1 1 1826 6 2 1 1 1858 7 2 1 1 1869 8 2 1 1 1773 9 2 1 1 1902 10 2 1 1 2174 1 2 1 1 2202 2 2 1 1 2172 3 2 1 1 223 5 2 1 1 2094 6 2 1 1 1949 7 2 1 1 1971 8 2 1 1 2123 9 2 1 1 2107 10 2 1 1 2231 1 2 1 1 2292 2 2 1 1 2216 3 2 1 1 2078 4 2 1 1 2192 5 2 1 1 2034 6 2 1 1 211 7 2 1 1 2123 8 2 1 1 2066 9 2 1 1 2046 10 2 1 1 181 1 2 1 1 1865 2 2 1 1 1806 3 2 1 1 1878 4 2 1 1 1824 5 2 1 1 1768 6 2 1 1 1811 7 2 1 1 1763 8 2 1 1 1865 9 2 1 1 1815 10 2 1 1

1806 1 2 1 1 1801 2 2 1 1 1786 3 2 1 1 1768 4 2 1 1 179 5 2 1 1 167 6 2 1 1 1749 7 2 1 1 1741 8 2 1 1 1781 9 2 1 1 1866 10 2 1 1 1784 1 2 1 1 1735 2 2 1 1 1778 3 2 1 1 1835 4 2 1 1 1794 5 2 1 1 17 6 2 1 1 1799 7 2 1 1 1765 8 2 1 1 1754 9 2 1 1 1792 10 2 1 1 1821 1 2 1 1 1813 2 2 1 1 1813 3 2 1 1 1749 4 2 1 1 1848 5 2 1 1 1711 6 2 1 1 1779 7 2 1 1 1769 8 2 1 1 1725 9 2 1 1 1814 10 2 1 1 1718 1 2 1 1 1753 2 2 1 1 1773 3 2 1 1 1777 4 2 1 1 1803 5 2 1 1 1718 6 2 1 1 1802 7 2 1 1 1834 8 2 1 1 1731 9 2 1 1 1842 10 2 1 1 1768 1 2 1 1 1698 2 2 1 1 1792 3 2 1 1 174 4 2 1 1 1795 5 2 1 1 1693 6 2 1 1 1743 7 2 1 1 1745 8 2 1 1 1714 9 2 1 1 1791 1 2 1 2

1746 2 2 1 2 1803 3 2 1 2 1767 4 2 1 2 1745 5 2 1 2 1824 6 2 1 2 1752 7 2 1 2 1813 8 2 1 2 1648 9 2 1 2 1818 10 2 1 2 177 1 2 1 2 182 2 2 1 2 1803 3 2 1 2 1739 4 2 1 2 1773 5 2 1 2 1729 6 2 1 2 1744 7 2 1 2 1754 8 2 1 2 1671 9 2 1 2 1741 10 2 1 2 179 1 2 2 2 1712 1 2 2 2 1693 1 2 2 2 1754 1 2 2 2 1757 1 2 2 2 1697 1 2 2 2 1722 1 2 2 2 1884 1 2 2 2 1713 1 2 2 2 1759 1 2 2 2 1749 1 2 2 2 1748 1 2 2 2 1632 1 2 2 2 1752 1 2 2 2 1871 1 2 2 2 1688 1 2 2 2 1789 1 2 2 2 1826 1 2 2 2 1751 1 2 2 2 1738 1 2 2 2 1742 1 2 2 2 1761 1 2 2 2 1909 1 2 2 2 1719 1 2 2 2 1755 1 2 2 2 1745 1 2 2 2 1687 1 2 2 2 1709 1 2 2 2 1625 1 2 2 2 1802 1 2 2 2 1734 1 2 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

61

1809 1 2 2 2 1782 1 2 2 2 1712 1 2 2 2 1668 1 2 2 2 1535 1 2 2 2 1835 1 2 2 2 1725 1 2 2 2 1769 1 2 2 2 178 1 2 2 2 1759 1 2 2 2 1678 1 2 2 2 1836 1 2 2 2 1743 1 2 2 2 1885 1 2 2 2 1704 1 2 2 2 1749 1 2 2 2 1739 1 2 2 2 1735 1 2 2 2 1714 1 2 2 2 1703 1 2 2 2 1772 1 2 2 2 1727 1 2 2 2 1738 1 2 2 2 174 1 2 2 2 1718 1 2 2 2 1701 1 2 2 2 1673 1 2 2 2 1887 1 2 2 2 1743 1 2 2 2 1825 1 2 2 2 1849 1 2 2 2 188 1 2 2 2 1768 1 2 2 2 1713 1 2 2 2 166 1 2 2 2 1775 2 2 2 2 1824 2 2 2 2 167 2 2 2 2 1683 2 2 2 2 1693 2 2 2 2 169 2 2 2 2 1864 2 2 2 2 1709 2 2 2 2 1764 2 2 2 2

1741 2 2 2 2 1797 2 2 2 2 1626 2 2 2 2 1724 2 2 2 2 177 2 2 2 2 1744 2 2 2 2 1716 2 2 2 2 1805 2 2 2 2 1827 2 2 2 2 1752 2 2 2 2 1746 2 2 2 2 1666 2 2 2 2 187 2 2 2 2 1718 2 2 2 2 1483 2 2 2 2 1632 2 2 2 2 1759 2 2 2 2 1723 2 2 2 2 1684 2 2 2 2 1887 2 2 2 2 1756 2 2 2 2 1816 2 2 2 2 1792 2 2 2 2 1614 2 2 2 2 1566 2 2 2 2 1741 2 2 2 2 1752 2 2 2 2 1697 2 2 2 2 1762 2 2 2 2 1801 2 2 2 2 1766 2 2 2 2 1747 2 2 2 2 1744 2 2 2 2 1692 2 2 2 2 186 2 2 2 2 1756 2 2 2 2 1896 2 2 2 2 1814 2 2 2 2 1703 2 2 2 2 1722 2 2 2 2 1688 2 2 2 2 1821 2 2 2 2 1758 2 2 2 2 1787 2 2 2 2

1784 2 2 2 2 179 2 2 2 2 1681 2 2 2 2 1685 2 2 2 2 1798 2 2 2 2 171 2 2 2 2 1638 2 2 2 2 1931 2 2 2 2 1786 2 2 2 2 1768 2 2 2 2 1764 2 2 2 2 1819 2 2 2 2 1666 1 2 3 2 1626 1 2 3 2 188 1 2 3 2 1653 1 2 3 2 1733 1 2 3 2 1551 1 2 3 2 1884 1 2 3 2 1718 1 2 3 2 1549 1 2 3 2 1723 1 2 3 2 194 1 2 3 2 164 1 2 3 2 1565 1 2 3 2 1787 1 2 3 2 1655 2 2 3 2 167 2 2 3 2 1673 2 2 3 2 1705 2 2 3 2 1755 2 2 3 2 1625 2 2 3 2 1708 2 2 3 2 1705 2 2 3 2 1862 2 2 3 2 1686 2 2 3 2 1723 2 2 3 2 1909 2 2 3 2 1859 2 2 3 2 1897 2 2 3 2 1972 2 2 3 2 END

Table C-3 Example OpenBUGS Certification Summary for Component B

mean sd MC_error val25pc median val975pc start sample b 09337 03387 0007504 02695 09371 1592 100001 100000 mu[1] 1764 03101 0007483 1704 1763 1825 100001 100000 mu[2] 1748 008014 4719E-4 1733 1748 1764 100001 100000 mu[3] 1735 02338 0003093 1688 1736 1781 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

62

Certification data for Component C list(Ntot=280) lamb[] asy[] cmp[] typ[] pip[] 163 1 3 1 1 1596 2 3 1 1 1574 3 3 1 1 159 4 3 1 1 161 5 3 1 1 1589 6 3 1 1 1569 7 3 1 1 1565 8 3 1 1 1564 9 3 1 1 1548 10 3 1 1 1706 1 3 1 1 1576 2 3 1 1 1538 3 3 1 1 1555 4 3 1 1 1607 5 3 1 1 1548 6 3 1 1 1572 7 3 1 1 1681 8 3 1 1 1492 9 3 1 1 1655 10 3 1 1 1499 1 3 1 1 1576 2 3 1 1 1545 3 3 1 1 1584 4 3 1 1 1577 5 3 1 1 1599 6 3 1 1 1587 7 3 1 1 152 8 3 1 1 1479 9 3 1 1 1512 10 3 1 1 1497 1 3 1 1 1611 2 3 1 1 156 3 3 1 1 1546 4 3 1 1 1529 5 3 1 1 1478 6 3 1 1 1504 7 3 1 1 1551 8 3 1 1 1438 9 3 1 1 1494 10 3 1 1 1452 1 3 1 1 1347 2 3 1 1 1395 3 3 1 1 1442 4 3 1 1 1448 5 3 1 1 1463 6 3 1 1 1402 7 3 1 1 1474 8 3 1 1 1466 9 3 1 1 1439 10 3 1 1 1424 1 3 1 1 1397 2 3 1 1 135 3 3 1 1 1459 4 3 1 1 1602 5 3 1 1 1435 6 3 1 1 1433 7 3 1 1 1481 8 3 1 1 1482 9 3 1 1

1568 10 3 1 1 147 1 3 1 1 1411 2 3 1 1 1512 3 3 1 1 1452 4 3 1 1 1567 5 3 1 1 148 6 3 1 1 1442 7 3 1 1 143 8 3 1 1 1413 9 3 1 1 1486 10 3 1 1 1444 1 3 1 1 1427 2 3 1 1 1439 3 3 1 1 1468 4 3 1 1 1524 5 3 1 1 1424 6 3 1 1 1459 7 3 1 1 1467 8 3 1 1 1411 9 3 1 1 148 10 3 1 1 1485 1 3 1 1 1422 2 3 1 1 1438 3 3 1 1 1439 4 3 1 1 1515 5 3 1 1 1409 6 3 1 1 1469 7 3 1 1 1453 8 3 1 1 1465 9 3 1 1 1363 10 3 1 1 1458 1 3 1 1 1459 2 3 1 1 1465 3 3 1 1 1468 4 3 1 1 1457 5 3 1 1 1412 6 3 1 1 1474 7 3 1 1 1428 8 3 1 1 1425 9 3 1 1 1393 10 3 1 1 1357 1 3 1 2 1447 2 3 1 2 1448 3 3 1 2 1409 4 3 1 2 1389 5 3 1 2 1415 6 3 1 2 1431 7 3 1 2 1381 8 3 1 2 1403 9 3 1 2 1435 10 3 1 2 1364 1 3 1 2 1361 2 3 1 2 1404 3 3 1 2 1411 4 3 1 2 1429 5 3 1 2 1417 6 3 1 2 138 7 3 1 2 1336 8 3 1 2 1387 9 3 1 2

1372 10 3 1 2 1493 1 3 2 2 1516 1 3 2 2 1451 1 3 2 2 144 1 3 2 2 1523 1 3 2 2 1437 1 3 2 2 1485 1 3 2 2 1433 1 3 2 2 1463 1 3 2 2 1534 1 3 2 2 1533 1 3 2 2 145 1 3 2 2 1537 1 3 2 2 1515 1 3 2 2 1511 1 3 2 2 1449 1 3 2 2 1388 1 3 2 2 1475 1 3 2 2 1468 1 3 2 2 1549 1 3 2 2 153 1 3 2 2 1442 1 3 2 2 1385 1 3 2 2 1513 1 3 2 2 1528 1 3 2 2 1486 1 3 2 2 1472 1 3 2 2 1493 1 3 2 2 1477 1 3 2 2 1405 1 3 2 2 1499 1 3 2 2 1458 1 3 2 2 1514 1 3 2 2 1525 1 3 2 2 1527 1 3 2 2 149 1 3 2 2 1568 1 3 2 2 1452 1 3 2 2 1421 1 3 2 2 1455 1 3 2 2 1525 1 3 2 2 1531 1 3 2 2 144 1 3 2 2 1487 1 3 2 2 151 1 3 2 2 1577 1 3 2 2 1506 1 3 2 2 152 1 3 2 2 1483 1 3 2 2 1522 1 3 2 2 1463 1 3 2 2 1478 1 3 2 2 1505 1 3 2 2 1454 1 3 2 2 1555 1 3 2 2 1497 1 3 2 2 1515 1 3 2 2 1494 1 3 2 2 1594 1 3 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

63

1471 1 3 2 2 1394 1 3 2 2 1445 1 3 2 2 1446 1 3 2 2 1499 1 3 2 2 1484 1 3 2 2 1436 1 3 2 2 1565 2 3 2 2 149 2 3 2 2 1485 2 3 2 2 1529 2 3 2 2 1513 2 3 2 2 1492 2 3 2 2 1475 2 3 2 2 1497 2 3 2 2 1503 2 3 2 2 1448 2 3 2 2 1488 2 3 2 2 1496 2 3 2 2 1499 2 3 2 2 1469 2 3 2 2 1515 2 3 2 2 1543 2 3 2 2 1452 2 3 2 2 1458 2 3 2 2 1435 2 3 2 2 1499 2 3 2 2 1499 2 3 2 2 1589 2 3 2 2 1562 2 3 2 2 1547 2 3 2 2 1505 2 3 2 2 1508 2 3 2 2 152 2 3 2 2

1481 2 3 2 2 1547 2 3 2 2 1501 2 3 2 2 1491 2 3 2 2 1465 2 3 2 2 147 2 3 2 2 1483 2 3 2 2 1479 2 3 2 2 148 2 3 2 2 1543 2 3 2 2 1448 2 3 2 2 1531 2 3 2 2 1478 2 3 2 2 1465 2 3 2 2 1454 2 3 2 2 1591 2 3 2 2 1466 2 3 2 2 157 2 3 2 2 1471 2 3 2 2 1506 2 3 2 2 1491 2 3 2 2 1461 2 3 2 2 1485 2 3 2 2 1495 2 3 2 2 1516 2 3 2 2 1474 2 3 2 2 1483 2 3 2 2 1519 2 3 2 2 1475 2 3 2 2 1466 2 3 2 2 1565 2 3 2 2 1491 2 3 2 2 1435 2 3 2 2 1459 2 3 2 2

1421 2 3 2 2 1468 2 3 2 2 1556 2 3 2 2 1424 1 3 3 2 1427 1 3 3 2 1381 1 3 3 2 1394 1 3 3 2 1473 1 3 3 2 133 1 3 3 2 1446 1 3 3 2 1424 1 3 3 2 133 1 3 3 2 1452 1 3 3 2 1512 1 3 3 2 1314 1 3 3 2 1525 1 3 3 2 1475 1 3 3 2 1542 1 3 3 2 1461 2 3 3 2 1443 2 3 3 2 1547 2 3 3 2 1394 2 3 3 2 1364 2 3 3 2 1395 2 3 3 2 1522 2 3 3 2 1483 2 3 3 2 1452 2 3 3 2 1482 2 3 3 2 1439 2 3 3 2 1458 2 3 3 2 155 2 3 3 2 1502 2 3 3 2 152 2 3 3 2 END

Table C-4 Example OpenBUGS Certification Summary for Component C

mean sd MC_error val25pc median val975pc start sample b 09626 01701 0002428 06285 09633 1293 100001 100000 mu[1] 1399 01566 0002413 1369 1399 143 100001 100000 mu[2] 1491 005701 2386E-4 148 1491 1503 100001 100000 mu[3] 1449 0132 0001181 1423 1449 1475 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

64

34 Results Table C-5 lists the results in CopiesnL for the three components

Table C-5 Parameter Values for Components A B and C

Posterior Distribution Summary Statistics Component Parameter Mean SD 25 50 975

A

β 260 029 203 260 318 μ1 1415 027 1362 1415 1467 μ2 1559 010 1539 1559 158 μ3 1533 016 1501 1533 1565

B

β 095 034 029 095 162 μ1 1762 031 1701 1762 1822 μ2 1748 010 1728 1748 1769 μ3 1735 024 1687 1735 1783

C

β 097 017 064 097 130 μ1 1399 016 1368 1399 1429 μ2 1491 008 1476 1491 1507 μ3 1449 014 1421 1449 1477

Since in each case the three data set types represent independent measurements the three estimates can be combined to obtain a consensus value In this way a between data set type variability is included in the final standard uncertainty Table C-6 lists the results in Copies per Nanoliter obtained using the NIST Consensus Builder (NICOB)C3

Table C-6 Certified Values for Components A B and C

Component Mean u(Mean) Uk=2(Mean) A 151 07 14 B 175 03 06 C 145 05 10

4 Mito-to-nuclear DNA ratios (mtDNAnDNA)

41 Data Measurements were also obtained on the same samples using three mitochondrial detection assays mtND1 mtBatz and mtKav and a single nDNA assay (genomic) to compute the ratios mtDNAnDNA For the complete mtDNAnDNA data see

Table 16 mtDNAnDNA Ratios

42 Model For each component the data was analyzed using the following random effects model to account for between assay variability

119903119903119894119894~119873119873120572120572119894119894 1205901205901198861198861198861198861198861198861198861198861198861198862

where 119894119894 denotes assay (for A and B 1 = mtND1 2 = mtBatz 3 = mtKav for C 1 = mtBatz 2 = mtKav) and 120572120572119894119894~119873119873micro120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119886119886119886119886119886119886119886119886119886119886

2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

65

43 OpenBUGS code inits and data Two versions of the same model are required one for the results from the three mtDNA assays available for components A and B mtND1 mtBatz and mtKav and the other for the results from the two mtDNA assays available for component C mtBatz and mtKav mtDNAnDNA code for components A and B ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 13)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for components A and B list(sigt=c(111)) Data for Component A list(Ndata=102) set[] rep[] asy[] rat[] 1 1 1 1931 1 2 1 187 1 1 2 1845 1 2 2 1764 2 1 1 181 2 2 1 1906 2 1 2 1785 2 2 2 1837 3 1 1 1743 3 2 1 1761 3 1 2 1829 3 2 2 1847 4 1 1 1716 4 2 1 180 4 1 2 174 4 2 2 1813 5 1 1 1923 5 2 1 1835 5 1 2 1935 5 2 2 1851 6 1 1 1757 6 2 1 1724 6 3 1 1768 6 1 2 1743 6 2 2 1671 6 3 2 1766 6 1 3 1692 6 2 3 1662 6 3 3 1767 7 1 1 1924 7 2 1 1944 7 3 1 1893 7 1 2 1786

7 2 2 1752 7 3 2 1743 7 1 3 153 7 2 3 1542 7 3 3 1523 7 1 2 158 7 2 2 1557 7 3 2 1592 7 1 3 2103 7 2 3 210 7 3 3 2049 7 1 1 1749 7 2 1 1769 7 3 1 1777 7 1 2 1965 7 2 2 1889 7 3 2 1947 7 1 3 1813 7 2 3 1864 7 3 3 1864 8 1 1 178 8 2 1 1776 8 3 1 1706 8 1 2 1804 8 2 2 1751 8 3 2 1738 8 1 3 1769 8 2 3 1786 8 3 3 1761 9 1 2 1584 9 2 2 1544 10 1 1 1947 10 2 1 1928 10 3 1 1922 10 1 1 1829

10 2 1 1895 10 3 1 1851 10 1 1 1674 10 2 1 1617 11 1 2 1667 11 2 2 1813 11 3 2 1708 11 4 2 1598 11 1 2 1566 11 2 2 1687 11 3 2 1663 11 4 2 1589 11 1 2 1643 11 2 2 1668 11 3 2 1604 11 4 2 1573 12 1 2 1719 12 2 2 1716 12 3 2 167 12 4 2 1666 12 1 2 1581 12 2 2 1573 12 3 2 1601 12 4 2 1612 12 1 2 1546 12 2 2 1542 12 3 2 1538 12 4 2 1587 13 1 1 1617 13 2 1 1599 13 1 1 1622 13 2 1 1579 13 1 1 1575 13 2 1 1528 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

66

Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A

mean sd MC_error val25pc median val975pc start sample mu 1730 1443 004419 1703 1730 1759 100001 100000

data for Component B list(Ndata=87) set[] rep[] asy[] rat[] 1 1 1 2461 1 2 1 2401 1 1 2 2555 1 2 2 2608 2 1 1 2229 2 2 1 2200 2 1 2 2108 2 2 2 2192 3 1 1 2127 3 2 1 2198 3 1 2 2455 3 2 2 2506 4 1 1 2211 4 2 1 2199 4 1 2 2380 4 2 2 2249 5 1 1 2105 5 2 1 1943 5 1 2 2136 5 2 2 2119 6 1 1 1990 6 2 1 1921 6 3 1 1982 6 1 2 1990 6 2 2 1924 6 3 2 1958 6 1 3 1903 6 2 3 1831

6 3 3 1919 7 1 1 2304 7 2 1 2341 7 3 1 2332 7 1 2 2048 7 2 2 2078 7 3 2 2067 7 1 3 2399 7 2 3 2479 7 3 3 2493 7 1 1 2065 7 2 1 2166 7 3 1 2164 7 1 2 2314 7 2 2 2280 7 3 2 2258 7 1 3 1867 7 2 3 1987 7 3 3 1971 8 1 1 2046 8 2 1 2057 8 3 1 2024 8 1 2 1998 8 2 2 2043 8 3 2 1952 8 1 3 2077 8 2 3 2117 8 3 3 2041 9 1 2 1890 9 2 2 1886

10 1 2 1928 10 2 2 1916 10 3 2 1915 10 1 2 1947 10 2 2 1946 10 3 2 1961 10 4 2 1909 10 1 2 1953 10 2 2 1952 10 3 2 1941 10 4 2 1881 11 1 2 1865 11 2 2 1863 11 3 2 1828 11 4 2 1828 11 1 2 1944 11 2 2 1964 11 3 2 1947 11 4 2 1895 11 1 2 1908 11 2 2 1909 11 3 2 1940 11 4 2 1839 12 1 1 1688 12 2 1 1783 12 1 1 1823 12 2 1 1631 12 1 1 1920 12 2 1 1887 END

Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B

mean sd MC_error val25pc median val975pc start sample mu 2045 2246 0111 1999 2045 2088 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

67

mtDNAnDNA code for component C ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 12)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for component C list(sigt=c(11)) mtDNAnDNA data for component C list(Ndata=66) set[] rep[] asy[] rat[] 1 1 1 3201 1 2 1 3299 2 1 1 3347 2 2 1 3287 3 1 1 3217 3 2 1 3384 4 1 1 3200 4 2 1 3363 5 1 1 3294 5 2 1 3013 6 1 1 2805 6 2 1 2825 6 3 1 2829 6 1 2 2709 6 2 2 2738 6 3 2 2727 7 1 1 2797 7 2 1 2642 7 3 1 2792 7 1 2 2804 7 2 2 2789

7 3 2 2842 7 1 1 2715 7 2 1 2606 7 3 1 2687 7 1 2 2799 7 2 2 2628 7 3 2 2559 7 1 1 2639 7 2 1 2535 7 3 1 2657 7 1 2 2953 7 2 2 2757 7 3 2 2802 8 1 1 2927 8 2 1 2776 8 3 1 2762 8 1 2 3044 8 2 2 2830 8 3 2 2825 9 1 1 2501 9 2 1 2522 10 1 1 2832 10 2 1 2840

10 3 1 2788 10 4 1 2725 10 1 1 2835 10 2 1 2790 10 3 1 2736 10 4 1 2679 10 1 1 2803 10 2 1 2773 10 3 1 2757 10 4 1 2752 11 1 1 2599 11 2 1 2508 11 3 1 2551 11 4 1 2609 11 1 1 2617 11 2 1 2483 11 3 1 2626 11 4 1 2602 11 1 1 2665 11 2 1 2535 11 3 1 2570 11 4 1 2650 END

Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C

mean sd MC_error val25pc median val975pc start sample mu 2774 2367 01176 2729 2774 2821 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

68

Table C-10 Results for mtDNAnDNA Posterior Distribution Summary Statistics

Component Mean Std Dev 25 50 975 A 1730 14 1703 1730 1758 B 2045 23 1994 2047 2087 C 2774 23 2729 2774 2821

5 ReferencesC1 Lunn DJ Spiegelhalter D Thomas A Best N Statistics in Medicine 2009283049ndash3082C2 Polson NG Scott JG Bayesian Analysis 20127887ndash902C3 NIST Consensus Builder Software and userrsquos guide freely available at

httpsconsensusnistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

69

Appendix D Performance in Commercial qPCR Chemistries The performance of the SRM 2372a materials in commercial qPCR chemistries was assessed using manufacturerrsquos recommended protocols Each panel in Figures D-1 to D-4 displays the average cycle threshold (Ct) as a function of DNA concentration for the 5-point dilutions of SRM 2372a components In all cases the external calibrant was SRM 2372 component A Error bars represent approximate 95 confidence intervals Table D-1 lists the estimated [DNA] for the three SRM 2372a components for all commercial assays evaluated This Table also lists the Component BComponent A (BA) and Component CComponent A (CA) ratios calculated from the [DNA] estimates

Figure D-1 Results for the Innogenomics Commercial qPCR assays

20

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantLong(IQ_h)

ABCStd

[A] = (384 plusmn 26) ngμL[B] = (441 plusmn 24) ngμL[C] = (516 plusmn 24) ngμL

Rsup2 = 0999 se(Ct) = 01016

17

18

19

20

21

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantShort(IQ_d)

ABCStd

[A] = (435 plusmn 28) ngμL[B] = (543 plusmn 12) ngμL[C] = (510 plusmn 28) ngμL

Rsup2 = 0999 se(Ct) = 007

24

25

26

27

28

29

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYLong(IQHY_h)

ABCStd

[A] = (387 plusmn 24) ngμL[B] = (448 plusmn 21) ngμL[C] = (559 plusmn 31) ngμL

Rsup2 = 0998 se(Ct) = 01314

16

18

20

22

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYShort(IQHY_d)

ABCStd

[A] = (353 plusmn 23) ngμL[B] = (267 plusmn 15) ngμL[C] = (390 plusmn 18) ngμL

Rsup2 = 1000 se(Ct) = 00420

22

24

26

28

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYY(IQHY_y)

ABCStd

[A] = (572 plusmn 13) ngμL[B] = 0 ngμL[C] = (137 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

70

Figure D-2 Results for the Thermo Fisher Commercial qPCR assays

21

22

23

24

25

26

27

28

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HumanHuman(QFH_h)

ABCStd

[A] = (464 plusmn 13) ngμL[B] = (529 plusmn 19) ngμL[C] = (472 plusmn 22) ngμL

Rsup2 = 1000 se(Ct) = 004

22

23

24

25

26

27

28

29

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoHuman(QFD_h)

ABCStd

[A] = (701 plusmn 53) ngμL[B] = (561 plusmn 11) ngμL[C] = (472 plusmn 11) ngμL

Rsup2 = 0994 se(Ct) = 01922

24

26

28

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoMale(QFD_y)

ABCStd

[A] = (706 plusmn 49) ngμL[B] = 0 ngμL[C] = (122 plusmn 02) ngμL

Rsup2 = 0999 se(Ct) = 007

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPLarge(QFHP_h)

ABCStd

[A] = (340 plusmn 18) ngμL[B] = (432 plusmn 05) ngμL[C] = (342 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 00720

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPSmall(QFHP_d)

ABCStd

[A] = (434 plusmn 15) ngμL[B] = (514 plusmn 10) ngμL[C] = (466 plusmn 07) ngμL

Rsup2 = 1000 se(Ct) = 005

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioLarge(QFT_h)

ABCStd

[A] = (362 plusmn 23) ngμL[B] = (430 plusmn 06) ngμL[C] = (378 plusmn 07) ngμL

Rsup2 = 0999 se(Ct) = 00920

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioSmall(QFT_d)

ABCStd

[A] = (468 plusmn 08) ngμL[B] = (505 plusmn 08) ngμL[C] = (523 plusmn 14) ngμL

Rsup2 = 1000 se(Ct) = 00318

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioY(QFT_y)

ABCStd

[A] = (372 plusmn 11) ngμL[B] = 0 ngμL[C] = (110 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 003

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

71

Figure D-3 Results for the Qiagen Commercial qPCR assays

Figure D-4 Results for the Promega Commercial qPCR assays

19

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresAutosomal(QPH_h)

ABCStd

[A] = (554 plusmn 21) ngμL[B] = (479 plusmn 05) ngμL[C] = (381 plusmn 09) ngμL

Rsup2 = 1000 se(Ct) = 00418

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresY(QPH_y)

ABCStd

[A] = (358 plusmn 10) ngμL[B] = 0 ngμL[C] = (118 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 007

16

17

18

19

20

21

22

23

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProHuman(QPP_h)

ABCStd

[A] = (525 plusmn 49) ngμL[B] = (480 plusmn 42) ngμL[C] = (467 plusmn 45) ngμL

Rsup2 = 0999 se(Ct) = 01019

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProDegradation(QPP_d)

ABCStd

[A] = (567 plusmn 18) ngμL[B] = (558 plusmn 08) ngμL[C] = (484 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 00410

15

20

25

30

35

-50 -40 -30 -20 -10 00 10

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProY(QPP_y)

ABCStd

[A] = (538 plusmn 27) ngμL[B] = (00 plusmn 00) ngμL[C] = (162 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 003

16

18

20

22

24

-05 00 05 10 15 20

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorAutosomal(P_h)

ABCStd

[A] = (759 plusmn 64) ngμL[B] = (775 plusmn 24) ngμL[C] = (1056 plusmn 17) ngμL

Rsup2 = 0995 se(Ct) = 01818

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorY(P_y)

ABCStd

[A] = (596 plusmn 51) ngμL[B] = 0 ngμL[C] = (117 plusmn 01) ngμL

Rsup2 = 0994 se(Ct) = 020

18

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantAutosomal(PQ_h)

ABCStd

[A] = (436 plusmn 13) ngμL[B] = (606 plusmn 20) ngμL[C] = (504 plusmn 10) ngμL

Rsup2 = 1000 se(Ct) = 00218

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantDegradation(PQ_d)

ABCStd

[A] = (420 plusmn 18) ngμL[B] = (529 plusmn 16) ngμL[C] = (530 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 00516

18

20

22

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantY(PQ_y)

ABCStd

[A] = (410 plusmn 20) ngμL[B] = 0 ngμL[C] = (139 plusmn 05) ngμL

Rsup2 = 0999 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

72

Table D-1 Summary [DNA] Results for the Commercial qPCR Assays Estimated [DNA] ngμL Ratios

A B C BA CA Assaya xb u(x)c xb u(x)c xb u(x)c yd u(y)e yd u(y)e

IQd 435 28 543 12 510 28 125 008 117 010 IQh 384 26 441 24 516 24 115 010 134 011

IQHYd 353 23 267 15 390 18 076 006 111 009 IQHYh 387 24 448 21 559 31 116 009 144 012

Ph 759 64 775 24 1056 16 102 009 139 012 PQd 420 18 529 16 530 06 126 007 126 006 PQh 436 13 606 20 504 10 139 006 116 004

QFDh 701 53 561 11 472 11 080 006 067 005 QFHh 464 13 529 19 472 22 114 005 102 006

QFHPd 434 15 514 10 466 07 118 005 107 004 QFHPh 340 18 432 05 342 03 127 007 101 005

QFTd 468 08 505 08 523 14 108 003 112 004 QFTh 362 23 430 06 378 07 118 008 104 007 QPHh 554 21 479 05 380 09 086 003 069 003 QPPd 567 18 558 08 484 11 098 003 085 003 QPPh 525 49 480 42 467 45 092 012 089 012

IQHYy 572 13 NDf 137 06

NRg

024 001 Py 596 51 NDf 117 01 020 002

PQy 410 20 NDf 139 05 034 002 QFDy 706 49 NDf 122 02 017 001 QFTy 372 11 NDf 110 06 030 002 QPHy 358 10 NDf 118 03 033 001 QPPy 538 27 0002 0001 162 11 030 003

a Assay codes are listed in Table 8 b Mean of dilution-adjusted results from five dilutions of the SRM 2372a component c Standard uncertainty of the mean estimated from the stdanrd deviation of the five dilutions d Ratio of the estimated [DNA] of component B or C relative to that of component A e Standard uncertainty of the ratio estimated from the standard uncertainties of the individual components f Not detected Component B was prepared using tissue from only female donors it is not expected to

respond to male-specific assays g Not relevant

  • Abstract
  • Keywords
  • Technical Information Contact for this SRM
  • Table of Contents
  • Table of Tables
  • Table of Figures
  • Purpose and Description
  • Warning SRM 2372a is a Human Source Material
  • Storage and Use
  • History and Background
    • Figure 1 Sales History of SRM 2372
      • Experimental Methods
        • Polymerase Chain Reaction Assays
          • Table 1 NIST-Developed Human Nuclear DNA Assays
          • Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays
          • Table 2 NIST-Optimized Human Mitochondrial DNA Assays
          • Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays
            • Droplet Digital Polymerase Chain Reaction Measurements
              • Nuclear DNA (nDNA)
                • Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays
                  • Mitochondrial DNA (mtDNA)
                    • Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays
                    • Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays
                        • Sample Preparation
                        • Absorbance Measurements
                        • SRM Unit Production
                        • Component Homogeneity
                          • Figure 4 Plate Layout Design for Homogeneity Measurements
                            • Component Stability
                              • Figure 5 Plate Layout Design for Stability Measurements
                                • Component Certification Measurements
                                  • Table 6 Tubes used for Certification Measurements
                                  • Figure 6 Basic Plate Layout Design for Certification Measurements
                                    • Mitochondrial DNA Measurements
                                      • Table 7 Run Parameters for Mitochondrial DNA
                                        • Performance in Commercial qPCR Chemistries
                                          • Table 8 Commercial qPCR Chemistries Tested with SRM 2372a
                                          • Figure 7 Plate Layout Design for Performance Assessments
                                              • Measurement Results
                                                • Absorbance Spectrophotometry
                                                  • Table 9 Absorbance at Selected Wavelengths
                                                  • Figure 8 Absorbance Spectra of SRM 2372a Components A B and C
                                                    • Droplet Volume
                                                    • Certification Measurements
                                                      • Homogeneity Results
                                                        • Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component
                                                        • Table 11 Within- and Between-Tube Relative Standard Deviations
                                                          • Stability Results
                                                            • Figure 9 ddPCR Stability Measurements as Function of Time
                                                            • Table 12 Stability Data as λacute Copy Number per Nanoliter of Component
                                                              • Ten-Assay Certification Results
                                                                • Figure 10 Comparison of Assay Performance
                                                                • Table 13 Certification Data as λacute Copy Number per Nanoliter of Component
                                                                  • Estimation of Copy Number Concentration
                                                                    • Table 14 Measured Copy Number Concentrations
                                                                      • Conversion of Copy Number to Mass Concentration
                                                                        • Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution
                                                                            • Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)
                                                                              • Figure 11 mtDNAnDNA for Components A B and C
                                                                              • Table 16 mtDNAnDNA Ratios
                                                                              • Table 17 mtDNAnDNA Summary Results
                                                                                • Performance in Commercial qPCR Chemistries
                                                                                  • Figure 12 qPCR Autosomal Quantitation Performance Summary
                                                                                  • Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary
                                                                                      • Qualification of ddPCR as Fit-For-Purpose
                                                                                        • Figure 14 Absorbance Spectra of SRM 2372 Components A and B
                                                                                        • Digital Polymerase Chain Reaction Assays
                                                                                        • Heat Treatments (Melting of the DNA into Single Strands)
                                                                                        • Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a
                                                                                          • Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a
                                                                                            • Quantitative Evaluation of ssDNA Content
                                                                                              • ddPCR Measurement of ssDNA Proportion
                                                                                              • Proof of Principle
                                                                                                • Table 19 Comparison of Snap-Cooling to Slow-Cooling
                                                                                                  • Time at Denaturation Temperature
                                                                                                    • Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures
                                                                                                      • Temperature
                                                                                                        • Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a
                                                                                                          • DNA Concentration in the ddPCR Solution
                                                                                                            • Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution
                                                                                                              • Equivalence of Assays
                                                                                                                • Table 23 Comparison of Assays
                                                                                                                    • ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials
                                                                                                                      • Table 24 Evaluation of Limiting Ratio
                                                                                                                          • Conclusions and Recommendations
                                                                                                                            • Recommended Certification Values
                                                                                                                              • Table 25 Recommended Values and 95 Uncertainties for Certification
                                                                                                                                • Use of SRM 2372a for Value-Assigning In-House Reference Materials
                                                                                                                                  • Certificate of Analysis
                                                                                                                                  • References
                                                                                                                                  • Appendix A DNA Extraction Method
                                                                                                                                    • Procedure
                                                                                                                                    • Reagents
                                                                                                                                    • Reference
                                                                                                                                      • Appendix B Spectrophotometric Measurement Protocol
                                                                                                                                        • Spectrophotometer Calibration
                                                                                                                                          • Establish Baseline
                                                                                                                                          • Enable Verification of Wavelength Calibration
                                                                                                                                          • Enable Verification of Absorbance Calibration
                                                                                                                                            • For ldquonativerdquo dsDNA Samples
                                                                                                                                              • Verify Optical Balance of Cuvettes
                                                                                                                                                • For NaOH denatured ssDNA Samples
                                                                                                                                                  • Prepare a 04 molL NaOH Stock Solution
                                                                                                                                                  • Prepare the Sample Solutions
                                                                                                                                                  • Measure Sample Solutions
                                                                                                                                                    • After All Sample Measurements are Complete
                                                                                                                                                      • Enable Verification of Spectrophotometer Stability
                                                                                                                                                      • Clean Up
                                                                                                                                                        • Evaluation
                                                                                                                                                        • Reference
                                                                                                                                                          • Appendix C SRM 2372a Statisticianrsquos report
                                                                                                                                                            • 1 Introduction
                                                                                                                                                            • 11 OpenBUGS Analysis
                                                                                                                                                            • 12 OpenBUGS Exemplar Output
                                                                                                                                                            • 2 Homogeneity assessment
                                                                                                                                                            • 21 Data
                                                                                                                                                            • 22 Model
                                                                                                                                                            • 23 OpenBUGS code and data
                                                                                                                                                              • Table C-1 Example OpenBUGS Homogeneity Summary
                                                                                                                                                                • 24 Results
                                                                                                                                                                  • Figure C-1 Homogeneity Assessment
                                                                                                                                                                    • 3 Certification measurements
                                                                                                                                                                    • 31 Data
                                                                                                                                                                    • 32 Model
                                                                                                                                                                    • 33 OpenBUGS code inits and data
                                                                                                                                                                      • Table C-2 Example OpenBUGS Certification Summary for Component A
                                                                                                                                                                      • Table C-3 Example OpenBUGS Certification Summary for Component B
                                                                                                                                                                      • Table C-4 Example OpenBUGS Certification Summary for Component C
                                                                                                                                                                        • 34 Results
                                                                                                                                                                          • Table C-5 Parameter Values for Components A B and C
                                                                                                                                                                          • Table C-6 Certified Values for Components A B and C
                                                                                                                                                                            • 4 Mito-to-nuclear DNA ratios (mtDNAnDNA)
                                                                                                                                                                            • 41 Data
                                                                                                                                                                            • 42 Model
                                                                                                                                                                            • 43 OpenBUGS code inits and data
                                                                                                                                                                              • Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A
                                                                                                                                                                              • Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B
                                                                                                                                                                              • Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C
                                                                                                                                                                              • Table C-10 Results for mtDNAnDNA
                                                                                                                                                                                • 5 References
                                                                                                                                                                                  • Appendix D Performance in Commercial qPCR Chemistries
                                                                                                                                                                                    • Table D-1 Summary [DNA] Results for the Commercial qPCR Assays
Page 7: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

iv

Table of Tables Table 1 NIST-Developed Human Nuclear DNA Assays 5 Table 2 NIST-Optimized Human Mitochondrial DNA Assays 6 Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays 8 Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays 8 Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays 8 Table 6 Tubes used for Certification Measurements 12 Table 7 Run Parameters for Mitochondrial DNA 13 Table 8 Commercial qPCR Chemistries Tested with SRM 2372a 14 Table 9 Absorbance at Selected Wavelengths 15 Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component 17 Table 11 Within- and Between-Tube Relative Standard Deviations 18 Table 12 Stability Data as λacute Copy Number per Nanoliter of Component 19 Table 13 Certification Data as λacute Copy Number per Nanoliter of Component 21 Table 14 Measured Copy Number Concentrations 22 Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution 23 Table 16 mtDNAnDNA Ratios 24 Table 17 mtDNAnDNA Summary Results 25 Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a 29 Table 19 Comparison of Snap-Cooling to Slow-Cooling 32 Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures 34 Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 36 Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution 37 Table 23 Comparison of Assays 39 Table 24 Evaluation of Limiting Ratio 41 Table 25 Recommended Values and 95 Uncertainties for Certification 43 Table C-1 Example OpenBUGS Homogeneity Summary 55 Table C-2 Example OpenBUGS Certification Summary for Component A 60 Table C-3 Example OpenBUGS Certification Summary for Component B 61 Table C-4 Example OpenBUGS Certification Summary for Component C 63 Table C-5 Parameter Values for Components A B and C 64 Table C-6 Certified Values for Components A B and C 64 Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A 66 Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B 66 Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C 67 Table C-10 Results for mtDNAnDNA 68 Table D-1 Summary [DNA] Results for the Commercial qPCR Assays 72

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

i

Table of Figures Figure 1 Sales History of SRM 2372 3 Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays 6 Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays 7 Figure 4 Plate Layout Design for Homogeneity Measurements 11 Figure 5 Plate Layout Design for Stability Measurements 11 Figure 6 Basic Plate Layout Design for Certification Measurements 12 Figure 7 Plate Layout Design for Performance Assessments 14 Figure 8 Absorbance Spectra of SRM 2372a Components A B and C 16 Figure 9 ddPCR Stability Measurements as Function of Time 18 Figure 10 Comparison of Assay Performance 20 Figure 11 mtDNAnDNA for Components A B and C 24 Figure 12 qPCR Autosomal Quantitation Performance Summary 26 Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary 26 Figure 14 Absorbance Spectra of SRM 2372 Components A and B 27 Figure 15 Comparison of Snap-Cooling to Slow-Cooling 32 Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures 33 Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 35 Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution 37 Figure 19 Comparison of Assays 38 Figure 20 Evaluation of Limiting Ratio 40 Figure 21 Degrees of Equivalence 42 Figure C-1 Homogeneity Assessment 56 Figure D-1 Results for the Innogenomics Commercial qPCR assays 69 Figure D-2 Results for the Thermo Fisher Commercial qPCR assays 70 Figure D-3 Results for the Qiagen Commercial qPCR assays 70 Figure D-4 Results for the Promega Commercial qPCR assays 71

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

1

Purpose and Description Standard Reference Material (SRM) 2372a is intended for use in the value assignment of human genomic deoxyribonucleic acid (DNA) forensic quantitation materials A unit of SRM 2372a consists of three well-characterized human genomic DNA materials solubilized in 10 mmolL 2-amino-2-(hydroxymethyl)-13 propanediol hydrochloride (Tris HCl) and 01 mmolL ethylenediaminetetraacetic acid disodium salt (disodium EDTA) pH 80 buffer (TE-4) The three component genomic DNA materials labeled A B and C are respectively derived from a single male donor a single female donor and 13 mixture of a male and a female donor A unit of the SRM consists of one 05 mL tube of each component each tube containing approximately 55 microL of DNA solution Each of these tubes is labeled and is sealed with a color-coded screw cap NIST is guided by and adheres to the ethical principles set forth in the Belmont Report1 SRM 2372a was developed after an appropriate human subject research determination by the NIST Human Subjects Protections Office

Warning SRM 2372a is a Human Source Material The three SRM 2372a components are human genomic DNA extracts prepared at NIST from buffy coat white blood cells from single-source anonymous donors All buffy coats were purchased from Interstate Blood Bank (Memphis TN) The supplier reported that each donor unit used in the preparation of this product was tested by FDA-licensed tests and found to be negative for human immunodeficiency virus (HIV-1 RNA and Anti-HIV 12) hepatitis B (HBV DNA and HBsAg) hepatitis C (HCV RNA and Anti-HCV) West Nile virus (WNV RNA) and syphilis Additionally each donor was tested according to FDA guidelines for Trypanosoma cruzi (Chagas) and found to be negativenon-reactive However no known test method can offer complete assurance that infectious agents are absent from this material Accordingly this human blood-based product should be handled at the Biosafety Level 12 SRM 2372a components and derived solutions should be disposed of in accordance with local state and federal regulations

Storage and Use Until required for use SRM 2372a should be stored in the dark between 2 degC and 8 degC The SRM 2372a component tubes should be mixed briefly and centrifuged (without opening the tube cap) prior to removing sample aliquots for analysis For the certified values to be applicable materials should be withdrawn immediately after opening the tubes and processed without delay Certified values do not apply to any material remaining in recapped tubes The certification only applies to the initial use and the same results are not guaranteed if the remaining material is used later

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

2

History and Background SRM 2372a is the second member of the SRM 2372 Human DNA Quantitation Standard series The supply of the initial material SRM 2372 was exhausted in May 2017 SRM 2372a is derived from commercially obtained buffy coat cells from different donors than those used in SRM 2372 A series of interlaboratory studies coordinated by what was then the Biotechnology Division of the National Institute of Standards and Technology (NIST) during the 1990s demonstrated that the DNA quantitation technologies then in use by the forensic human identification community adequately met the needs of the DNA typing tools then available34 However our 2004 study demonstrated that the increasingly sensitive and informative polymerase chain reaction (PCR)-based multiplex typing tools required more reliable determinations of DNA input quantity for best performance5 Quantitative PCR (qPCR) methods for DNA quantitation require calibration to an external standard of defined DNA concentration ([DNA]) expressed in units of nanograms of DNA per microliter of extract solution The true [DNA] of a commercial secondary standard may vary from its nominal concentration depending on supplier and lot number To enable providers of secondary standards to more reliably evaluate their materials and forensic testing laboratories to validate their use of commercial quantitation methods and supplies in 2007 we delivered the human DNA calibration standard SRM 2372 Human DNA Quantitation Standard6 The quantity of DNA in each of the three components of SRM 2372 was originally assigned based on the absorbance of double-stranded DNA (dsDNA) at 260 nm7 However by 2012 the absorbance properties of the materials had increased beyond the certified uncertainty limits due to slow but progressive changes in DNA tertiary structure Sales of the SRM were restricted until the DNA quantities were reassigned based on measurements of single-stranded DNA (ssDNA) in early 20138-10 Figure 1 displays the sales history of this relatively high-selling SRM While SRM 2372 was fit for the practical purpose of harmonizing DNA quantitation measurements to an accessible and reproducible reference material the conversion of both dsDNA and ssDNA absorbance to mass [DNA] is through conventional proportionality constants that are not metrologically traceable to the modern reference system for measurements the International System of Units (SI) Responding to requests from the forensic and clinical communities for calibration standards that provide traceability to the SI we and others have established droplet digital polymerase chain reaction (ddPCR) as a potentially primary measurement process that can provide traceable results11-18 The copy numbers of human nuclear DNA (nDNA) in the reaction mixture per partition symbolized as λ were determined by ddPCR using ten PCR assays optimized for use with our ddPCR platform These assays target loci on eight chromosomes with three assays on widely separated locations on one chromosome (see Table 1)

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

3

Figure 1 Sales History of SRM 2372

Each small black dot represents the sale of one unit of SRM 2372 The red lines summarize the average rate of sales as originally certified (2007 to 2012) and after re-certification (2013 to 2017) The magenta labels state the average yearly sales rate during these two periods

The nDNA copy number concentration human haploid genome equivalents (HHGE) per nanoliter component solution and symbolized here as λacute is calculated as

copy number [DNA] = λ = λ(VF) HHGEnL

where V is the average droplet volume and F is the total dilution factor the volume fraction of the component solution in the ddPCR reaction mixture The manufacturer of the ddPCR platform used a Bio-Rad QX200 Droplet Digital PCR System (Hercules CA) system does not provide metrologically traceable droplet volume information This lack of information prevented metrologically traceable conversion of the number per partition measurements to the desired number per reaction volume units In consequence NIST developed a measurement method that provides metrologically traceable droplet volumes19 We show elsewhere that the mass [DNA] nanograms of human DNA per microliter component can be traceably estimated as

mass [DNA] = 3031timesλacute ngmicroL

where the constant is the product of the number of nucleotide basepairs (bp) in the human reference genome the average bp molecular weight Avogadrorsquos number of bp per mole and

2372

2372(Recertif ied)

050

010

00

1713 unitsyear

1416 unitsyear

2007 2009 2011 2013 2015 2017

Uni

ts S

old

Calendar Year

Human DNA Quantitation StandardSRM 2372

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

4

unit scaling factors18 The three components of SRM 2372a are certified for both copy number and mass [DNA] Mitochondrial DNA (mtDNA) analysis within the forensic community is useful when the nDNA may be degraded or is in low-to-undetectable quantity The challenges of mtDNA testing include contamination of the standard due to routine laboratory use degradation of plasmids or synthetic oligomers or improper commutability of the standard relative to casework samples Commonly cell line materials are used as standards for qPCR with human genomic DNA which may lead to an incorrect mtDNA-to-nDNA ratio (mtDNAnDNA) for unknown samples Non-certified values of mtDNAnDNA for the three components are provided to bridge the current gap between well characterized mitochondrial DNA quantification assays and the availability of commercial standards With the understanding that the forensic DNA community primarily uses qPCR as a method for DNA quantification the performance of the three SRM 2372a materials was evaluated in eleven commercial qPCR chemistries currently used employed within the forensic DNA community

Experimental Methods Production and value-assignment of any certified reference material (CRM) is a complex process In addition to acquiring processing and packaging the materials included in the CRM the materials must be sufficiently homogenous across the units stable over time within its packaging and well-characterized to deliver results that are fit-for-purpose within the intended measurement community20 The process has been particularly complicated for SRM 2372a because we needed to determine the metrological properties of the required ddPCR measurement systems before we could trust the results they provide The following sections describe the methods and processes used to produce and characterize the SRM 2372a materials including sample preparation SRM unit production PCR assays ddPCR measurement processes spectrophotometric absorbance measurements homogeneity testing stability testing ddPCR nDNA and mtDNAnDNA measurements and performance in commercial qPCR assessments Many of these processes were mutually dependent therefore the sections are presented in a ldquodefinitionalrdquo order that sequentially introduces terminology Polymerase Chain Reaction Assays

Ten PCR assays were developed for characterization of SRM 2372a Each assay was confirmed to target one locus per HHGE using the National Center for Biotechnology Informationrsquos BLASTblastn system21 Primers were purchased from Eurofins Operon (Huntsville AL) TaqMan probes labeled with 6-carboxyfluorescein (FAM) were purchased from Thermo Fisher (Waltham MA) FAM-labeled Blackhole Quencher and Blackhole Quencher+ probes were purchased from LGC Biosearch Technologies (Novato CA) Table 1 details the ten nDNA assays used to certify copy number and mass [DNA] of the SRM 2372a components FAM-labeled TaqMan probes were purchased from Thermo Fisher

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

5

and used in monoplex reactions VIC (a proprietary dye)-labeled probes were purchased from the same source and were used when nDNA and mtDNA were analyzed in the same reaction mixture Figure 2 displays exemplar droplet patterns for the assays All assays were developed at NIST and have been optimized for ddPCR18 Table 2 details three mitochondrial DNA (mtDNA) assays used to estimate mtDNAnDNA in the SRM 2372a components Figure 3 displays exemplar droplet patterns for the assays These assays were optimized for ddPCR from real-time quantitative PCR (qPCR) assays from the literature All primers probes were diluted with TE-4 buffer purchased from Affymetrix-USB Products (Affymetrix Inc Cleveland OH) to a nominal 5 micromol working solution for all single-plex assays (individual nDNA and mtDNA assays) For duplex assays (combined nDNA and mtDNA) primer and probe working solutions were diluted to a nominal 10 micromol

Table 1 NIST-Developed Human Nuclear DNA Assays

Assay Target

Chromosome Band Accession Primers and Probe a

Amplicon Length bp

NEIF Gene EIF5B

Chr 2 p111-q111 NC_00000212

F gccaaacttcagccttctcttc R ctctggcaacatttcacactaca PB+ tcatgcagttgtcagaagctg

67

2PR4 Gene RPS27A

Chr 2 p16 NC_00000212

F cgggtttgggttcaggtctt R tgctacaatgaaaacattcagaagtct PB tttgtctaccacttgcaaagctggccttt

97

POTP STR TPOX

Chr 2 p253 NC_00000212

F ccaccttcctctgcttcacttt R acatgggtttttgcctttgg PT caccaactgaaatatg

60

NR4Q Gene DCK

Chr 4 q133-q211 NC_00000412

F tggtgggaatgttcttcagatga R tcgactgagacaggcatatgtt PB+ tgtatgagaaacctgaacgatggt

83

D5 STR D5S2500

Chr 5 q112 NC_00000510

F ttcatacaggcaagcaatgcat R cttaaagggtaaatgtttgcagtaatagat PT ataatatcagggtaaacaggg

75

ND6 STR D6S474

Chr 6 q21-22 NC_00000612

F gcatggctgagtctaaagttcaaag R gcagcctcagggttctcaa PB+ cccagaaccaaggaagatggt

82

D9 STR D9S2157

Chr 9 q342 NC_00000912

F ggctttgctgggtactgctt R ggaccacagcacatcagtcact PT cagggcacatgaat

60

HBB1 Gene HBB

Chr 11 p155 NC_00001110

F gctgagggtttgaagtccaactc R ggtctaagtgatgacagccgtacct PT agccagtgccagaagagccaagga

76

ND14 STR D14S1434

Chr 14 q3213C NC_0000149

F tccaccactgggttctatagttc R ggctgggaagtcccacaatc PB+ tcagactgaatcacaccatcag

109

22C3 Gene PMM1

Chr 22 q132 NC_00002210

F cccctaagaggtctgttgtgttg R aggtctggtggcttctccaat PB caaatcacctgaggtcaaggccagaaca

78

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

6

a F Forward primer R Reverse primer PB Blackhole quencher probe (FAM labeled) PB+ Blackhole Plus quencher probe (FAM labeled) PT TaqMan MGB probe (FAM labeled)

Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays

Table 2 NIST-Optimized Human Mitochondrial DNA Assays

Assay Location Primers and Probe a Amplicon Length bp Ref

mtND1 34853504 F ccctaaaacccgccacatct

69 22 35333553 R gagcgatggtgagagctaaggt 35063522 PT ccatcaccctctacatc

mtBatz 84468473 F aatattaaacacaaactaccacctacct

79 23 85038524 R tggttctcagggtttgttataa 84758491 PT cctcaccaaagcccata

mtKav 1318813308 F ggcatcaaccaaccacaccta

105 24 1336913392 R attgttaaggttgtggatgatgga 1331013325 PT cattcctgcacatctg

a F Forward primer R Reverse primer PT TaqMan MGB probe (FAM labeled)

2PR4

POTP

NEI

F

R4Q

5

D5

ND6

D9

HBB1

ND1

4

22C3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

7

Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays

Droplet Digital Polymerase Chain Reaction Measurements

Nuclear DNA (nDNA) All nDNA ddPCR measurements used in this study followed the same measurement protocol

bull Droplets were generated on the Auto Droplet Generator using the Bio-Rad ldquoddPCR ldquoSupermix for Probes (No dUTP)rdquo (Bio-Rad cat 186-3024 Control 64079080) and ldquoDroplet Generation Oil for Probesrdquo (Bio-Rad cat 186-4110 Control 64046051) Note desoxyuridine 5-triphosphatase (dUTP) is an enzyme sometimes included in PCR reaction mixtures to minimize PCR carryover contamination

bull Table 3 details the composition of the ddPCR mastermix setup for a given assay and sample where ldquomastermix refers to reaction mixture of manufacturerrsquos proprietary PCR reagent ldquoSupermixrdquo forward and reverse PCR primers for a given assay probe for that assay the sample DNA and sometimes additional water to produce the desired value dilution factor F This setup was used for all assays and samples

bull Non-Template Controls (NTCs) for each assay were included in every analysis bull Droplets were thermal cycled on a ProFlex thermal cycler (Applied Biosystems) for

95 degC for 10 min followed by 60 cycles of 94 degC for 05 min and 61 degC for 1 min then 98 degC for 10 min before a 4 degC hold until the plate was removed

bull Droplets were read on the QX200 Droplet Reader and analysis was performed using the QuantaSoft Analysis Software version 1740917

mtND1 mtBatz mtKav

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

8

bull The numbers of positive and negative droplets at the end of 60 cycles were determined and were exported into a spreadsheet for further analysis Assay-specific intensity thresholds were determined by visual inspection for each assay for each measurement session

Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays

Concentration nDNA Mastermix Microliter

per Reaction 2X ldquoSupermix for Probes (no dUTPs)rdquo 125

5 μmolL Forward Primer 1875 5 μmolL Reverse Primer 1875 5 μmolL Probe (FAM) 125

PCR Grade Water 50 14 Diluted Sample 25 Total Volume 250

Mitochondrial DNA (mtDNA) All the mtDNA measurements used in this study followed the nDNA protocol described above but with different mastermix setups The duplex mtDNA measurements used the setup described in Table 4 The monoplex mtDNA measurements used the setup described in Table 5

Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 125

10 μmolL Forward Primer 09375 10 μmolL Reverse Primer 09375 10 μmolL Probe (VIC) 0625

PCR Grade Water 95 nDNA Mastermix from Table 3 05 Total Volume 250

Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 120

10 μmolL Forward Primer 09 10 μmolL Reverse Primer 09 10 μmolL Probe (FAM) 06

PCR Grade Water 72 Diluted Sample (See Table 7) 24

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

9

Total Volume 240 Sample Preparation

The buffy coat white blood cells used to prepare SRM 2372a materials were purchased from Interstate Blood Bank These materials were from four single-source anonymous individuals supplied to us labeled as Female1 Female2 Male1 and Male2 In our laboratories each donor buffy coat bag was aliquoted (5 mL per aliquot) into sterile 50 mL conical tubes and stored at 4 degC prior to extraction The modified salt-out manual extraction protocol described in Appendix A was performed for each aliquot with rehydration of the DNA in TE-4 buffer From 150 mL to 250 mL of solubilized DNA for each component was accumulated through the extraction of the 5 mL aliquots The DNA was stored in perfluoroalkoxy polymer (PFA) containers that had been bleach sterilized thoroughly rinsed with deionized water rinsed with ethanol and air-dried The minimum volume for production of 2000 units was 120 mL Preliminary estimates of the mass [DNA] of the re-solubilized components were obtained using a Nanodrop microvolume spectrophotometer Dilutions of the component materials were made until the absorbance at 260 nm was approximately 1 corresponding to a nominal mass [DNA] of 50 ngμL of dsDNA7 After addition of the TE-4 buffer the containers were placed on an orbital shaker and gently rotated for several hours prior to refrigerated storage at 4 ordmC The four single-source components were initially screened with the nDNA ddPCR measurement process using four nDNA assays 2PR4 NR4Q ND6 and 22C3 Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate The gravimetric 1rarr4 dilution was prepared by weighing an empty PFA vial and recording the weight then pipetting 30 microL TE-4 into the vial and recording the weight then pipetting 10 microL DNA and recording the final weight The dilution factor was calculated in a spreadsheet These screening measurements verified that the individual component materials were fit-for-purpose The four single-source components were evaluated with the nDNA ddPCR measurement process using the ten nDNA assays Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate Quantitative information was derived from the average of the ten nDNA assays to prepare the component C mixture Thirty milliliters of sample Male2 (estimated at 452 ngμL mass [DNA]) were combined with 80 mL of sample Female2 (estimated at 507 ngμL mass [DNA]) and 10 mL of TE-4 buffer to create a 1 male nDNA to 3 female nDNA producing a 14 malehuman mixture After addition of the TE-4 buffer the component C container was gently rotated on the orbital shaker for several hours prior to refrigerated storage at 4 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

10

Absorbance Measurements

In February 2017 absorbance spectra of the recently prepared fully diluted nominally dsDNA and ssDNA solutions of the three components were acquired with a BioCary 100 spectrophotometer using the measurement protocol described in Appendix B In December 2017 additional spectra were acquired for the remaining bulk solution of components A and B and for SRM 2372 components A and B These solutions had been stored at 4 degC in well-sealed PFA containers There was no remaining bulk component C solution SRM Unit Production

In the morning of the day a component was transferred into the sterile component tubes (Sarstedt Biosphere SC micro tube-05 mL Newton NC) used to package the SRM the container holding the diluted component was removed from the refrigerator and placed on a slow orbital shaker for 2 h The tubes were opened and placed under the laminar flow hood in rows of 5 within 80-hole tube racks for a total of 30 tubes per rack Tubes were placed in every second row of the racks to facilitate filling with an Eppendorf Repeater Xstream pipette (Eppendorf North America Inc Hauppauge NY) fitted with a 1-mL positive displacement tip set to dispense 55 microL per tube The component container was removed from the shaker a magnetic stir-bar was added the container was placed on a magnetic stir plate in the laminar flow hood and the material was stirred gently Filling proceeded until 2200 vials were filled with pipet tip replacement after filling every 200 tubes The filling process consisted of two individuals manually filling each vial within a biosafety cabinet (one physically pipetting each sample the other verifying that each tube was filled) and four individuals tightly closing the lids to each tube on a sterilized and covered bench When all tubes in a rack were filled the rack was moved out of the hood the tubes checked for proper filling their lids closed and the tubes transferred to 100-unit pre-labeled storage boxes As the 100-unit boxes were filled they were transferred to the labeling room where SRM labels were applied by hand and colored inserts were added to the lids red for component A white for component B and blue for component C After all tubes for a given component were labeled the 22 boxes were placed in a locked refrigerator Filling of each of the components took 60 min to 90 min All units were transferred to the refrigerator within 4 hours of beginning the filling process All units of the three materials were allowed to equilibrate for a minimum of two weeks prior to beginning homogeneity analysis Component Homogeneity

The ddPCR measurement protocol described above was used to assess the nDNA homogeneity of the three components One tube from the approximate center of each of the 22 storage boxes was assayed using the NEIF and ND6 assays (see Table 1) A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Three technical replicates were obtained for every tube A total

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

11

of six plates were used two plates per component one with the NEIF assay and one with the ND6 assay Figure 4 details the layout of the samples

Figure 4 Plate Layout Design for Homogeneity Measurements

Component Stability

Stability of the materials was evaluated at 4 degC room temperature (RT asymp 22 degC) and 37 degC Tubes from the homogeneity study were divided among the three temperatures for the stability study Additionally two never-opened tubes (one from box 8 and one from box 12 of each component) were held at each temperature Two tubes at each temperature for each component were opened sampled and resealed 4 times over the course of 8 weeks Measurements were made at week 1 week 3 week 5 and week 8 The measurement on week 8 was from an unopened tube (from box 8) for each of the components at all temperatures The data from the homogeneity study was used as timepoint 0 in the stability study anchored to 4 degC by the storage recommendations A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used for both the NEIF and ND6 assays All temperatures components and both assays fit onto one plate for timepoints 1 to 5 Twelve NTCs placed in rows G and H were included for each assay Figure 5 details the sample layout

Figure 5 Plate Layout Design for Stability Measurements

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

ND6 ND6 ND6

NEIFNEIFNEIFComponent A

Component A

Component B

Component B

Component C

Component C

1 2 3 4 5 6 7 8 9 10 11 12A A-7 A-7 B-7 B-7 C-7 C-7 A-7 A-7 B-7 B-7 C-7 C-7B A-10 A-10 B-10 B-10 C-10 C-10 A-10 A-10 B-10 B-10 C-10 C-10C A-9 A-9 B-9 B-9 C-9 C-9 A-9 A-9 B-9 B-9 C-9 C-9D A-12 A-12 B-12 B-12 C-12 C-12 A-12 A-12 B-12 B-12 C-12 C-12E A-8 A-8 B-8 B-8 C-8 C-8 A-8 A-8 B-8 B-8 C-8 C-8F A-11 A-11 B-11 B-11 C-11 C-11 A-11 A-11 B-11 B-11 C-11 C-11G NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTCH NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

NEIF

4C

RT

37C

ND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

12

Component Certification Measurements

After establishing homogeneity and stability measurements intended to be used to certify nDNA values and inform mtDNA values were started using randomly selected SRM tubes from the boxes listed in Table 7 The ten nDNA assays described in Table 1 and two of the three mtDNA assays described in Table 2 were run simultaneously on the same sample plate A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used in each of the six measurement campaigns

Table 6 Tubes used for Certification Measurements

Component Plate A B C

dd20170711 Box 1 Box 1 Box 1 dd20170712a Box 10 tube a Box 10 tube a Box 10 tube a dd20170712b Box 10 tube b Box 10 tube b Box 10 tube b dd20170713 Box 20 Box 20 Box 20 dd20170714 Box 5 Box 5 Box 5 dd20170801 Box 5 Box 5 Box 5

Every sample was analyzed in duplicate with every assay in all six plates Component A was always located in rows B and C component B in rows D and E component C in rows F and G and NTCs in rows A and H Figure 6 displays the basic layout of the ten nDNA and two mtDNA assays In addition to the plate column assignment scheme shown the three assay groupings were also ordered as Group 2 Group 3 Group 1 and Group 3 Group 1 Group 2 to address potential positional bias

Group 1 Group 2 Group 3 2PR4 POTP NEIF NR4Q D5 ND6 D9 HBB1 ND14 22C3 mtND1 mtBatz 1 2 3 4 5 6 7 8 9 10 11 12 A NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Comp A B Comp A C Comp B D Comp B E Comp C F Comp C G

H NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Figure 6 Basic Plate Layout Design for Certification Measurements

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

13

Mitochondrial DNA Measurements

A ldquoduplexed reactionrdquo was used to determine mtDNAnDNA where 1 μL of one of the nDNA mastermix was added to 49 μL of the mtDNA mastermix resulting in a 1rarr50 volumetric dilution of the nDNA The nDNA was targeted with a FAM-labeled probe while the mtDNA was targeted with a VIC labeled probe The primer and probe concentrations were increased from 5 micromol to 10 micromol in the duplexed reactions Additional testing of the mtDNA assays were performed in monoplex with FAM-labeled probes to determine the additional sources of bias between sampling assay setup and dilution factors Table 7 outlines the different methods examined to determine mtDNAnDNA for the three components

Table 7 Run Parameters for Mitochondrial DNA Droplet

Generation nDNA mtDNA

Plate Component Assay Dilution Reps Assay Dilution Reps dd20170824 A manualDG HBB1 1rarr4 3 mtND1 1rarr200 9 dd20170825 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170829 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170831 AB autoDG HBB1 1rarr3 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr6 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr12 2 mtND1 1rarr10rarr100 3 dd20170907 ABC manualDG HBB1 1rarr3 2 mtBatz 1rarr10rarr100 3 dd20170913 AB manualDG NEIF 1rarr4 3 mtND1 mtBatz mtKav 1rarr10rarr100 3 dd20170913 C manualDG NEIF 1rarr4 3 mtBatz mtKav 1rarr10rarr100 3 Performance in Commercial qPCR Chemistries

Table 8 lists the 11 qPCR chemistries that were commercially available in late 2017 All three of the SRM 2372a components were evaluated with these chemistries using component A of SRM 2372 as the reference standard For each component of SRM 2372a four tubes were pulled from box 22 and combined into one tube A 1rarr5 dilution of all samples was prepared followed by three serial 1rarr2 dilutions A five-point serial dilution of each SRM 2372a component was amplified in triplicate for each chemistry This consisted of the neat material the 1rarr5 dilution then the three serial 1rarr2 dilutions A six-point serial dilution of SRM 2372 Component A was also created for the standard curve for the commercial qPCR chemistries This consisted of the neat material followed by a 1rarr5 dilution then followed by four 1rarr2 dilutions serially The six-point serial dilution series of SRM 2372 Component A was also amplified in triplicate These commercial chemistries were amplified in individual plates on an Applied Biosystems 7500 HID Real-Time PCR Instrument (Foster City CA) following the manufacturersrsquo recommended protocols for reaction mix setup and cycling conditions Figure 7 details the plate layout for the qPCR plates Data was analyzed with the HID Real-Time PCR Analysis Software v12 and further analyzed in a spreadsheet

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

14

Table 8 Commercial qPCR Chemistries Tested with SRM 2372a

Manufacturer Commercial Kit Target Codea

Innogenomics

InnoQuant Long IQh Short IQd

InnoQuantHY Long IQHYh Short IQHYd Y IQHYy

Thermo Fisher Quantifiler Duo Human QFDh Male QFDy

Quantifiler HP Large QFHPh Small QFHPd

Quantifiler Human Human QFHh

Quantifiler Trio Large QFTh Small QFTd Y QFTy

Qiagen

Quantiplex Autosomal QPh

Quantiplex Hyres Autosomal QPHh Y QPHy

Quantiplex Pro Human QPPh Degradation QPPd Y QPPy

Promega

Plexor Autosomal Ph Y Py

PowerQuant Autosomal PQh Degradation PQd Y PQy

a Assays are coded with a short acronym derived from the kit name followed by the following codes for the type of target ldquohrdquo autosomal non-degradation described as ldquoAutosomalrdquo ldquoHumanrdquo or ldquoLongrdquo ldquodrdquo autosomal degradation described as ldquoDegradationrdquo or ldquoShortrdquo ldquoyrdquo Y-chromosome described as ldquoMalerdquo or ldquoYrdquo

1 2 3 4 5 6 7 8 9 10 11 12

A

B 2372-A_57 2372-A_57 2372-A_57 2372a-neat_A

2372a-neat_A

2372a-neat_A

2372a-neat_B

2372a-neat_B

2372a-neat_B

2372a-neat_C

2372a-neat_C

2372a-neat_C

C 2372-A_11

4 2372-A_11

4 2372-A_11

4 2372a-15_A

2372a-15_A

2372a-15_A

2372a-15_B

2372a-15_B

2372a-15_B

2372a-15_C

2372a-15_C

2372a-15_C

D 2372-A_5

7 2372-A_5

7 2372-A_5

7 2372a-110_A

2372a-110_A

2372a-110_A

2372a-110_B

2372a-110_B

2372a-110_B

2372a-110_C

2372a-110_C

2372a-110_C

E 2372-A_2

85 2372-A_2

85 2372-A_2

85 2372a-120_A

2372a-120_A

2372a-120_A

2372a-120_B

2372a-120_B

2372a-120_B

2372a-120_C

2372a-120_C

2372a-120_C

F 2372-A_1

425 2372-A_1

425 2372-A_1

425 2372a-140_A

2372a-140_A

2372a-140_A

2372a-140_B

2372a-140_B

2372a-140_B

2372a-140_C

2372a-140_C

2372a-140_C

G 2372-A_0

7125 2372-A_0

7125 2372-A_0

7125

H NTC NTC NTC

Figure 7 Plate Layout Design for Performance Assessments

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

15

Measurement Results After preparing the SRM 2372a bulk solutions their approximate 50 ngmicroL mass [DNA] were confirmed spectrophotometrically The bulk solutions were then packaged into individual tubes The materials contained in the tubes were characterized for homogeneity stability nDNA copy number [DNA] and mtDNAnDNA ratio The following sections detail the results of these measurements Absorbance Spectrophotometry

By widely accepted convention an aqueous DNA solution having an absorbance of 10 at 260 nm corresponds to a mass [DNA] of 50 ngmicroL for dsDNA and (37 to 40) ngmicroL for ssDNA7 Absorbance at (230 270 280 and 330) nm are traditionally used in the assessment of DNA quality725 Due to minor baseline shifts during acquisition of some spectra it was necessary to bias-adjusted all spectra to have absorbance of 00 at 330 nm Table 9 lists the resulting bias-adjusted values for selected wavelengths and the conventional mass [DNA] conversion of the absorbance at 260 nm The close agreement between the mass concentration values for the nominally dsDNA and the converted-to-ssDNA spectra confirm that the bulk solutions were correctly prepared to be asymp 50 ngμL dsDNA The slight increase in absorbance over the 10 months between acquiring the spectra suggests that like SRM 2372 the tertiary structure of the dsDNA in the SRM 2372a materials may be changing slowly with time Figure 8 displays the spectra

Table 9 Absorbance at Selected Wavelengths

Wavelength nm Form Date Component 230 260 270 280 ngμL dsDNA 2142017 A 0394 0931 0758 0488 466

B 0480 1112 0905 0583 556 C 0451 1006 0815 0524 503

ssDNA 2142017 A 0390 0614 0540 0316 454 B 0457 0723 0634 0372 535 C 0411 0641 0563 0330 474

dsDNA 12142017 A 0402 0956 0775 0503 478 B 0484 1128 0914 0589 564

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

16

Figure 8 Absorbance Spectra of SRM 2372a Components A B and C

dsDNA is denatured to ssDNA by diluting with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to DNA mass concentration

Droplet Volume

Droplet volumes for the two lots of ldquoSupermixrdquo used in the studies described in this report were determined using NISTrsquos Special Test Method 11050S-D as described in Dagata et al19 The lot used for most of the ddPCR analyses produced (07349 plusmn 00085) nL droplets where the ldquoplusmnrdquo value is the combined standard uncertainty The lot used for analyses performed after October 11 2017 produced (07376 plusmn 00085) nL droplets NIST measurements have demonstrated that droplet volumes are not influenced by the presence or absence of DNA in the sample solution Measurements repeated over time on an earlier lot of the Supermix product used in these studies suggest that droplet volumes remain constant over the shelf-life of the lot However droplet volumes for different lots of this product as well as for different Supermix products have been observed to differ by several percent We intend to document these measurements in a future report Other researchers have demonstrated that non-constant droplet volume distributions may contribute to ddPCR measurement imprecision and bias2627 However these effects become significant only with distributions wider than observed using Special Test Method 11050S-D19 and at copydroplet values larger than the λ le 05 copydroplet typically used in our measurements Certification Measurements

Homogeneity Results Table 10 lists the λacute nDNA copy number values adjusted for both droplet volume and sample dilution from the homogeneity measurements The data was analyzed using the three-level Gaussian hierarchical ANOVA model described in Appendix C Table 11 lists the values of the within- and between-tube variance components expressed as percent relative standard deviations At the 95 confidence level all components are homogeneous

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372a-A2141721417-NaOH121417

220 260 300 340

Wavelength nm

2372a-B2141721417-NaOH121517

220 260 300 340

Wavelength nm

2372a-C21417

21417-NaOH

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

17

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 NEIF A-1-01 1632 1529 1511 B-1-01 1790 1909 1885 C-1-01 1493 1385 1510 ND6 1616 1640 1582 NAe 1870 1860 1565 1562 1466

NEIF A-1-02 1531 1525 1549 B-1-02 1712 1719 1704 C-1-02 1516 1513 1577 ND6 1615 1589 1601 1775 1718 1756 1490 1547 1570

NEIF A-1-03 1524 1540 1533 B-1-03 1693 1755 1749 C-1-03 1451 1528 1506 ND6 1582 1152 1555 1824 1483 1896 1485 1505 1471

NEIF A-1-04 1454 1458 1427 B-1-04 1754 1745 1739 C-1-04 1440 1486 1520 ND6 1627 1605 1607 1670 1632 1814 1529 1508 1506

NEIF A-1-05 1465 1540 1504 B-1-05 1757 1687 1735 C-1-05 1523 1472 1483 ND6 1607 1633 1586 1683 1759 1703 1513 1520 1491

NEIF A-1-06 1512 1491 1491 B-1-06 1697 1709 1714 C-1-06 1437 1493 1522 ND6 1511 1483 1524 1693 1723 1722 1492 1481 1461

NEIF A-1-07 1418 1480 1455 B-1-07 1722 1625 1703 C-1-07 1485 1477 1463 ND6 1479 1557 1656 1690 1684 1688 1475 1547 1485

NEIF A-1-08 1496 1560 1501 B-1-08 1884 1802 1772 C-1-08 1433 1405 1478 ND6 1514 1615 1540 1864 1887 1821 1497 1501 1495

NEIF A-1-09 1588 1581 1637 B-1-09 1713 1734 1727 C-1-09 1463 1499 1505 ND6 1474 1438 1454 1709 1756 1758 1503 1491 1516

NEIF A-1-10 1574 1513 1570 B-1-10 1759 1809 1738 C-1-10 1534 1458 1454 ND6 1594 1713 1669 1764 1816 1787 1448 1465 1474

NEIF A-1-11 1557 1508 1633 B-1-11 1749 1782 1740 C-1-11 1533 1514 1555 ND6 1571 1611 1631 1741 1792 1784 1488 1470 1483

NEIF A-1-12 1500 1476 1515 B-1-12 1748 1712 1718 C-1-12 1450 1525 1497 ND6 1626 1586 1608 1797 1614 1790 1496 1483 1519

NEIF A-1-13 1672 1634 1804 B-1-13 1632 1668 1701 C-1-13 1537 1527 1515 ND6 1490 1516 1497 1626 1566 1681 1499 1479 1475

NEIF A-1-14 1551 1568 1538 B-1-14 1752 1535 1673 C-1-14 1515 1490 1494 ND6 1617 1614 1639 1724 1741 1685 1469 1480 1466

NEIF A-1-15b 1793 1722 1733 B-1-15 1871 1835 1887 C-1-15 1511 1568 1594 ND6 1818 1753 1762 1770 1752 1798 1515 1543 1565

NEIF A-1-16 1573 1520 1617 B-1-16 1688 1725 1743 C-1-16 1449 1452 1471 ND6 1634 1627 1575 1744 1697 1710 1543 1448 1491

NEIF A-1-17 1516 1516 1628 B-1-17 1789 1769 1825 C-1-17 1388 1421 1394 ND6 1613 1567 1567 1716 1762 1638 1452 NAe 1435

NEIF A-1-18 1612 1535 1672 B-1-18 1826 1780 1849 C-1-18 1475 1455 1445 ND6 1529 1495 1528 1805 1801 1931 1458 1531 1459

NEIF A-1-19 1598 1514 1596 B-1-19 1751 1759 1880 C-1-19 1468 1525 1446 ND6 1571 1535 1511 1827 1766 1786 1435 1478 1421

NEIF A-1-20 1575 1568 1586 B-1-20 1738 1678 1768 C-1-20 1549 1531 1499 ND6 1652 1545 1627 1752 1747 1768 1499 1465 1468

NEIF A-1-21 1658 1611 1578 B-1-21 1742 1836 1713 C-1-21 1530 1440 1484 ND6 1647 1607 1612 1746 1744 1764 1499 1454 NAe

NEIF A-1-22 1547 1528 1532 B-1-22 1761 1743 1660 C-1-22 1442 1487 1436 ND6 1537 1513 1536 1666 1692 1819 1589 1591 1556

NEIF A-1-15c 1456 1448 1523 ND6 1525 1402 1436

NEIF A-2-15d 1679 1599 1674 ND6 1581 1550 1584

a Tube coded as Component-Tube-Box b These results were identified as a probable error in the preparation or volumetric delivery of the sample

solution into the microfluidic device and were not used in the homogeneity evaluations c Repeat assessment of the solution in sample A-1-15 d Results for a second tube from box 15 analyzed at the same time as the repeat assessment of A-1-15 e Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

18

Table 11 Within- and Between-Tube Relative Standard Deviations

Within-Tube Component NEIF ND6 Between-Tube

A 28 40 33 B 27 38 25 C 23 18 19

Stability Results Table 12 lists the λacute volume- and dilution-adjusted nDNA copy number values from the stability measurements Figure 9 summarizes these results The results for the tubes held at 4 degC room temperature (RT asymp 22 degC) and 37 degC are very similar for all three components The absence of between-temperature differences identifies the between-week differences as attributable to variation in the measurement processes rather than sample instability These results indicate that the SRM 2372a genomic copy number is thermally stable from 4 degC to 37 degC over an extended period

Figure 9 ddPCR Stability Measurements as Function of Time

Each dot-and-bar symbol represents the mean λacute and the approximate 95 expanded uncertainty for all results for a given component at a given storage temperature The thick black horizontal lines represent an approximate 95 confidence interval on the λacute values estimated from the homogeneity data

13

14

15

16

17

18

19

0 1 2 3 4 5 6 7 8

λacute H

aplo

id n

DNA

Copi

es μL

Weeks

A

0 1 2 3 4 5 6 7 8

Weeks

B

4 degC 22 degC 37 degC

0 1 2 3 4 5 6 7 8

Weeks

C

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

19

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Week Temp Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3

1 4 degC NEIF A-10 1568 1441 B-10 1666 1653 C-10 1424 1394 ND6 A-10 1474 1494 B-10 1655 1705 C-10 1461 1394

NEIF A-07 1500 1463 B-07 1718 1640 C-07 1446 1452 ND6 A-07 1479 1479 B-07 1708 1686 C-07 1522 1482

RT (22 degC) NEIF A-12 1674 1608 B-12 1805 1777 C-12 1506 1495 ND6 A-12 1522 1507 B-12 1759 1788 C-12 1478 1461 NEIF A-09 1635 1539 B-09 1791 1793 C-09 1411 NAb ND6 A-09 1543 1519 B-09 1661 1682 C-09 1482 1496

37 degC NEIF A-11 1614 1637 B-11 1748 1744 C-11 1515 1464 ND6 A-11 1494 1502 B-11 1740 1697 C-11 1495 1535 NEIF A-08 1592 1614 B-08 1691 1725 C-08 1528 1500 ND6 A-08 1581 1595 B-08 1727 1699 C-08 1458 1483 3 4 degC NEIF A-10 1520 1706 B-10 NAb 1733 C-10 1427 1473 ND6 A-10 1543 1624 B-10 1670 1755 C-10 1443 1364 NEIF A-07 1537 1570 B-07 1549 1565 C-07 1424 1512 ND6 A-07 1517 1479 B-07 1705 1723 C-07 1483 1439

RT (22 degC) NEIF A-12 1521 1607 B-12 1817 1842 C-12 1566 1471 ND6 A-12 1644 1597 B-12 1845 1786 C-12 1504 1511 NEIF A-09 1432 1626 B-09 1825 1856 C-09 1392 1387 ND6 A-09 1539 1528 B-09 1734 1737 C-09 1535 1465

37 degC NEIF A-11 1651 1685 B-11 1759 1779 C-11 1437 1440 ND6 A-11 1680 1594 B-11 1806 1806 C-11 1471 1509 NEIF A-08 1580 1613 B-08 1740 1791 C-08 1582 1590 ND6 A-08 1592 1524 B-08 1769 1694 C-08 1468 1520 5 4 degC NEIF A-10 1486 1440 B-10 1626 1551 C-10 1381 1330 ND6 A-10 1555 1577 B-10 1673 1625 C-10 1547 1395 NEIF A-07 1387 1429 B-07 1723 1787 C-07 1330 1314 ND6 A-07 1538 1522 B-07 1862 1909 C-07 1452 1458

RT (22 degC) NEIF A-12 1601 1631 B-12 1718 1752 C-12 1546 1541 ND6 A-12 1699 1677 B-12 1772 1721 C-12 1490 1471 NEIF A-09 1381 1484 B-09 1675 1694 C-09 1371 1376 ND6 A-09 1625 1615 B-09 1904 1815 C-09 1468 1472

37 degC NEIF A-11 1456 1517 B-11 1720 1704 C-11 1505 1493 ND6 A-11 1650 1658 B-11 1148 1152 C-11 1527 1600 NEIF A-08 1590 1666 B-08 1745 1723 C-08 1469 1540 ND6 A-08 1235 1244 B-08 1809 1749 C-08 1508 1529 8 4 degC NEIF A-08a 1667 1632 1589 B-08a 1880 1884 1940 C-08a 1525 1475 1542 ND6 A-08a 1599 1593 1590 B-08a 1859 1897 1972 C-08a 1550 1502 1520

RT (22 degC) NEIF A-08b 1693 1666 1634 B-08b 1769 1884 1771 C-08b 1595 1557 1545 ND6 A-08b 1644 1620 1647 B-08b 1862 1738 1811 C-08b 1611 1585 1535

37 degC NEIF A-08c 1620 1576 1608 B-08c 1934 1952 1956 C-08c 1595 1600 1569 ND6 A-08c 1576 1553 1532 B-08c 1889 1844 1889 C-08c 1569 1514 1581

a Tube coded as Component-Tube-Box b Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

20

Ten-Assay Certification Results The original intention was to certify the copy number [DNA] of the three components with results from five independently prepared sets of samples analyzed over four days Each sample of the three components was from a new tube from different boxes and was evaluated with 10 nDNA assays The assay positions were alternated across these five ldquo10-assay certificationrdquo plates The λ results for these five plates are systematically somewhat larger than those for the homogeneity and stability data On review we realized that a different workspace and its suite of pipettes had been used to prepare the 1rarr10 dilution of the 1rarr4 diluted components into the reaction mastermix The ldquohighrdquo workstation was used for stability and homogeneity measurements while the ldquolowrdquo workstation was used for certification measurements Gravimetric testing of the pipettes established that the volumes delivered by these alternate pipettes was sufficiently biased to account for the observed λ differences A sixth ldquocertificationrdquo plate was prepared using the original (and now re-verified) pipettes The λ results or this plate agreed well with the homogeneity and stability data Table 13 lists the λacute volume- and dilution-adjusted nDNA copy number values for all six of the ten-assay datasets Figure 10 displays the extent of agreement among the ten assays While there are small-but-consistent differences between the assays there is only a 4 difference between the assays providing smallest and largest λ

Figure 10 Comparison of Assay Performance

The blue dot-and-bars represent the mean and its combined standard uncertainty for the A B and C components for each assay relative to the over-all grand mean for each component The black dot-and-bars represent combination of three component results for each assay The solid horizontal line represents the median difference of the assay means from the grand mean the dotted horizontal lines bound an approximate 95 confidence region about the median

2PR4POTPNEIFR4Q5 D5ND6 D9 HBB1ND14 22C3-10

-8

-6

-4

-2

0

2

4

6

8

10

Rela

tive

Diffe

renc

e fr

om M

ean

Assays

Component MeansAssay MeansMedianU95

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

21

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

7112017 2PR4 1667 1698 1869 1960 1630 1706 POTP 1655 1685 1850 1781 1596 1576 NEIF 1647 1665 1709 1707 1574 1538 R4Q5 1594 1658 1856 1787 1590 1555 D5 1655 1683 1848 1829 1610 1607 ND6 1629 1647 1828 1826 1589 1548 D9 1655 1652 1861 1858 1569 1572 HBB1 1644 1611 1825 1869 1565 1681 ND14 1645 1707 1871 1773 1564 1492 22C3 1545 1603 1776 1902 1548 1655

7122017 2PR4 1668 1739 2174 2231 1499 1497 POTP 1587 1702 2202 2292 1576 1611 NEIF 1678 1614 2172 2216 1545 1560 R4Q5 1628 1762 NA 2078 1584 1546 D5 1738 1649 2230 2192 1577 1529 ND6 1667 1678 2094 2034 1599 1478 D9 1695 1653 1949 2110 1587 1504 HBB1 1757 1750 1971 2123 1520 1551 ND14 1771 1764 2123 2066 1479 1438 22C3 1762 1759 2107 2046 1512 1494

7122017 2PR4 1638 1667 1810 1806 1452 1424 POTP 1676 1638 1865 1801 1347 1397 NEIF 1616 1624 1806 1786 1395 1350 R4Q5 1559 1587 1878 1768 1442 1459 D5 1608 1652 1824 1790 1448 1602 ND6 1611 1589 1768 1670 1463 1435 D9 1517 1607 1811 1749 1402 1433 HBB1 1557 1582 1763 1741 1474 1481 ND14 1604 1625 1865 1781 1466 1482 22C3 1675 1616 1815 1866 1439 1568

7132017 2PR4 1888 1939 1784 1821 1470 1444 POTP 1868 1815 1735 1813 1411 1427 NEIF 1953 1902 1778 1813 1512 1439 R4Q5 1779 1871 1835 1749 1452 1468 D5 1924 2124 1794 1848 1567 1524 ND6 1861 1787 1700 1711 1480 1424 D9 1824 1844 1799 1779 1442 1459 HBB1 1809 1799 1765 1769 1430 1467 ND14 1864 1857 1754 1725 1413 1411 22C3 1930 1942 1792 1814 1486 1480

7142017 2PR4 1597 1590 1718 1768 1485 1458 POTP 1533 1404 1753 1698 1422 1459 NEIF 1462 1533 1773 1792 1438 1465 R4Q5 1493 1519 1777 1740 1439 1468 D5 1538 1586 1803 1795 1515 1457 ND6 1492 1516 1718 1693 1409 1412 D9 1528 1538 1802 1743 1469 1474 HBB1 1548 1581 1834 1745 1453 1428 ND14 1513 1514 1731 1714 1465 1425 22C3 1562 NA 1842 NA 1363 1393

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

22

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

101217 2PR4 1431 1411 1791 1770 1357 1364 POTP 1407 1395 1746 1820 1447 1361 NEIF 1487 1398 1803 1803 1448 1404 R4Q5 1450 1439 1767 1739 1409 1411 D5 1394 1422 1745 1773 1389 1429 ND6 1396 1393 1824 1729 1415 1417 D9 1414 1390 1752 1744 1431 1380 HBB1 1401 1382 1813 1754 1381 1336 ND14 1412 1460 1648 1671 1403 1387 22C3 1435 1410 1818 1741 1435 1372

Estimation of Copy Number Concentration It was decided that all three sets of results (homogeneity stability and certification) would be used to derive the certified values with the λ of the first five ldquoten-assayrdquo plates adjusted by a data-driven estimate of the volumetric bias This bias was incorporated into the statistical model providing better estimates of the λ means while capturing all of the variance components (homogeneity stability between-assay and between data sets) as random effects using Bayesian MCMC programmed in OpenBUGS28 Table 14 lists the estimated values and their uncertainties See Appendix C for details

Table 14 Measured Copy Number Concentrations

Genomic Copies per Nanoliter of Solution Component Mean a u(Mean) b U95(Mean) c CV95

d A 151 07 14 93 B 175 03 06 34 C 145 05 10 69

a Mean the experimentally determined most probable estimate of the true value b u(Mean) the standard uncertainty of the Mean c Mean plusmnU95(Mean) the interval which is believed with about 95 confidence to contain the true value d CV95 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the counting unit one29 2) the validity of the Poisson endpoint transformation for digital PCR endpoint assays when applied to samples providing le 05 copies per droplet 3) the determination that the copies are dispersed as dsDNA and 4) calibrated mean droplet volume measurements made at NIST during sample dilution and mastermix preparation Conversion of Copy Number to Mass Concentration As described in Reference 18 metrologically valid conversion from copies per nanoliter to nanograms per microliter requires quantitative estimates for the number of nucleotide bases per reference human genome and the average nucleotide mass Using the values estimated in that report and asserting that each target copy corresponds to one HHGE the required transformation is

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

23

[DNA] ngmicroL

= [DNA] HHGE

nLtimes

3301 ngHHGE

The combined standard uncertainty is

119906119906 [DNA] ng

microL = 1199061199062 [DNA] HHGE

nL + 0531100

times[DNA] HHGE

nL 2

Both the measured and the conversion factor uncertainties are associated with ldquolargerdquo degrees of freedom Therefor the approximate 95 confidence uncertainty is estimated

11988011988095 [DNA] ng

microL = 2 times 119906119906

[DNA] ngmicroL

Table 15 lists the transformation of the Table 14 copy number values These values agree well with the conventional 260 nm absorbance values averaging (4 plusmn 5) larger

Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution

ngmicroL Component Value a u(Measurement)b u(Conversion)c U95(Value)d CV95

e A 498 23 03 47 93 B 578 10 03 21 36 C 479 17 03 33 70

a Estimated true value b Standard uncertainty component from the ddPCR measurements c Standard uncertainty estimated from the relative uncertainty of the conversion factor d Value plusmnU95(Value) is believed with about 95 confidence to contain the true value e 100U95(Value)Value the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the measured estimates of copies per nanoliter and 2) determinations made by us from internationally recognized literature resources of the average number of nucleotide bases per HHGE and the mean molecular weights of the bases18 Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)

While not intended to provide certified values mtDNAnDNA was measured as part of 13 experiments over 5 months Measurements were made using dilution-adjusted λ for the mtND1 mtBatz andor mtKav mitochondrial assays referenced against the dilution-adjusted λ of either the NEIF or HBB1 nDNA assays By chance both DNAs in the component C mixture have a binding site mutation that silences the mtND1 probe Figure 11 graphically summarizes the results by assay and component Table 16 lists the data Table 17 summarizes results from a random effects model that accounts for between-assay variability See Appendix C for details

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

24

Due to the limited number of mtDNA assays used in these measurements these results are metrologically traceable only to the assays used

Figure 11 mtDNAnDNA for Components A B and C

Each dot-and-bar symbol represents a mean mtDNAnDNA and its approximate 95 expanded uncertainty The dotted horizontal lines represent approximate 95 confidence intervals on the ratios

Table 16 mtDNAnDNA Ratios

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

20170711a 1 mtND1 1931 1870 2461 2401 1 mtBatz 1845 1764 2555 2608 3201 3299

20170711b 1 mtND1 1810 1906 2229 2200 1 mtBatz 1785 1837 2108 2192 3347 3287

20170712 1 mtND1 1743 1761 2127 2198 1 mtBatz 1829 1847 2455 2506 3217 3384

20170713 1 mtND1 1716 1800 2211 2199 1 mtBatz 1740 1813 2380 2249 3200 3363

20170714 1 mtND1 1923 1835 2105 1943 1 mtBatz 1935 1851 2136 2119 3294 3013

20170814 1 mtND1 1780 1776 1706 2046 2057 2024 1 mtBatz 1804 1751 1738 1998 2043 1952 2927 2776 2762 1 mtKav 1769 1786 1761 2077 2117 2041 3044 2830 2825

20170824 1 mtND1 1947 1928 1922 2 mtND1 1829 1895 1851 3 mtND1 1674 1617

20170825 1 mtBatz 1667 1813 1708 1598 1928 1916 1915 2832 2840 2788 2725 2 mtBatz 1566 1687 1663 1589 1947 1946 1961 1909 2835 2790 2736 2679 3 mtBatz 1643 1668 1604 1573 1953 1952 1941 1881 2803 2773 2757 2752

20170829 1 mtBatz 1719 1716 1670 1666 1865 1863 1828 1828 2599 2508 2551 2609 2 mtBatz 1581 1573 1601 1612 1944 1964 1947 1895 2617 2483 2626 2602 3 mtBatz 1546 1542 1538 1587 1908 1909 1940 1839 2665 2535 2570 2650

20170831 1 mtND1 1617 1599 1688 1783

ND1

ND1

ND1Ka

v

Kav

Kav

Batz

Batz

Batz

A

B

C

160

210

260

mtD

NAn

DNA

Ratio

Assay

Bind

ing s

ite m

utat

ion

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

25

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

2 mtND1 1622 1579 1823 1631 3 mtND1 1575 1528 1920 1887

20170907 1 mtBatz 1584 1544 1890 1886 2501 2522 20170913 1 mtND1 1757 1724 1768 1990 1921 1982

1 mtBatz 1743 1671 1766 1990 1924 1958 2805 2825 2829 1 mtKav 1692 1662 1767 1903 1831 1919 2709 2738 2727

20171016 1 mtND1 1924 1944 1893 2304 2341 2332 1 mtBatz 1786 1752 1743 2048 2078 2067 2797 2642 2792 1 mtKav 1530 1542 1523 2399 2479 2493 2804 2789 2842 2 mtBatz 1580 1557 1592 2715 2606 2687 2 mtKav 2103 2100 2049 2799 2628 2559 3 mtND1 1749 1769 1777 2065 2166 2164 3 mtBatz 1965 1889 1947 2314 2280 2258 2639 2535 2657 3 mtKav 1813 1864 1864 1867 1987 1971 2953 2757 2802

Table 17 mtDNAnDNA Summary Results

mtDNAnDNA Component Value a u(Value)b U95(Value)c CV95

d A 1737 16 38 17 B 2058 22 44 21 C 2791 23 47 17

a Estimated true value b Standard uncertainty c Value plusmnU95(Value) is believed with about 95 confidence to contain the true value d 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) Performance in Commercial qPCR Chemistries

Traditional qPCR assays only quantify relative to a reference standard Thus qPCR assays do not provide absolute [DNA] for the individual materials but they can provide [DNA] of one component relative to another Figure 12 summarizes the BA and CA ratios for the commercial autosomal (Human) targets from the eleven assays evaluated (see Table 8) The Y-chromosome (Male) assays cannot be summarized in this manner as the (all female) component B does not include a Y chromosome Figure 13 summaries the mass [DNA] results for the (all male) component A and (male-female mixture) component C Since this material was prepared as a nominal 13 malefemale mixture the expected malehuman (total) ratio is 14 sjc As displayed in these figures there is intrinsic variability between the SRM 2372a components and between the commercial qPCR chemistries examined This reflects the difference in the targets and amplification chemistries for each of the commercial kits Many of the commercial chemistries on the market to date contain multi-copy targets for both autosomal and Y-chromosomal quantification to increase the sensitivity of the assays

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

26

Figure 12 qPCR Autosomal Quantitation Performance Summary

Open circles denote the BA CA ratios the 17 commercial qPCR assays evaluated each labeled with the code listed in Table 8 The solid black circle represents ratios as estimated by ddPCR The red lines bound the approximate 95 confidence intervals on the ddPCR-estimated ratios

Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary

Open squares denote the ngμL values for SRM 2372a Components A (male) and C (mixture of male and female) for the seven commercial Y-chromosome qPCR assays evaluated each labeled with the code listed in Table 8 The dotted red lines represent malehuman ratios of 13 14 and 15

Appendix D provides a graphical analysis of all the data generated in the performance assessments for each of the targets amplified in the eleven commercial qPCR chemistries and the [DNA] values for all of the assays

dd

IQ_h

IQ_d

IQHY_h

IQHY_d

QFD_h

QFH_h QFHP_h

QFHP_dQFT_h

QFT_d

QPH_h

QPP_hQPP_d

P_h

PQ_h

PQ_d

06

08

10

12

14

07 09 11 13

CA

BA

IQHY_y

QFD_y

QFT_y

QPH_y

QPP_y

P_y

PQ_y

1413

15

10

12

14

16

18

30 40 50 60 70

[DNA

] C n

gμL

[DNA]A nguL

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

27

Qualification of ddPCR as Fit-For-Purpose We have documented the requirements for establishing metrological traceability of ddPCR human genomic assay results expressed as copies per nanoliter of sample in a series of three peer-reviewed papers151617 Through the experimental evidence they are based upon these papers demonstrate that ddPCR can accurately enumerate the number of independently migrating genomic copies in an aqueous extract Our most recent paper establishes the metrological traceability of these ddPCR results when transformed into the nanogram per microliter units used by the forensic and clinical communities18 However our previous studies have not addressed a fundamental requirement for correctly interpreting digital PCR measurements In broad simplification end-point assay systems count the number of independently partitioning entities regardless of the number of copies the entities contain With assays that target just one site per HHGE entities that migrate as ssDNA will yield counts that are twice the value had they migrated as dsDNA30 Should some originally dsDNA be converted over time to ssDNA ddPCR measurements of the number of independently migrating entities would increase although the number of HHGE would be unchanged Measurements of mass [DNA] provided by absorbance spectrophotometry and ddPCR reflect different properties absorbance at 260 nm is sensitive to the spatial configuration of the nucleotides as well as their number while ddPCR counts independently migrating entities The SRM 2372a materials have absorbance spectra shown in Figure 8 consistent with being predominantly dsDNA The SRM 2372 materials now have the absorbance spectra shown in Figure 14 which are consistent with them being entirely ssDNA

Figure 14 Absorbance Spectra of SRM 2372 Components A and B dsDNA is denatured to ssDNA by diluting the source material with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to mass [DNA]

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372-A200720122012-NaOH121417

0

4

6

0

4

6

220 260 300 340

Wavelength nm

2372-B200720122012-NaOH121517

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

28

To ensure the validity of ddPCR-based value assignments it is thus necessary to evaluate the nature of the partitioning entities Should a solution of what is thought to be dsDNA contain an appreciable fraction of ssDNA then the ddPCR results would require adjustment31 To avoid further delay in making SRM 2372a available to the forensic community the following sections detail a series of experiments that have enabled us to conclude with high confidence that the SRM 2372a components have a dsDNA conformation and that for this material ldquoentitiesrdquo can be considered dsDNA ldquocopiesrdquo The remaining stock of the SRM 2372 components also have a dsDNA conformation however the strands in the old materials are much less robustly connected than they are in the new materials Digital Polymerase Chain Reaction Assays

Three nDNA assays were used HBB1 NEIF and ND6 The HBB1 and NEIF assays reproducibly yield very similar results for DNA extracts from many different donors The HBB1 target is at the tip of the short (p) arm of chromosome 11 The NEIF target is in the centromere region of chromosome 2 The ND6 assay was selected for use because it typically yields the lowest results of the ten nDNA assays used to certify the SRM 2372a components (see Figure 10 below) The ND6 target is about halfway down the long (q) arm of chromosome 6 Heat Treatments (Melting of the DNA into Single Strands)

In all experiments the non-heated control and the heat-treated diluted solutions were stored in the same type of PCR reaction tube The starting point for the heat treatment was to determine if there was a difference between snap-cooling on ice and slower cooling at 25 degC Two tubes of the diluted material were placed in a GeneAmp 9700 thermal cycler (Applied Biosystems) and cycled up to 98 degC and held for 15 min After 15 min one tube was removed and placed on ice while the remaining tube was cooled at 1 degCs in the thermal cycler to 25 degC then placed on ice until prepared for ddPCR Additional heat treatments included cycling samples on a GeneAmp 9700 thermal cycler

a) from 98 degC for 5 min then cooling to 25 degC for 2 min During the 25 degC hold a randomly selected tube was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 45 min at 98 degC

b) from 94 degC for 1 min then cooling to 25 degC for 1 min During the 1 min hold at 25 degC a random sample was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 9 min at 94 degC

c) from 90 degC for 1 min then cooling to 25 degC for 1 min for a total of 20 cycles Samples were pulled at 25 degC after 1 min 3 min 5 min 10 min and 20 min and placed on ice

After the above experiments heat treatments were performed using a ProFlex thermal cycler This thermal cycler has six temperature zones that can be le 5 degC apart This enables heating

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

29

tubes at the same time but at different temperatures After treatment tubes were cooled at 5 degCs to 4 degC in the thermal cycler While designed for use with PCR plates a MicroAmp 96-well tray for the VeriFlex block enabled its use with single PCR reaction tubes Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a

As a first step in evaluating whether the DNA in the SRM materials disperse as dsDNA or ssDNA a direct comparison of the five materials was made using the HBB1 and NEIF assays Table 18 reports summary statistics for the ddPCR measurements reported as λ copiesdroplet and the estimated mass [DNA] values from the ddPCR results18 from the conventional dsDNA relationship7 and from the conventional ssDNA relationship after denaturing with 04 molL NaOH89 While the mass [DNA] values differ somewhat between the three methods there is no evidence for the SRM materials segregating exclusively or even mostly as ssDNA The extent of the hypochromic interactions among the nucleotides does not accurately reflect the extent to which the two strands of dsDNA have separated

Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a

λ copies per droplet λ copies per droplet ngmicroL Sample Assay Nrep

a Mean CVb ncmbc Mean CVb ddPCRd dsDNAe ssDNAf

2372-A HBB1 4 03645 19 8 03644 20 566g 524 57 2372-A NEIF 4 03644 23 2372-B HBB1 4 03513 23 8 03487 35 580h 536 61 2372-B NEIF 4 03461 47 2372a-A HBB1 4 02603 27 8 02649 30 474 465 454 2372a-A NEIF 4 02694 24 2372a-B HBB1 4 02908 34 8 02891 37 517 556 534 2372a-B NEIF 4 02873 45 2372a-C HBB1 4 02757 28 8 02711 28 485 502 474 2372a-C NEIF 4 02665 16

a Number of technical replicates for the given assay b CV = 100(standard deviation)mean c Combined number of technical replicates ignoring potential differences between the assays d Estimated from the ddPCR λ measurements listed in this Table18 e Estimated from the conventional relationship between the mass [DNA] of dsDNA and absorbance at

260 nm7 using absorbance spectra acquired shortly after solutions were produced f Estimated from the relationship between the mass [DNA] of ssDNA and absorbance at 260 nm after 11

dilution with 04 molL NaOH89 g Values adjusted to measured (132 plusmn19) volume loss of archived units due to evaporation h Values adjusted to measured (71 plusmn25) volume loss of archived due to evaporation

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

30

Quantitative Evaluation of ssDNA Content

Absorbance spectrophotometry is routinely used to quantitatively assess DNA tertiary structure3233 On storage ldquoat physiological pH the intrastrand repulsion between the negatively charged phosphate groups forces the double helix into more rigid rodlike conformationrdquo34 suggesting a mechanism for the observed slow changes in the SRM 2372 materials As demonstrated in Table 18 absorbance at 260 nm does not necessarily correlate with ddPCR behavior Yet ldquofurthermore the repulsion between phosphate groups on opposite strands tends to separate the complementary strandsrdquo34 Therefore a method that is independent of hypochromism is needed to adequately evaluate the proportion (if any) of ldquotruerdquo (completely disjoint) ssDNA in the SRM materials Circular dichroism spectroscopy is sensitive to the conformational structures of DNA in solution35 and different DNA configurations migrate differentially in agarose gel electrophoresis7 Both techniques have been used to qualitatively discriminate ssDNA from dsDNA in ldquopurerdquo single-sequence materials but neither technique appears to be able to quantify small proportions of ssDNA in a complex multi-chromosome genomic material Wilson and Ellison have proposed a real-time chamber digital PCR (cdPCR) technique that can estimate excess proportions of ssDNA copies from the crossing-threshold distribution31 However their method is insufficiently sensitive for low ssDNA proportions It is well known that heating DNA beyond the melting temperature separates mammalian dsDNA into ssDNA36 albeit we have shown using cdPCR that boiling dsDNA for some minutes also renders some target sites non-amplifiable15 In the following section we document a ddPCR-based approach for estimating the proportion if any of ssDNA in a sample that involves less destructive heating While the method is not yet optimized the results from these preliminary studies demonstrate that the SRM 2372 and 2372a components disperse into the ddPCR droplets as dsDNA ddPCR Measurement of ssDNA Proportion Given two otherwise identical aliquots of a given DNA material one subjected to some treatment and the other untreated the ratio between the average number of copies per droplet of treated material and that of its untreated sibling λtλu quantitatively estimates the change in the number of independently dispersing entities Should the treatment reduce the number of accessible amplifiable targets (AAT) λtλu will be less than 1 Reduction in AAT may result from damage to the target itself andor conformational changes to the fragment containing the target that render the target inaccessible or non-amplifiable16 Assuming no denaturation of dsDNA to ssDNA either because of an ineffective treatment or the material having started as ldquopurerdquo ssDNA if the treatment does not itself damage targets then λtλu will equal 1 Should treatment convert at least some dsDNA to ssDNA without damaging targets the number of AAT will increase and λtλu will be greater than 1 Assuming complete conversion of ldquopurerdquo dsDNA to ldquopurerdquo ssDNA without other damage then λtλu will equal 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

31

Values greater than 1 but less than 2 indicate some conversion of dsDNA to ssDNA but also some combination of (a) incomplete denaturation of dsDNA to ssDNA (b) partial renaturation of ssDNA to dsDNA after treatment but before ddPCR analysis (c) damage to targets during the process or ndash of greatest interest ndash (d) the original sample material was an impure mixture of dsDNA and ssDNA The following discussions graphically summarize sets of λtλu followed by tabular listings of the relevant experimental conditions ddPCR measurements and the numerical λtλu values The information in these tables includes

Plate Index Experiment identifier indicates date performed in YYYYMMDD format Material Sample material SRM 2372-A 2372-B 2372a-A 2372a-B or 2372a-C Assay ddPCR assay HBB1 NEIF or ND6 degC Denaturation temperature in degC Min Length of time the treated material was held at denaturation temperature in

minutes Cool Method used to cool treated material from denaturation temperature Snap

Slow R1 or R5 where ldquoSnaprdquo indicates cooling on ice ldquoSlowrdquo cooling at room temperature to 25 degC ldquoR1rdquo cooling in a thermal cycler to 25 degC at 1 degCs and ldquoR5rdquo cooling in a thermal cycler to 4 degC at 5 degCs

F Dilution factor expressed as (volume Material) per (final volume of diluted solution) 12 14 or 18

n Number of technical replicates Mean Mean of technical replicates CV Coefficient of Variation expressed as a percentage 100 Standard Deviation

Mean

λtλu the copies per droplet result of the treated solution λt divided by the copies per droplet result of its untreated sibling λu

u(λtλu) Standard uncertainty of λtλu 119906119906(120582120582t 120582120582ufrasl ) = 120582120582t 120582120582ufrasl CV(120582120582t)100

2

+ CV(120582120582u)100

2

Proof of Principle Figure 15 summarizes two replicate sets of measurements performed to assess whether the SRM 2372 and 2372a materials respond similarly whether the denatured materials renature if allowed to cool slowly and whether results are (reasonably) repeatable Table 19 details the experimental conditions The λtλu values for the four materials are similar all much greater than 1 but not approaching 2 Most of the ratios are on or close to the ldquoslow cooling is the same as snap coolingrdquo line suggesting that renaturation is not a significant concern While the replicate results for the SRM 2372-A and 2372a-A materials do not overlap as closely as those for SRM 2372-B and 2372a-B similar treatment conditions produce similar results

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

32

Figure 15 Comparison of Snap-Cooling to Slow-Cooling

Symbols mark the means of replicate measurements made using the same dilutions on successive days with bars representing plusmn one standard combined uncertainty Results are labeled by material ldquoArdquo and ldquoBrdquo for SRM 2372 components and ldquoaArdquo ldquoaBrdquo and ldquoaCrdquo for SRM 2372a components The diagonal black line represents equality between the snap- and slow-cooling treatments

Table 19 Comparison of Snap-Cooling to Slow-Cooling

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171218 2372-A HBB1 14 4 03830 29

dd20171218 2372-A HBB1 98 15 Slow 14 2 05644 18 1474 0051 dd20171218 2372-A HBB1 98 15 Snap 14 4 05852 15 1528 0051

dd20171218 2372a-A HBB1 14 4 03029 18

dd20171218 2372a-A HBB1 98 15 Slow 14 4 04430 21 1463 0040 dd20171218 2372a-A HBB1 98 15 Snap 14 4 04454 33 1471 0055

dd20171218 2372-B HBB1 14 4 03594 31

dd20171218 2372-B HBB1 98 15 Slow 14 4 05806 18 1615 0058 dd20171218 2372-B HBB1 98 15 Snap 14 4 05825 13 1620 0055

dd20171218 2372a-B HBB1 14 4 03383 15

dd20171218 2372a-B HBB1 98 15 Slow 14 4 05043 38 1491 0061 dd20171218 2372a-B HBB1 98 15 Snap 14 4 05254 20 1553 0039

dd20171219 2372-A HBB1 14 4 03357 23

dd20171219 2372-A HBB1 98 15 Slow 14 4 05340 16 1591 0046 dd20171219 2372-A HBB1 98 15 Snap 14 4 05340 12 1591 0042

A

aA

B

aB

A

aA

B

aB

14

15

16

17

14 15 16 17

λ tλ

u Sn

ap C

oole

d

λtλu Slow Cooled

Rep 1Rep 2Snap = Slow

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

33

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171219 2372a-A HBB1 14 4 02735 24

dd20171219 2372a-A HBB1 98 15 Slow 14 4 04151 03 1518 0037 dd20171219 2372a-A HBB1 98 15 Snap 14 4 03918 21 1433 0045 dd20171219 2372-B HBB1 14 4 03346 22

dd20171219 2372-B HBB1 98 15 Slow 14 4 05476 17 1637 0045 dd20171219 2372-B HBB1 98 15 Snap 14 4 05480 29 1638 0059 dd20171219 2372a-B HBB1 14 4 03158 19

dd20171219 2372a-B HBB1 98 15 Slow 14 4 04798 16 1519 0038 dd20171219 2372a-B HBB1 98 15 Snap 14 4 04856 17 1538 0039

Time at Denaturation Temperature Figure 16 summarizes λtλu for the SRM 2372-B material as a function of denaturation temperature and the cumulative time the solution is held at that temperature Table 20 details the experimental conditions

Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures

The symbols represent the ratio of the mean result of three technical replicates of the untreated and each of the treated solutions the bars represent combined standard uncertainties The lines denote regression fits to the function λtλu = atimese-bt where t is the cumulative time Values and the standard uncertainties of the coefficients are listed in the legend

12

14

16

18

20

0 5 10 15 20

λ t λ

u

Cummulative Treatment Time min

degC a u (a ) b u (b )90 1804 0039 00198 0000494 1788 0029 00099 0002998 1855 0019 00067 00019

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

34

For all three of the denaturation temperatures studied the relationship between λtλu and cumulative time t can be modeled as

λtλu = ae-bt

where a is the extrapolated value of λtλu at t = 0 and b is the rate of change in AAT per minute As expected the rate of loss due to the treatment increases with increasing treatment temperature and cumulative time at that temperature37 Given the much longer treatment time for the 98 degC series the agreement among the three t = 0 parameter (a) values is remarkable This suggests that the first 1 min of treatment efficiently separates dsDNA into ssDNA The pooled relative standard uncertainty of these regression parameter is 17 somewhat smaller than the 25 pooled relative standard uncertainties of the individual λtλu However the plusmnstandard uncertainty interval of the 1-min λtλu results for the 94 degC and 90 degC both overlap the extrapolated t = 0 value suggesting that reliable results can be obtained using a single treatment of short duration

Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171228 2372-B HBB1 14 3 03650 17 dd20171228 2372-B HBB1 98 5 R1 14 3 06157 12 1687 0035 dd20171228 2372-B HBB1 98 10 R1 14 3 05493 08 1505 0028 dd20171228 2372-B HBB1 98 15 R1 14 3 05019 13 1375 0029 dd20171228 2372-B HBB1 98 20 R1 14 3 04583 21 1256 0034 dd20171228 2372-B HBB1 98 25 R1 14 3 04154 32 1138 0041 dd20171228 2372-B HBB1 98 30 R1 14 3 03784 14 1037 0023 dd20171228 2372-B HBB1 98 35 R1 14 3 03408 10 0934 0019 dd20171228 2372-B HBB1 98 40 R1 14 3 02980 01 0816 0014 dd20171228 2372-B HBB1 98 40 R1 14 3 02815 08 0771 0015

dd20171229 2372-B HBB1 14 3 03708 27

dd20171229 2372-B HBB1 94 1 R1 14 3 06598 26 1779 0056 dd20171229 2372-B HBB1 94 2 R1 14 3 06553 32 1767 0063 dd20171229 2372-B HBB1 94 3 R1 14 3 06301 20 1699 0045 dd20171229 2372-B HBB1 94 4 R1 14 3 06252 14 1686 0037 dd20171229 2372-B HBB1 94 5 R1 14 3 06577 14 1773 0039 dd20171229 2372-B HBB1 94 6 R1 14 3 06156 07 1660 0030 dd20171229 2372-B HBB1 94 7 R1 14 3 06118 10 1650 0033 dd20171229 2372-B HBB1 94 8 R1 14 3 06256 12 1687 0035 dd20171229 2372-B HBB1 94 9 R1 14 3 05999 10 1618 0032

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

35

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180104 2372-B HBB1 14 4 03713 25 dd20180104 2372-B HBB1 90 1 R1 14 3 06916 22 1863 0052 dd20180104 2372-B HBB1 90 3 R1 14 3 06465 06 1741 0032 dd20180104 2372-B HBB1 90 5 R1 14 3 06431 20 1732 0045 dd20180104 2372-B HBB1 90 10 R1 14 3 06014 25 1620 0049 dd20180104 2372-B HBB1 90 15 R1 14 3 06024 10 1623 0032 dd20180104 2372-B HBB1 90 20 R1 14 3 06003 14 1617 0035

Temperature Figure 17 summarizes the effect of denaturation temperature using a single treatment of 05 min on the 2372-B and 2372a-B materials Table 21 details the experimental conditions

Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

The symbols represent the ratio of the mean result of four technical replicates of the untreated and three replicates of the treated solutions the bars represent combined standard uncertainties In all cases the duration of the treatment was 05 min The lines denote regression fits to the linear function λtλu = a +bT where Tis the denaturation temperature

The SRM 2372 and 2372a materials respond very differently to changes in denaturation temperature The older spectrophotometrically ssDNA material suffers somewhat greater loss in λtλu as temperature increases The new spectrophotometrically dsDNA material suffers much greater loss in λtλu as temperature decreases

14

15

16

17

18

19

20

82 86 90 94 98

λ tλ

u

Temperature degC

2372-B

2372a-B

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

36

Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180109 2372-B HBB1 14 4 03577 25 dd20180109 2372-B HBB1 100 05 R5 14 3 06314 16 1765 0051

dd20180104a 2372-B HBB1 14 4 03577 05

dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032 dd20180104a 2372-B HBB1 94 05 R5 14 3 06459 25 1806 0046 dd20180104a 2372-B HBB1 90 05 R5 14 2 06591 08 1842 0017 dd20180104a 2372-B HBB1 88 05 R5 14 3 06589 05 1842 0013 dd20180104a 2372-B HBB1 84 05 R5 14 3 06729 10 1881 0022 dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

We hypothesize that in the older SRM 2372 material the between-strand bond is relatively fragile and can be completely severed at relatively low temperature while avoiding much temperature-dependent AAT loss In contrast the between-strand bond in the SRM 2372a material requires higher temperature to completely disassociate the strands This suggests that there is not a sample-independent ldquooptimumrdquo combination of denaturation temperature and treatment duration However 98 degC for 05 min appears to provide comparable results for these two materials DNA Concentration in the ddPCR Solution Figure 18 summarizes the effect of [DNA] in the ddPCR solution with SRM 2372-B at a denaturation temperature of 98 degC and treatment duration of 05 min Table 22 details the experimental conditions

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

37

Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution

The solid black symbols represent the ratio of the mean result of four technical replicates of the untreated and treated solutions with the λtλu values shown on the left-hand vertical axis the bars represent combined standard uncertainties The solid blue circles represent λu and the open red circles λt with values shown on the right-hand vertical axis at this graphical scale the standard deviations of the technical replicates are covered by the symbol The diagonal blue and red lines represent the log-linear relationships between just the 18 and 14 λ values

The primary effect of [DNA] appears to be the loss of linearity in λt with copiesdroplet values greater than about 1 copiesdroplet There may be some advantage to use somewhat greater dilutions than the routine 14 however the small increase in λtλu may be offset by the increased variability with small λu This suggests the use of dilutions that give λu values of (01 to 04) copiesdroplet

Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180117 2372-B HBB1 12 4 07127 11 dd20180117 2372-B HBB1 98 05 R5 12 4 12331 14 1730 0031

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 3 06151 15 1843 0040

dd20180117 2372-B HBB1 18 4 01549 18 dd20180117 2372-B HBB1 98 05 R5 18 4 02897 20 1870 0051

12 14 18

170

175

180

185

190

02

04

06

08

101214

λ T

arge

t D

ropl

et

λ tλ

u

Dilution Factor

λ₊λᵤλᵤλ₊

λtλu

λu

λt

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

38

Equivalence of Assays Figure 19 summarizes the agreement between the HBB1 NEIF and ND6 assays at a denaturation temperature of 98 degC and treatment duration of 05 min Table 23 details the experimental conditions

Figure 19 Comparison of Assays

The symbols compared the λtλu results from the HBB1 and either the NEIF or ND6 assays Each λtλu value is the ratio of the mean result of four technical replicates of an untreated and three or four replicates of the treated solutions The bars represent combined standard uncertainties The diagonal line represents equality between the assays

There is little or no difference between the λtλu results from the HBB1 NEIF and ND6 assays suggesting that the chromosomal location of the assay target has little influence Since the NEIFHBB1 comparison includes results for both the SRM 2372 and 2372a materials this equivalence is not material-specific

14

16

18

20

14 16 18 20

λ tλ

u NE

IF o

r ND6

λtλu HBB1

NEIFHBB1

ND6HBB1

2nd Assay = HBB1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

39

Table 23 Comparison of Assays

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu)

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045 dd20180105 2372a-B ND6 94 05 R5 14 3 05526 36 1711 0074 dd20180105 2372a-B ND6 90 05 R5 14 3 05326 18 1649 0051 dd20180105 2372a-B ND6 88 05 R5 14 3 05077 22 1572 0052 dd20180105 2372a-B ND6 86 05 R5 14 3 05038 23 1560 0053 dd20180105 2372a-B ND6 84 05 R5 14 3 04906 10 1519 0041

dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

40

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials

Figure 20 summarizes the currently available λtλu results obtained by denaturing at 98 degC for 05 min rapidly cooling in the thermal cycler with dilutions that provide asymp 03 copiesdroplet The figure also displays the consensus result and its 95 confidence region estimated using the Hierarchical Bayes method provided in the NIST Consensus Builder (NICOB) system3839 Table 24 details the experimental conditions

Figure 20 Evaluation of Limiting Ratio

The symbols represent the ratio of the mean result of three or four technical replicates of the untreated and treated solutions the bars represent combined standard uncertainties The thick horizontal line represents a consensus value covering all displayed results the dotted horizontal lines bracket the 95 confidence interval about the consensus value

The Hierarchical Bayes consensus result for these values is 183 with a 95 confidence interval of plusmn003 All but three of the 16 λtλu values are within one standard uncertainty of this consensus value with most of the λtλu means lying within the 95 confidence interval This suggests that there are no statistically significant differences between the SRM 2372 and 2372a materials with regard to ssDNA proportion

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

17

18

19

20

λ tλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

41

Table 24 Evaluation of Limiting Ratio

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180104a 2372-B HBB1 14 4 03577 05 dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032

dd20180108 2372-B HBB1 14 3 03463 09 dd20180108 2372-B HBB1 98 05 R5 14 3 06562 14 1895 0031

dd20180109 2372-B HBB1 14 3 03577 25 dd20180109 2372-B HBB1 98 05 R5 14 3 06495 21 1816 0059

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 4 06151 15 1843 0040

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

42

The agreement among the measured λtλu is more directly visualized using estimated unilateral ldquoDegrees of Equivalencerdquo d Each d is the absolute difference between the measured and consensus values

119889119889 = |120582120582119905119905 120582120582119906119906frasl minus consensus|

and is associated with a 95 confidence interval U95(d) that combines the uncertainty estimates of the measured and consensus values The U95(d) calculation details are specific to the consensus estimation method39 Values of d plusmn U95(d) that include 0 indicate that the λtλu result does not differ from the majority at about a 95 level of confidence Figure 21 displays the d for the λtλu results for the Hierarchical Bayes consensus value39 By this figure of merit all the λtλu can be considered as members of a single population

Figure 21 Degrees of Equivalence

The symbols represent the Degree of Equivalence d for the λtλu results relative to the Hierarchical Bayes consensus value The bars represent 95 level of confidence uncertainty intervals d plusmn U95(d) The horizontal line marks the d = 0 ldquono differencerdquo value Results with d plusmn U95(d) that include this line are not significantly different from the majority

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

-024

-018

-012

-006

000

006

012

018

024

Degr

ees o

f Equ

ival

ence

λtλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

43

Conclusions and Recommendations With high confidence all tubes of the SRM 2372a solution have the same nDNA copy number content within measurement repeatability Additionally within measurement repeatability all tubes of the SRM 2372a have the same mtDNAnDNA With high confidence the nDNA copy number content of the SRM 2372a solution is thermally stable from 4 degC to 37 degC over an extended period The solution should not be shipped or stored below 4 degC With high confidence the ddPCR measurements used to establish the entity and mass concentrations and the mtDNAnDNA ratios are adequately precise and unbiased With high confidence the DNA in the SRM 2372a solutions is now dsDNA and can be expected to remain dsDNA for at least ten years when stored at 4 degC Recommended Certification Values

Since the component materials are believed to be homogeneous are of very similar composition and were prepared and processed similarly the observed differences in the measured entity concentration uncertainties are likely artifacts of the measurement process To limit over-interpretation of these differences we recommend that the copy per nanoliter values of the three components be assigned same 95 relative uncertainty interval (CV95) based on the largest of the measured CV95 values (93 ) rounded up to 10 Since the ngμL transformation does not appreciably inflate this relative uncertainty we recommend that the mass concentration values also be assigned CV95 of 10 Similarly we recommend that the mtDNAnDNA ratios be rounded to integer values all three components be assigned the same CV95 and that this uncertainty be based on the largest (21 ) of the measured CV95 rounded up to 25 Table 25 lists the values that we recommend be used in the Certificate of Analysis

Table 25 Recommended Values and 95 Uncertainties for Certification

Component Units Value U95(Value) a A Copies per nanoliter 151 15 B Copies per nanoliter 175 18 C Copies per nanoliter 145 15 A ngμL 498 50

B ngμL 578 58 C ngμL 479 48 A mtDNAnDNA 174 4

B mtDNAnDNA 206 5 C mtDNAnDNA 279 7

a The interval Value plusmn U95(Value) is believed with about 95 confidence to contain the true value of the measurand

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

44

Use of SRM 2372a for Value-Assigning In-House Reference Materials

The within-component and between-assay variability observed for the commercial qPCR assays is not unexpected since different qPCR methods exploit qualitatively different molecular targets Given the intrinsic variability of the human genome the prospect of developing a human DNA source material that responds equally to all current and future qPCR assays is improbable All three components should be used to elucidate potential bias when using these materials to value assign in-house DNA reference and control materials for qPCR assays

Certificate of Analysis In accordance with ISO Guide 31 2000 a NIST SRM certificate is a document containing the name description and intended purpose of the material the logo of the US Department of Commerce the name of NIST as a certifying body instructions for proper use and storage of the material certified property value(s) with associated uncertainty(ies) method(s) used to obtain property values the period of validity if appropriate and any other technical information deemed necessary for its proper use A Certificate is issued for an SRM certified for one or more specific physical or engineering performance properties and may contain NIST reference information or both values in addition to certified values A Certificate of Analysis is issued for an SRM certified for one or more specific chemical properties Note ISO Guide 31 is updated periodically check with ISO for the latest version [httpswwwnistgovsrmsrm-definitions] For the most current version of the Certificate of Analysis for NIST SRM 2372a Human DNA Quantification Standard please visit httpswww-snistgovsrmorsview_detailcfmsrm=2372a

References

1 National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research Department of Health Education and Welfare (1978) Belmont Report Ethical Principles and Guidelines for the Protection of Human Subjects of Research DHEW Publication No (OS) 78-0012 httpvideocastnihgovpdfohrp_belmont_reportpdf

2 CDCNIH Biosafety in Microbiological and Biomedical Laboratories 5th ed HHS publication No (CDC) 21-1112 Chosewood LC Wilson DE Eds US Government Printing Office Washington DC (2009) httpwwwcdcgovbiosafetypublicationsbmbl5indexhtm

3 Duewer DL Kline MC Redman JW Newall PJ Reeder DJ NIST Mixed Stain Studies 1 and 2 Interlaboratory Comparison of DNA Quantification Practice and Short Tandem Repeat Multiplex Performance with Multiple-Source Samples J Forensic Sci 200146(5)1199-1210 PMID 11569565

4 Kline MC Duewer DL Redman JW Butler JM NIST Mixed Stain Study 3 DNA quantitation accuracy and its influence on short tandem repeat multiplex signal intensity Anal Chem 200375(10)2463-2469 httpsdoiorg101021ac026410i

5 Kline MC Duewer DL Redman JW Butler JM Results from the NIST 2004 DNA Quantitation Study J Forensic Sci 200550(3)571-578 PMID 15932088

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

45

6 Kline MC Duewer DL Travis JC Smith MV Redman JW Vallone PM Decker AE Butler JM Production and Certification of Standard Reference Material 2372 Human DNA Quantitation Standard Anal Bioanal Chem 2009394(4)1183-1192 httpsdoiorg101007s00216-009-2782-0

7 Sambrook J and Russell DW (2001) Molecular Cloning a Laboratory Manual Cold Spring Harbor Laboratory Press Cold Spring Harbor New York

8 EU-JRC Protocol NK603 ndash Community Reference Laboratory Event-specific method for the quantification of maize line NK603 using real-time PCR Section 33 Spectrophotometric Measurement of DNA Concentration httpgmo-crljrceceuropaeusummariesNK603-WEB-ProtocolValidationpdf

9 ISO 215712005(E) Foodstuffs ndash Methods of analysis for the detection of genetically modified organisms and derived products ndash Nucleic acid extraction Annex B Methods for quantitation of extracted DNA pp 34-36 International Organization for Standardization Geneva (2005)

10 Certificate of Analysis Standard Reference Material 2372 Human DNA Quantitation Standard (2013) Gaithersburg MD USA Standard Reference Materials Program National Institute of Standards and Technology httpswww-snistgovsrmorsview_certcfmsrm=2372

11 Vogelstein B Kinzler KW Digital PCR Proc Natl Acad Sci USA 1999969236-9241 12 Hindson BJ Ness KD Masquelier DA Belgrader P Heredia NJ Makarewicz AJ Bright IJ Lucero

MY Hiddessen AL Legler TC Kitano TK Hodel MR Petersen JF Wyatt PW Steenblock ER Shah PH Bousse LJ Troup CB Mellen JC Wittmann DK Erndt NG Cauley TH Koehler RT So AP Dube S Rose KA Montesclaros L Wang S Stumbo DP Hodges SP Romine S Milanovich FP White HE Regan JF Karlin-Neumann GA Hindson CM Saxonov S Colston BW High-Throughput Droplet Digital PCR System for Absolute Quantitation of DNA Copy Number Anal Chem 2011838604-8610 httpsdoiorg101021ac202028g

13 Pinheiro LB Coleman VA Hindson CM Herrmann J Hindson BJ Bhat S Emslie KR Evaluation of a Droplet Digital Polymerase Chain Reaction Format for DNA Copy Number Quantification AnalChem 2012841003-1011 httpsdoiorg101021ac202578x

14 Corbisier P Pinheiro L Mazoua S Kortekaas AM Chung PY Gerganova T Roebben G Emons H Emslie K DNA copy number concentration measured by digital and droplet digital quantitative PCR using certified reference materials Anal Bioanal Chem 2015407(7)1831-1840 httpsdoiorg101007s00216-015-8458-z

15 Duewer DL Kline MC Romsos EL Real-time cdPCR opens a window into events occurring in the first few PCR amplification cycles Anal Bioanal Chem 2015407(30)9061-9069 httpsdoiorg101007s00216-015-9073-8

16 Kline MC Romsos EL Duewer DL Evaluating Digital PCR for the Quantification of Human Genomic DNA Accessible Amplifiable Targets Anal Chem 201688(4)2132-2139 httpsdoiorg101021acsanalchem5b03692

17 Kline MC Duewer DL Evaluating Droplet Digital Polymerase Chain Reaction for the Quantification of Human Genomic DNA Lifting the Traceability Fog Anal Chem 201789(8)4648-4654 httpsdoiorg101021acsanalchem7b00240

18 Duewer DL Kline MC Romsos EL Toman B Evaluating Droplet Digital PCR for the quantification of human genomic DNA Conversion from Copies per Nanoliter to Nanogram Genomic DNA per Microliter Anal Bioanal Chem 2018in press httpsdoiorg101007s00216-018-0982-1

19 Dagata JA Farkas N Kramer JA Method for measuring the volume of nominally 100 μm diameter spherical water in oil emulsion droplets NIST Special Publication 260-184 2016 httpsdoiorg106028NISTSP260-184

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

20 ISO 170342016 General requirements for the competence of reference material producers International Organization for Standardization Geneva (2016)

21 NCBI Standard Nucleotide BLAST httpsblastncbinlmnihgovBlastcgiPAGE_TYPE=BlastSearch

22 Timken MD Swango KL Orrego C Buoncristiani MR A Duplex Real-Time qPCR Assay for the Quantification of Human Nuclear and Mitochondrial DNA in Forensic Samples Implications for Quantifying DNA in Degraded Samples J Forensic Sci 200550(5)1044-60 PMID 16225209

23 Walker JA Hedges DJ Perodeau BP Landry KE Stoilova N Laborde ME Shewale J Sinha SK Batzer MA Multiplex polymerase chain reaction for simultaneous quantitation of human nuclear mitochondrial and male Y-chromosome DNA application in human identification Anal Biochem 200533789ndash97 httpsdoiorg101016jab200409036

24 Kavlick MF Lawrence HS Merritt RT Fisher C Isenberg A Robertson JM Budowle B Quantification of Human Mitochondrial DNA Using Synthesized DNA Standards J Forensic Sci 201156(6) 1457ndash1463 httpsdoiorg101111j1556-4029201101871x

25 Stulnig TM Amberger A Exposing contaminating phenol in nucleic acid preparations BioTechniques 199416(3) 403-404

26 Majumdar N Wessel T Marks J Digital PCR Modeling for Maximal Sensitivity Dynamic Range and Measurement Precision PLoS ONE 201510(3)e0118833 httpsdoi101371journalpone0118833

27 Tellinghuisen J Partition Volume Variability in Digital Polymerase Chain Reaction Methods Polydispersity Causes Bias but Can Improve Precision Anal Chem 2016881283-12187 httpsdoiorg101021acsanalchem6b03139

28 Lunn DJ Spiegelhalter D Thomas A Best N (2009) Statistics in Medicine 283049ndash3082 29 De Biegravevre P Dybkaer R Fajgelj A Hibbert DB Metrological traceability of measurement results in

chemistry Concepts and implementation Pure Appl Chem 201183(10)1873-1935 httpsdoiorg101007s00769-011-0805-y

30 Huggett JF Foy CA Benes V Emslie K Garson JA Haynes R Hellemans J Kubista M Mueller RD Nolan T Pfaffl MW Shipley GL Vandesompele J Wittwer CT Bustin SA The Digital MIQE Guidelines Minimum Information for Publication of Quantitative Digital PCR Experiments Clin Chem 201359(6)892-902 httpsdoiorg101373clinchem2013206375

31 Wilson PJ Ellison SLR Extending digital PCR analysis by modelling quantification cycle data BMC Bioinfo 201617421 httpsdoiorg101186s12859-016-1275-3

32 Wang X Son A Effects of pretreatment on the denaturation and fragmentation of genomic DNA for DNA hybridization Environ Sci Processes Impacts 2013152204 httpsdoiorg101039c3em00457k

33 Nwokeoji AO Kilby PM Portwood DE Dickman MJ Accurate Quantification of Nucleic Acids Using Hypochromicity Measurements in Conjunction with UV Spectrophotometry Anal Chem 201789(24)13567-13574 https101021acsanalchem7b04000

34 Farkas DH Kaul KL Wiedbrauk DL Kiechle FL Specimen collection and storage for diagnostic molecular pathology investigation Arch Pathol Lab Med 1996120(6)591-6

35 Kypr J Kejnovskaacute I Renčiuk D Vorličkovaacute M Circular dichroism and conformational polymorphism of DNA Nucleic Acids Res 200937(6)1713-1725 httpsdoiorg101093nargkp026

36 Strachan T Read AP (1999) Human Molecular Genetics 2nd Ed John Wiley amp Sons NY

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

37 Bhat S McLaughlin JL Emslie KR Effect of sustained elevated temperature prior to amplification on template copy number estimation using digital polymerase chain reaction Analyst 2011136724ndash32 httpsdoiorg101039c0an00484g

38 NIST Consensus Builder (NICOB) httpsconsensusnistgov 39 Koepke A Lafarge T Toman B Possolo A NIST Consensus Builder Userrsquos Manual

httpsconsensusnistgovNISTConsensusBuilder-UserManualpdf

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

Appendix A DNA Extraction Method The following method has been adapted from Miller et al (1988)A1 Procedure

1 Transfer approximately 5 mL of buffy coat into a 50 mL sterile polypropylene centrifuge tube and add 30 mL Red Blood Cell Lysis Buffer

2 Mix by gentle inversion and let stand at room temperature for 15 min

3 Centrifuge at approximately 209 rads (2000 rpm) for 20 min

4 Discard supernatant

5 Add 9 mL of Nucleic Lysis Buffer and re-suspend the pelleted cells by agitating with a plastic transfer pipette

6 Add 300 microL of 20 sodium dodecyl sulfate (SDS) and 18 mL of Protease K Solution (freshly prepared 2 mg to 5 mg Protease K per 1 mL Protease K Buffer)

7 Mix by inversion followed by incubation overnight or longer at 37 ordmC to 40 ordmC in a shaking water bath The length of incubation is determined by the time required to solubilize the cells present in the tube

8 Add 4 mL saturated ammonium acetate solution (in deionized water stored at 2 ordmC to 8 ordmC) Shake vigorously for 15 s

9 Centrifuge at approximately 471 rads (4500 rpm) for 10 min

10 Pipette supernatant into a 50 mL sterile polypropylene centrifuge tube and add two volumes of room temperature absolute ethanol Mix by gentle inversion until the DNA aggregates and floats to the top of the tube

11 Spool the aggregated DNA from the current tube into a fresh tube of 80 ethanol using a sterile inoculation loop At this step multiple tubes of precipitated DNA may be pooled to make a large lot of the purified material Mix by gentle inversion

12 Transfer the aggregated DNA to a fresh dry 50 mL polypropylene centrifuge tube Remove as much ethanol as possible by squeezing the DNA against the wall of the tube with a sterile inoculation loop Air-dry the material in a laminar flow hood

13 Resolubilize DNA in TE-4 buffer and store material at 4 ordmC Reagents

Red Blood Cell Lysis Buffer 144 mmolL ammonium chloride 1 mmolL sodium bicarbonate in deionized (DI) water Store at 2 ordmC to 8 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

Nucleic Lysis Buffer 10 mmolL Tris-HCl 400 mmolL sodium chloride 20 mmolL disodium ethylenediaminetetraacetic acid in DI Water adjust pH to 82 Store at 2 ordmC to 8 ordmC

Protease K Buffer

20 mmolL disodium ethylenediaminetetraacetic acid 10 (mass per volume) Sodium Dodecyl Sulfate (SDS) in DI water Store at 2 ordmC to 8 ordmC

Protease K

Sigma-Aldrich cat P6556 Tritirachium album lyophilized powder Store at ndash20 ordmC

Ammonium Acetate

192 molL ammonium acetate (saturated) in DI water Store at 2 ordmC to 8 ordmC

Absolute Ethanol (200 proof)

Store at room temperature Reference

A1 Miller SA Dykes DD Polesky HF A simple salting out procedure for extracting DNA from human nucleated cells Nucleic Acids Res 198816(3)1215 httpsdoiorg101093nar1631215

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

48

Appendix B Spectrophotometric Measurement Protocol The following is for use with a scanning recording spectrophotometer The procedures include the spectrophotometer calibration measurement of the ldquonativerdquo dsDNA and measurement of the NaOH denatured ssDNA The procedure for evaluating solutions after treatment with NaOH is adapted from references89 Spectrophotometer Calibration

Establish Baseline Set the spectrophotometer to acquire data from 220 nm to 345 nm in steps of 01 nm Zero the instrument with nothing (air) in the sample and reference cells Acquire an air-air baseline of the instrument with nothing (air) in the sample and reference cells If the trace has significant structure or excessive noise take corrective action If the trace is acceptable proceed with data collection Enable Verification of Wavelength Calibration SRM 2034 is a solution of holmium oxide in 10 perchloric acid contained in a sealed quartz cuvette delivering an intrinsic UVVis wavelength standard for absorption spectrophotometryB1 Evaluation of the location of peak maxima in the spectra of this solution enables verification of the spectrophotometers wavelength calibration Place SRM 2034 in the sample cell Create and name a file to hold the verification spectra Perform three replicate SRM 2034-to-air scans saving each scan to the file Remove SRM 2034 from the spectrophotometer and store appropriately Enable Verification of Absorbance Calibration SRM 2031 consists of three metal-on-fused silica filters held in cuvette-sized filter holders plus an empty filter holder The three filters are certified for transmittance and transmittance density (-log10(transmittance) effectively absorbance) at several wavelengths with nominal wavelength-independent transmittances of 10 30 and 90 Evaluation of the absorbance of each of these three filters at the certified 250 nm 280 nm and 340 nm wavelengths enables verification of the spectrophotometers absorbance calibration Place the empty SRM 2031 filter holder in the reference cell Remove the protective covers from the 10 transmittance filter and place in the sample cell Perform three replicate scans saving each scan to file Remove the filter from the spectrophotometer replace the protective covers and store appropriately Repeat for the 30 and 90 filters Close the file containing the verification spectra For ldquonativerdquo dsDNA Samples

Verify Optical Balance of Cuvettes The measurements of the DNA-containing samples are made using a 100 microL quartz cuvette in the sample cell compared to an optically matched cuvette in the reference cell The optical equivalence of these two cuvettes can be verified by filling both cuvettes with appropriate volumes of a sample solution matrix (TE-4 buffer) that does not contain DNA Fill two dry

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

49

clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes followed by a methanol rinse in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquonativerdquo dsDNA remove and clean both the reference cuvette and the sample cuvette For NaOH denatured ssDNA Samples

Prepare a 04 molL NaOH Stock Solution To a 15 mL disposable centrifuge tube add 016 g NaOH pellets add deionized (DI) water to the 10 mL line and mix until all the NaOH pellets have dissolved Centrifuge at 524 rads (5000 rpm) for 5 min For every 9 samples to be analyzed aliquot 1 mL of this 04 molL NaOH stock solution into a 15 mL screw top tube Use these tubes to prepare the NaOH blanks and DNA samples In a 15 mL screw cap vial mix 200 microL TE-4 buffer with 200 microL 04 molL sodium hydroxide solution to achieve a final NaOH concentration of 02 molL Vortex 5 s then centrifuge at 1257 rads (12 000 rpm) for 1 min Fill two dry clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Prepare the Sample Solutions For each unit of material to be measured (Tube or vial)

a) Vortex the unit for 5 s b) Centrifuge at 1257 rads (12 000 rpm) for 1 min c) Remove a 100 microL aliquot and place in a new labeled vial d) Add 100 microL of the freshly prepared 04 molL NaOH solution (with the same pipet) e) Vortex 5 s f) Centrifuge at 1257 rads (12 000 rpm) for 1 min

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

50

Measure Sample Solutions Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes rinse with methanol in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the NaOH denatured DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquodenaturedrdquo ssDNA close the file holding the spectra for this series Remove and clean both the reference cuvette and the sample cuvette After All Sample Measurements are Complete

Enable Verification of Spectrophotometer Stability The stability of the spectrophotometer over the course of the data collection can be evaluated by comparing spectra for SRM 2034 and 2031 taken at the start of data collection with spectra taken at the end of data collection Repeat the wavelength and absorbance verification measurements described above collecting the spectra in a suitably named file Clean Up Transfer all relevant data files from the spectrophotometers storage to a transfer device Clean the spectrophotometer work area as necessary and restore the SRMs and cuvettes to their appropriate storage locations Evaluation

Export the files to a spreadsheet Average the replicate spectra Subtract the average absorbance at A320 from the absorbance at A260 resulting in the corrected absorbance at A260 The conventional spectrophotometric estimate of the mass concentration of the dsDNA in the sample is achieved by multiplying the corrected absorbance by 50 The conventional estimate for the NaOH-treated ssDNA is achieved by multiplying the corrected absorbance by 37times2 where the ldquotimes2rdquo adjusts for the 1rarr2 volumetric dilution of the sample Note reliable spectrophotometric measurements require corrected A260 to be greater than 005 Reference

B1 Travis JC Acosta JC Andor G Bastie J Blattner P Chunnilall CJ Crosson SC Duewer DL Early EA Hengstberger F Kim C-S Liedquist L Manoocheri F Mercader F Metzdorf J Mito A Monardh LAG Nevask S Nilsson M Noeumll M Rodriguez AC Ruiacutez A Schirmacher A Smith MV Valencia G van Tonder N Zwinkels J Intrinsic Wavelength Standard Absorption Bands in Holmium Oxide Solution for UVvisible Molecular Absorption Spectrophotometry J Phys Chem Ref Data 200534(1)41-56

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

51

Appendix C SRM 2372a Statisticianrsquos report Extracted from Blaza Toman

November 30 2017 1 Introduction Candidate material for SRM 2372a is composed of three components labeled A B and C All components are human genomic DNA and prepared at NIST from buffy coat white blood cells from single-source anonymous donors Component A and B are single source (malefemale respectively) and Component C is a 1 male to 3 female mixture of two single-source donors Droplet digital PCR (ddPCR) assays were used to obtain measurements of the concentration of copies per microliter of sample Ten NIST-developed human nuclear DNA and three literature-derived human mitochondrial DNA PCR assays were used in this study

11 OpenBUGS Analysis All parameters in the following sections were estimated using Markov Chain Monte Carlo (MCMC) implemented in the OpenBUGS freewareC1 OpenBUGS is a software application for the Bayesian analysis of complex statistical models

12 OpenBUGS Exemplar Output All following sections list the OpenBUGs model (ldquocoderdquo) initialization values (ldquoinitsrdquo) if they are needed and the data used to evaluate the various data sets discussed in the body of this report homogeneity stability nDNA certification and mtDNAnDNA value assignment The provided information in each of the following sections is sufficient to produce estimates but the estimates may not exactly reproduce the exemplar results provided The MCMC paradigm does not produce ldquoanalyticalrdquo values but approaches them asymptotically as the number o model updates increases All exemplar results presented here are based upon the following bull 10-fold ldquothinningrdquo to minimize autocorrelation there were 10 MCMC estimates for every

model update used in parameter estimation bull a ldquoburn inrdquo of 5 000 updates before beginning parameter estimation and bull 10 000 model updates

The following statistics are provided in the tables listing the exemplar results

mean arithmetic mean of the parameterrsquos model update values sd standard deviation of the parameterrsquos model update values MC_error Monte Carlo standard deviation of the mean meanradic(number of model updates) val25pc 25th percentile of the model update values median 50th percentile of the model update values val975pc 975th percentile of the model update values startnumber of ldquoburn inrdquo updates sample number of model update values

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

52

2 Homogeneity assessment 21 Data

Samples of the material of all three components were drawn from 22 different boxes Two assays (NEIF ND6) were used to obtain 3 replicates of the measurement in copies per nanoliter for each sample and component For the complete homogeneity data see

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

22 Model The data were analyzed using a three-level Gaussian hierarchical ANOVA model

119910119910119894119894119894119894119894119894~119873119873120572120572119894119894119894119894119894119894 120590120590119908119908119894119894119905119905ℎ119894119894119894119894 119887119887119887119887119887119887 1198941198941198941198942

where ~N() denotes ldquodistributed as a normal distribution of some mean μ and variancerdquo 119894119894 denotes component (1 for A 2 for B 3 for C) 119895119895 denotes assay (1 for NEIF 2 for ND6) 119896119896 denotes box (1hellip22) and 120572120572119894119894119894119894119894119894~119873119873micro119894119894119894119894120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119887119887119887119887119887119887119894119894

2

All variance components were given a Half Cauchy prior as described in Reference C2 Note The Half Cauchy prior is specified by ldquo~dnorm(0 00016)I(0001)rdquo The ldquoIrdquo notation is for censoring It cuts off any random draws from the Normal distribution that are lower than 0001 An informative prior is needed when there are less than four data per distribution and the half Cauchy is a good general-purpose prior

23 OpenBUGS code and data Homogeneity code ModelBegin for(i in 13)xiNsb[i] ~ dnorm(0 00016)I(0001) chSqNsb[i] ~ dgamma(0505) sigb[i] lt- xiNsb[i]sqrt(chSqNsb[i]) for(i in 13)for(j in 12)mu[ij]~dnorm(010E-5) for(i in 13) muA[i]lt-mean(mu[i])sigbP[i]lt-100(muA[i]sqrt(sigb[i])) for(i in 13)for(j in 12)xiNs[ij] ~ dnorm(0 00016)I(0001) chSqNs[ij]~dgamma(0505) sigt[ij]lt-xiNs[ij]sqrt(chSqNs[ij]) sigtP[ij]lt-100(mu[ij]sqrt(sigt[ij])) for(i in 13)for(j in 12)for(k in 122)alpha[ijk]~dnorm(mu[ij]sigb[i]) for(idx in 1393)lamb[idx]~dnorm(alpha[cmp[idx]asy[idx]box[idx]]sigt[cmp[idx]asy[idx]]) ModelEnd Inits None required

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

53

Homogeneity Data lamb[] asy[] box[] cmp[] rep[] 1632 1 1 1 1 1529 1 1 1 2 1511 1 1 1 3 1616 2 1 1 1 1640 2 1 1 2 1582 2 1 1 3 1531 1 2 1 1 1525 1 2 1 2 1549 1 2 1 3 1615 2 2 1 1 1589 2 2 1 2 1601 2 2 1 3 1524 1 3 1 1 1540 1 3 1 2 1533 1 3 1 3 1582 2 3 1 1 1152 2 3 1 2 1555 2 3 1 3 1454 1 4 1 1 1458 1 4 1 2 1427 1 4 1 3 1627 2 4 1 1 1605 2 4 1 2 1607 2 4 1 3 1465 1 5 1 1 1540 1 5 1 2 1504 1 5 1 3 1607 2 5 1 1 1633 2 5 1 2 1586 2 5 1 3 1512 1 6 1 1 1491 1 6 1 2 1491 1 6 1 3 1511 2 6 1 1 1483 2 6 1 2 1524 2 6 1 3 1418 1 7 1 1 1480 1 7 1 2 1455 1 7 1 3 1479 2 7 1 1 1557 2 7 1 2 1656 2 7 1 3 1496 1 8 1 1 1560 1 8 1 2 1501 1 8 1 3 1514 2 8 1 1 1615 2 8 1 2 1540 2 8 1 3 1588 1 9 1 1 1581 1 9 1 2 1637 1 9 1 3 1474 2 9 1 1 1438 2 9 1 2 1454 2 9 1 3 1574 1 10 1 1 1513 1 10 1 2 1570 1 10 1 3 1594 2 10 1 1 1713 2 10 1 2 1669 2 10 1 3

1557 1 11 1 1 1508 1 11 1 2 1633 1 11 1 3 1571 2 11 1 1 1611 2 11 1 2 1631 2 11 1 3 1500 1 12 1 1 1476 1 12 1 2 1515 1 12 1 3 1626 2 12 1 1 1586 2 12 1 2 1608 2 12 1 3 1672 1 13 1 1 1634 1 13 1 2 1804 1 13 1 3 1490 2 13 1 1 1516 2 13 1 2 1497 2 13 1 3 1551 1 14 1 1 1568 1 14 1 2 1538 1 14 1 3 1617 2 14 1 1 1614 2 14 1 2 1639 2 14 1 3 1456 1 15 1 1 1448 1 15 1 2 1523 1 15 1 3 1525 2 15 1 1 1402 2 15 1 2 1436 2 15 1 3 1573 1 16 1 1 1520 1 16 1 2 1617 1 16 1 3 1634 2 16 1 1 1627 2 16 1 2 1575 2 16 1 3 1516 1 17 1 1 1516 1 17 1 2 1628 1 17 1 3 1613 2 17 1 1 1567 2 17 1 2 1567 2 17 1 3 1612 1 18 1 1 1535 1 18 1 2 1672 1 18 1 3 1529 2 18 1 1 1495 2 18 1 2 1528 2 18 1 3 1598 1 19 1 1 1514 1 19 1 2 1596 1 19 1 3 1571 2 19 1 1 1535 2 19 1 2 1511 2 19 1 3 1575 1 20 1 1 1568 1 20 1 2 1586 1 20 1 3 1652 2 20 1 1 1545 2 20 1 2 1627 2 20 1 3 1658 1 21 1 1

1611 1 21 1 2 1578 1 21 1 3 1647 2 21 1 1 1607 2 21 1 2 1612 2 21 1 3 1547 1 22 1 1 1528 1 22 1 2 1532 1 22 1 3 1537 2 22 1 1 1513 2 22 1 2 1536 2 22 1 3 1790 1 1 2 1 1909 1 1 2 2 1885 1 1 2 3 1870 2 1 2 2 1860 2 1 2 3 1712 1 2 2 1 1719 1 2 2 2 1704 1 2 2 3 1775 2 2 2 1 1718 2 2 2 2 1756 2 2 2 3 1693 1 3 2 1 1755 1 3 2 2 1749 1 3 2 3 1824 2 3 2 1 1483 2 3 2 2 1896 2 3 2 3 1754 1 4 2 1 1745 1 4 2 2 1739 1 4 2 3 1670 2 4 2 1 1632 2 4 2 2 1814 2 4 2 3 1757 1 5 2 1 1687 1 5 2 2 1735 1 5 2 3 1683 2 5 2 1 1759 2 5 2 2 1703 2 5 2 3 1697 1 6 2 1 1709 1 6 2 2 1714 1 6 2 3 1693 2 6 2 1 1723 2 6 2 2 1722 2 6 2 3 1722 1 7 2 1 1625 1 7 2 2 1703 1 7 2 3 1690 2 7 2 1 1684 2 7 2 2 1688 2 7 2 3 1884 1 8 2 1 1802 1 8 2 2 1772 1 8 2 3 1864 2 8 2 1 1887 2 8 2 2 1821 2 8 2 3 1713 1 9 2 1 1734 1 9 2 2 1727 1 9 2 3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

54

1709 2 9 2 1 1756 2 9 2 2 1758 2 9 2 3 1759 1 10 2 1 1809 1 10 2 2 1738 1 10 2 3 1764 2 10 2 1 1816 2 10 2 2 1787 2 10 2 3 1749 1 11 2 1 1782 1 11 2 2 1740 1 11 2 3 1741 2 11 2 1 1792 2 11 2 2 1784 2 11 2 3 1748 1 12 2 1 1712 1 12 2 2 1718 1 12 2 3 1797 2 12 2 1 1614 2 12 2 2 1790 2 12 2 3 1632 1 13 2 1 1668 1 13 2 2 1701 1 13 2 3 1626 2 13 2 1 1566 2 13 2 2 1681 2 13 2 3 1752 1 14 2 1 1535 1 14 2 2 1673 1 14 2 3 1724 2 14 2 1 1741 2 14 2 2 1685 2 14 2 3 1871 1 15 2 1 1835 1 15 2 2 1887 1 15 2 3 1770 2 15 2 1 1752 2 15 2 2 1798 2 15 2 3 1688 1 16 2 1 1725 1 16 2 2 1743 1 16 2 3 1744 2 16 2 1 1697 2 16 2 2 1710 2 16 2 3 1789 1 17 2 1 1769 1 17 2 2 1825 1 17 2 3 1716 2 17 2 1 1762 2 17 2 2 1638 2 17 2 3 1826 1 18 2 1 1780 1 18 2 2 1849 1 18 2 3 1805 2 18 2 1 1801 2 18 2 2 1931 2 18 2 3 1751 1 19 2 1 1759 1 19 2 2 1880 1 19 2 3 1827 2 19 2 1 1766 2 19 2 2

1786 2 19 2 3 1738 1 20 2 1 1678 1 20 2 2 1768 1 20 2 3 1752 2 20 2 1 1747 2 20 2 2 1768 2 20 2 3 1742 1 21 2 1 1836 1 21 2 2 1713 1 21 2 3 1746 2 21 2 1 1744 2 21 2 2 1764 2 21 2 3 1761 1 22 2 1 1743 1 22 2 2 1660 1 22 2 3 1666 2 22 2 1 1692 2 22 2 2 1819 2 22 2 3 1493 1 1 3 1 1385 1 1 3 2 1510 1 1 3 3 1565 2 1 3 1 1562 2 1 3 2 1466 2 1 3 3 1516 1 2 3 1 1513 1 2 3 2 1577 1 2 3 3 1490 2 2 3 1 1547 2 2 3 2 1570 2 2 3 3 1451 1 3 3 1 1528 1 3 3 2 1506 1 3 3 3 1485 2 3 3 1 1505 2 3 3 2 1471 2 3 3 3 1440 1 4 3 1 1486 1 4 3 2 1520 1 4 3 3 1529 2 4 3 1 1508 2 4 3 2 1506 2 4 3 3 1523 1 5 3 1 1472 1 5 3 2 1483 1 5 3 3 1513 2 5 3 1 1520 2 5 3 2 1491 2 5 3 3 1437 1 6 3 1 1493 1 6 3 2 1522 1 6 3 3 1492 2 6 3 1 1481 2 6 3 2 1461 2 6 3 3 1485 1 7 3 1 1477 1 7 3 2 1463 1 7 3 3 1475 2 7 3 1 1547 2 7 3 2 1485 2 7 3 3 1433 1 8 3 1

1405 1 8 3 2 1478 1 8 3 3 1497 2 8 3 1 1501 2 8 3 2 1495 2 8 3 3 1463 1 9 3 1 1499 1 9 3 2 1505 1 9 3 3 1503 2 9 3 1 1491 2 9 3 2 1516 2 9 3 3 1534 1 10 3 1 1458 1 10 3 2 1454 1 10 3 3 1448 2 10 3 1 1465 2 10 3 2 1474 2 10 3 3 1533 1 11 3 1 1514 1 11 3 2 1555 1 11 3 3 1488 2 11 3 1 1470 2 11 3 2 1483 2 11 3 3 1450 1 12 3 1 1525 1 12 3 2 1497 1 12 3 3 1496 2 12 3 1 1483 2 12 3 2 1519 2 12 3 3 1537 1 13 3 1 1527 1 13 3 2 1515 1 13 3 3 1499 2 13 3 1 1479 2 13 3 2 1475 2 13 3 3 1515 1 14 3 1 1490 1 14 3 2 1494 1 14 3 3 1469 2 14 3 1 1480 2 14 3 2 1466 2 14 3 3 1511 1 15 3 1 1568 1 15 3 2 1594 1 15 3 3 1515 2 15 3 1 1543 2 15 3 2 1565 2 15 3 3 1449 1 16 3 1 1452 1 16 3 2 1471 1 16 3 3 1543 2 16 3 1 1448 2 16 3 2 1491 2 16 3 3 1388 1 17 3 1 1421 1 17 3 2 1394 1 17 3 3 1452 2 17 3 1 1435 2 17 3 3 1475 1 18 3 1 1455 1 18 3 2 1445 1 18 3 3 1458 2 18 3 1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

55

1531 2 18 3 2 1459 2 18 3 3 1468 1 19 3 1 1525 1 19 3 2 1446 1 19 3 3 1435 2 19 3 1 1478 2 19 3 2 1421 2 19 3 3 1549 1 20 3 1

1531 1 20 3 2 1499 1 20 3 3 1499 2 20 3 1 1465 2 20 3 2 1468 2 20 3 3 1530 1 21 3 1 1440 1 21 3 2 1484 1 21 3 3 1499 2 21 3 1

1454 2 21 3 2 1442 1 22 3 1 1487 1 22 3 2 1436 1 22 3 3 1589 2 22 3 1 1591 2 22 3 2 1556 2 22 3 3 END

Table C-1 Example OpenBUGS Homogeneity Summary

mean sd MC_error val25pc median val975pc start sample mu[11] 15460 0122 000123 1522 1546 1570 5001 10000 mu[12] 15640 0137 000141 1537 1564 1591 5001 10000 mu[21] 17490 0113 000118 1727 1749 1772 5001 10000 mu[22] 17480 0129 000125 1722 1748 1774 5001 10000 mu[31] 14870 0073 000073 1473 1487 1501 5001 10000 mu[32] 14960 0069 000063 1482 1496 1509 5001 10000 muA[1] 15550 0092 000092 1536 1555 1573 5001 10000 muA[2] 17490 0085 000086 1732 1749 1765 5001 10000 muA[3] 14910 0050 000046 1481 1491 1501 5001 10000 sigbP[1] 3269 04985 000514 2376 3244 4324 5001 10000 sigbP[2] 2548 04385 000461 1733 2533 3472 5001 10000 sigbP[3] 1850 0292 000260 1329 1832 2477 5001 10000

sigtP[11] 2774 0298 000260 2268 2750 3442 5001 10000 sigtP[12] 4036 0429 000438 3290 4005 4986 5001 10000 sigtP[21] 2664 0301 000321 2155 2640 3313 5001 10000 sigtP[22] 3807 0395 000377 3116 3772 4662 5001 10000 sigtP[31] 2313 0250 000209 1880 2291 2873 5001 10000 sigtP[32] 1812 0195 000170 1473 1797 2243 5001 10000

24 Results The material appears to be homogeneous as can be seen in Figure C-1 Based on these results the between-tube variability is of the same order as within-tube variability

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

56

Component A Assay NEIF Component A Assay ND6 box plot gamma[11]

[111][112][113]

[114]

[115][116]

[117]

[118]

[119]

[1110][1111]

[1112]

[1113]

[1114]

[1115]

[1116][1117]

[1118]

[1119][1120]

[1121]

[1122]

114

box plot gamma[12]

[121][122]

[123]

[124][125]

[126]

[127][128]

[129]

[1210]

[1211][1212]

[1213]

[1214]

[1215]

[1216]

[1217]

[1218][1219]

[1220][1221]

[1222]

115

Component B Assay NEIF Component B Assay ND6

box plot gamma[21]

[211]

[212]

[213][214]

[215][216]

[217]

[218]

[219]

[2110][2111]

[2112]

[2113][2114]

[2115]

[2116]

[2117]

[2118]

[2119]

[2120]

[2121]

[2122]

129

box plot gamma[22]

[221]

[222][223]

[224][225][226][227]

[228]

[229]

[2210][2211]

[2212]

[2213]

[2214]

[2215]

[2216][2217]

[2218]

[2219]

[2220][2221][2222]

128

Component C Assay NEIF Component C Assay ND6

box plot gamma[31]

[311]

[312]

[313][314]

[315][316]

[317]

[318]

[319][3110]

[3111]

[3112]

[3113]

[3114]

[3115]

[3116]

[3117]

[3118]

[3119]

[3120]

[3121]

[3122]109

box plot gamma[32]

[321][322]

[323]

[324][325]

[326]

[327][328][329]

[3210]

[3211]

[3212]

[3213][3214]

[3215]

[3216]

[3217]

[3218]

[3219]

[3220][3221]

[3222]

11

Figure C-1 Homogeneity Assessment

Note These results were obtained using the 1-4 gravimetric dilution-adjusted λ but uncorrected for droplet volume

or the 1-10 volumetric dilution That is the vertical axis in each panel should be multiplied by 07349times10 = 7349 This constant difference in scale does not impact the homogeneity analysis Error bars span approximate 95 confidence intervals

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

57

3 Certification measurements 31 Data

Originally the experiment consisted of 5 plates containing the 3 components (A B C) and 12 assays Ten of the assays were for genomic DNA identification (for the certified value) and two of the assays were for mitochondrial detection which is used for informational values (and are not included in this data set) Each of the plates contained samples from a new vial from different boxes (noted in the ldquovial usedrdquo portion the number references the box) The assay positions were alternated across the five plates as well For the plate design and which tubes were used see

Figure 6 Basic Plate Layout Design for Certification Measurements It was later determined that the homogeneity and stability data sets were obtained using a different set of pipettes than the certification data set and since all data was to be combined for certification another plate was obtained using the original pipette For the complete certification data see

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component Stability data was measured at different temperatures over several weeks using assays NEIF and ND6 Only the 4degC data was also used for certification For the complete stability data see

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

32 Model It was decided that all three sets of data should be used to derive the certified values In this way the uncertainty of the certified values would better reflect any potential heterogeneity in the samples as well as any additional unknown factors that may have effected the results Because two different pipettes were used to obtain different sets of data there was a significant bias in some of the measurements Specifically the ldquocertificationrdquo data from the first five plates was obtained using pipette set ldquo1rdquo and the remaining measurements were obtained using pipette set ldquo2rdquo This bias was introduced into the statistical model as part of the mean Possible effects of different assays and type of data (heterogeneity certification and stability) were accounted for in the statistical model in terms of a random effect Specifically the response for each component (A B C) was represented as

119910119910119894119894119894119894119894119894~1198731198731205741205741198941198941198941198941198941198941205901205901198941198942 where ~N() denotes ldquodistributed as a normal distribution of some mean and variancerdquo i = 123 denotes type of measurement heterogeneity certification and stability j = 1hellip10 denotes assay 2PR4 POTP NEIF R4Q5 D5 ND6 D9 HBB1 ND14 22C3 k = 1 2 denotes pipette set 120574120574119894119894119894119894119894119894 = 120572120572119894119894119894119894 + 120573120573119894119894 120572120572119894119894119894119894~119873119873(micro1198941198941205901205901205721205722) 120573120573119894119894 = 120573120573 for data type 1 and 120573120573119894119894 = 0 for data types 2 and 3 For a given 2372a component the variance 1205901205901205721205722 measures the between assay variability and 1205901205901198941198942 measures the repeatability variability for each data type The data was analyzed using Bayesian MCMC programmed in OpenBUGSC2 with non-informative Gaussian priors for the means and Gamma priors for the variance components

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

58

33 OpenBUGS code inits and data Certification code ModelBegin for(i in 13)mu[i]~dnorm(010E-5) sig[i]~dgamma(10E-510E-5) sigalph~dgamma(10E-510E-5) for(i in 110)alpha[1i]~dnorm(mu[1] sigalph) for(i in 12)alpha[2i]~dnorm(mu[2] sigalph) alpha[3i]~dnorm(mu[3]sigalph) b~dnorm(010E-5) beta[1]lt-b beta[2]lt-00 for(i in 1Ntot)mean[i]lt-alpha[typ[i]asy[i]]+beta[pip[i]] for(i in 1Ntot)lamb[i]~dnorm(mean[i]sig[typ[i]]) ModelEnd Certification Inits list(sig=c(111) sigalph=1) Certification Data for Component A list(Ntot=275) lamb[] asy[] cmp[] typ[] pip[] 1667 1 1 1 1 1655 2 1 1 1 1647 3 1 1 1 1594 4 1 1 1 1655 5 1 1 1 1629 6 1 1 1 1655 7 1 1 1 1644 8 1 1 1 1645 9 1 1 1 1545 10 1 1 1 1698 1 1 1 1 1685 2 1 1 1 1665 3 1 1 1 1658 4 1 1 1 1683 5 1 1 1 1647 6 1 1 1 1652 7 1 1 1 1611 8 1 1 1 1707 9 1 1 1 1603 10 1 1 1 1668 1 1 1 1 1587 2 1 1 1 1678 3 1 1 1 1628 4 1 1 1 1738 5 1 1 1 1667 6 1 1 1 1695 7 1 1 1 1757 8 1 1 1 1771 9 1 1 1 1762 10 1 1 1 1739 1 1 1 1 1702 2 1 1 1 1614 3 1 1 1 1762 4 1 1 1 1649 5 1 1 1 1678 6 1 1 1 1653 7 1 1 1

175 8 1 1 1 1764 9 1 1 1 1759 10 1 1 1 1638 1 1 1 1 1676 2 1 1 1 1616 3 1 1 1 1559 4 1 1 1 1608 5 1 1 1 1611 6 1 1 1 1517 7 1 1 1 1557 8 1 1 1 1604 9 1 1 1 1675 10 1 1 1 1667 1 1 1 1 1638 2 1 1 1 1624 3 1 1 1 1587 4 1 1 1 1652 5 1 1 1 1589 6 1 1 1 1607 7 1 1 1 1582 8 1 1 1 1625 9 1 1 1 1616 10 1 1 1 1888 1 1 1 1 1868 2 1 1 1 1953 3 1 1 1 1779 4 1 1 1 1924 5 1 1 1 1861 6 1 1 1 1824 7 1 1 1 1809 8 1 1 1 1864 9 1 1 1 193 10 1 1 1 1939 1 1 1 1 1815 2 1 1 1 1902 3 1 1 1 1871 4 1 1 1 2124 5 1 1 1

1787 6 1 1 1 1844 7 1 1 1 1799 8 1 1 1 1857 9 1 1 1 1942 10 1 1 1 1597 1 1 1 1 1533 2 1 1 1 1462 3 1 1 1 1493 4 1 1 1 1538 5 1 1 1 1492 6 1 1 1 1528 7 1 1 1 1548 8 1 1 1 1513 9 1 1 1 1562 10 1 1 1 159 1 1 1 1 1404 2 1 1 1 1533 3 1 1 1 1519 4 1 1 1 1586 5 1 1 1 1516 6 1 1 1 1538 7 1 1 1 1581 8 1 1 1 1514 9 1 1 1 1431 1 1 1 2 1407 2 1 1 2 1487 3 1 1 2 145 4 1 1 2 1394 5 1 1 2 1396 6 1 1 2 1414 7 1 1 2 1401 8 1 1 2 1412 9 1 1 2 1435 10 1 1 2 1411 1 1 1 2 1395 2 1 1 2 1398 3 1 1 2 1439 4 1 1 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

59

1422 5 1 1 2 1393 6 1 1 2 139 7 1 1 2 1382 8 1 1 2 146 9 1 1 2 141 10 1 1 2 1632 1 1 2 2 1531 1 1 2 2 1524 1 1 2 2 1454 1 1 2 2 1465 1 1 2 2 1512 1 1 2 2 1418 1 1 2 2 1496 1 1 2 2 1588 1 1 2 2 1574 1 1 2 2 1557 1 1 2 2 15 1 1 2 2 1672 1 1 2 2 1551 1 1 2 2 1573 1 1 2 2 1516 1 1 2 2 1612 1 1 2 2 1598 1 1 2 2 1575 1 1 2 2 1658 1 1 2 2 1547 1 1 2 2 1529 1 1 2 2 1525 1 1 2 2 154 1 1 2 2 1458 1 1 2 2 154 1 1 2 2 1491 1 1 2 2 148 1 1 2 2 156 1 1 2 2 1581 1 1 2 2 1513 1 1 2 2 1508 1 1 2 2 1476 1 1 2 2 1634 1 1 2 2 1568 1 1 2 2 152 1 1 2 2 1516 1 1 2 2 1535 1 1 2 2 1514 1 1 2 2 1568 1 1 2 2 1611 1 1 2 2 1528 1 1 2 2 1511 1 1 2 2 1549 1 1 2 2 1533 1 1 2 2 1427 1 1 2 2 1504 1 1 2 2 1491 1 1 2 2 1455 1 1 2 2

1501 1 1 2 2 1637 1 1 2 2 157 1 1 2 2 1633 1 1 2 2 1515 1 1 2 2 1804 1 1 2 2 1538 1 1 2 2 1617 1 1 2 2 1628 1 1 2 2 1672 1 1 2 2 1596 1 1 2 2 1586 1 1 2 2 1578 1 1 2 2 1532 1 1 2 2 1616 2 1 2 2 1615 2 1 2 2 1582 2 1 2 2 1627 2 1 2 2 1607 2 1 2 2 1511 2 1 2 2 1479 2 1 2 2 1514 2 1 2 2 1474 2 1 2 2 1594 2 1 2 2 1571 2 1 2 2 1626 2 1 2 2 149 2 1 2 2 1617 2 1 2 2 1634 2 1 2 2 1613 2 1 2 2 1529 2 1 2 2 1571 2 1 2 2 1652 2 1 2 2 1647 2 1 2 2 1537 2 1 2 2 164 2 1 2 2 1589 2 1 2 2 1152 2 1 2 2 1605 2 1 2 2 1633 2 1 2 2 1483 2 1 2 2 1557 2 1 2 2 1615 2 1 2 2 1438 2 1 2 2 1713 2 1 2 2 1611 2 1 2 2 1586 2 1 2 2 1516 2 1 2 2 1614 2 1 2 2 1627 2 1 2 2 1567 2 1 2 2 1495 2 1 2 2 1535 2 1 2 2 1545 2 1 2 2 1607 2 1 2 2

1513 2 1 2 2 1582 2 1 2 2 1601 2 1 2 2 1555 2 1 2 2 1607 2 1 2 2 1586 2 1 2 2 1524 2 1 2 2 1656 2 1 2 2 154 2 1 2 2 1454 2 1 2 2 1669 2 1 2 2 1631 2 1 2 2 1608 2 1 2 2 1497 2 1 2 2 1639 2 1 2 2 1575 2 1 2 2 1567 2 1 2 2 1528 2 1 2 2 1511 2 1 2 2 1627 2 1 2 2 1612 2 1 2 2 1536 2 1 2 2 1568 1 1 3 2 152 1 1 3 2 1486 1 1 3 2 1441 1 1 3 2 1706 1 1 3 2 144 1 1 3 2 15 1 1 3 2 1537 1 1 3 2 1387 1 1 3 2 1463 1 1 3 2 157 1 1 3 2 1429 1 1 3 2 1667 1 1 3 2 1632 1 1 3 2 1589 1 1 3 2 1474 2 1 3 2 1543 2 1 3 2 1555 2 1 3 2 1599 2 1 3 2 1494 2 1 3 2 1624 2 1 3 2 1577 2 1 3 2 1593 2 1 3 2 1479 2 1 3 2 1517 2 1 3 2 1538 2 1 3 2 159 2 1 3 2 1479 2 1 3 2 1479 2 1 3 2 1522 2 1 3 2 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

60

Table C-2 Example OpenBUGS Certification Summary for Component A

mean sd MC_error val25pc median val975pc start sample b 2584 02948 0005376 2001 2588 3158 100001 100000 mu[1] 1417 02706 0005366 1364 1416 147 100001 100000 mu[2] 1559 008368 4776E-4 1543 1559 1576 100001 100000 mu[3] 1533 01504 0001419 1504 1533 1563 100001 100000

Certification Data for Component B list(Ntot=278) lamb[] asy[] cmp[] typ[] pip[] 1869 1 2 1 1 185 2 2 1 1 1709 3 2 1 1 1856 4 2 1 1 1848 5 2 1 1 1828 6 2 1 1 1861 7 2 1 1 1825 8 2 1 1 1871 9 2 1 1 1776 10 2 1 1 196 1 2 1 1 1781 2 2 1 1 1707 3 2 1 1 1787 4 2 1 1 1829 5 2 1 1 1826 6 2 1 1 1858 7 2 1 1 1869 8 2 1 1 1773 9 2 1 1 1902 10 2 1 1 2174 1 2 1 1 2202 2 2 1 1 2172 3 2 1 1 223 5 2 1 1 2094 6 2 1 1 1949 7 2 1 1 1971 8 2 1 1 2123 9 2 1 1 2107 10 2 1 1 2231 1 2 1 1 2292 2 2 1 1 2216 3 2 1 1 2078 4 2 1 1 2192 5 2 1 1 2034 6 2 1 1 211 7 2 1 1 2123 8 2 1 1 2066 9 2 1 1 2046 10 2 1 1 181 1 2 1 1 1865 2 2 1 1 1806 3 2 1 1 1878 4 2 1 1 1824 5 2 1 1 1768 6 2 1 1 1811 7 2 1 1 1763 8 2 1 1 1865 9 2 1 1 1815 10 2 1 1

1806 1 2 1 1 1801 2 2 1 1 1786 3 2 1 1 1768 4 2 1 1 179 5 2 1 1 167 6 2 1 1 1749 7 2 1 1 1741 8 2 1 1 1781 9 2 1 1 1866 10 2 1 1 1784 1 2 1 1 1735 2 2 1 1 1778 3 2 1 1 1835 4 2 1 1 1794 5 2 1 1 17 6 2 1 1 1799 7 2 1 1 1765 8 2 1 1 1754 9 2 1 1 1792 10 2 1 1 1821 1 2 1 1 1813 2 2 1 1 1813 3 2 1 1 1749 4 2 1 1 1848 5 2 1 1 1711 6 2 1 1 1779 7 2 1 1 1769 8 2 1 1 1725 9 2 1 1 1814 10 2 1 1 1718 1 2 1 1 1753 2 2 1 1 1773 3 2 1 1 1777 4 2 1 1 1803 5 2 1 1 1718 6 2 1 1 1802 7 2 1 1 1834 8 2 1 1 1731 9 2 1 1 1842 10 2 1 1 1768 1 2 1 1 1698 2 2 1 1 1792 3 2 1 1 174 4 2 1 1 1795 5 2 1 1 1693 6 2 1 1 1743 7 2 1 1 1745 8 2 1 1 1714 9 2 1 1 1791 1 2 1 2

1746 2 2 1 2 1803 3 2 1 2 1767 4 2 1 2 1745 5 2 1 2 1824 6 2 1 2 1752 7 2 1 2 1813 8 2 1 2 1648 9 2 1 2 1818 10 2 1 2 177 1 2 1 2 182 2 2 1 2 1803 3 2 1 2 1739 4 2 1 2 1773 5 2 1 2 1729 6 2 1 2 1744 7 2 1 2 1754 8 2 1 2 1671 9 2 1 2 1741 10 2 1 2 179 1 2 2 2 1712 1 2 2 2 1693 1 2 2 2 1754 1 2 2 2 1757 1 2 2 2 1697 1 2 2 2 1722 1 2 2 2 1884 1 2 2 2 1713 1 2 2 2 1759 1 2 2 2 1749 1 2 2 2 1748 1 2 2 2 1632 1 2 2 2 1752 1 2 2 2 1871 1 2 2 2 1688 1 2 2 2 1789 1 2 2 2 1826 1 2 2 2 1751 1 2 2 2 1738 1 2 2 2 1742 1 2 2 2 1761 1 2 2 2 1909 1 2 2 2 1719 1 2 2 2 1755 1 2 2 2 1745 1 2 2 2 1687 1 2 2 2 1709 1 2 2 2 1625 1 2 2 2 1802 1 2 2 2 1734 1 2 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

61

1809 1 2 2 2 1782 1 2 2 2 1712 1 2 2 2 1668 1 2 2 2 1535 1 2 2 2 1835 1 2 2 2 1725 1 2 2 2 1769 1 2 2 2 178 1 2 2 2 1759 1 2 2 2 1678 1 2 2 2 1836 1 2 2 2 1743 1 2 2 2 1885 1 2 2 2 1704 1 2 2 2 1749 1 2 2 2 1739 1 2 2 2 1735 1 2 2 2 1714 1 2 2 2 1703 1 2 2 2 1772 1 2 2 2 1727 1 2 2 2 1738 1 2 2 2 174 1 2 2 2 1718 1 2 2 2 1701 1 2 2 2 1673 1 2 2 2 1887 1 2 2 2 1743 1 2 2 2 1825 1 2 2 2 1849 1 2 2 2 188 1 2 2 2 1768 1 2 2 2 1713 1 2 2 2 166 1 2 2 2 1775 2 2 2 2 1824 2 2 2 2 167 2 2 2 2 1683 2 2 2 2 1693 2 2 2 2 169 2 2 2 2 1864 2 2 2 2 1709 2 2 2 2 1764 2 2 2 2

1741 2 2 2 2 1797 2 2 2 2 1626 2 2 2 2 1724 2 2 2 2 177 2 2 2 2 1744 2 2 2 2 1716 2 2 2 2 1805 2 2 2 2 1827 2 2 2 2 1752 2 2 2 2 1746 2 2 2 2 1666 2 2 2 2 187 2 2 2 2 1718 2 2 2 2 1483 2 2 2 2 1632 2 2 2 2 1759 2 2 2 2 1723 2 2 2 2 1684 2 2 2 2 1887 2 2 2 2 1756 2 2 2 2 1816 2 2 2 2 1792 2 2 2 2 1614 2 2 2 2 1566 2 2 2 2 1741 2 2 2 2 1752 2 2 2 2 1697 2 2 2 2 1762 2 2 2 2 1801 2 2 2 2 1766 2 2 2 2 1747 2 2 2 2 1744 2 2 2 2 1692 2 2 2 2 186 2 2 2 2 1756 2 2 2 2 1896 2 2 2 2 1814 2 2 2 2 1703 2 2 2 2 1722 2 2 2 2 1688 2 2 2 2 1821 2 2 2 2 1758 2 2 2 2 1787 2 2 2 2

1784 2 2 2 2 179 2 2 2 2 1681 2 2 2 2 1685 2 2 2 2 1798 2 2 2 2 171 2 2 2 2 1638 2 2 2 2 1931 2 2 2 2 1786 2 2 2 2 1768 2 2 2 2 1764 2 2 2 2 1819 2 2 2 2 1666 1 2 3 2 1626 1 2 3 2 188 1 2 3 2 1653 1 2 3 2 1733 1 2 3 2 1551 1 2 3 2 1884 1 2 3 2 1718 1 2 3 2 1549 1 2 3 2 1723 1 2 3 2 194 1 2 3 2 164 1 2 3 2 1565 1 2 3 2 1787 1 2 3 2 1655 2 2 3 2 167 2 2 3 2 1673 2 2 3 2 1705 2 2 3 2 1755 2 2 3 2 1625 2 2 3 2 1708 2 2 3 2 1705 2 2 3 2 1862 2 2 3 2 1686 2 2 3 2 1723 2 2 3 2 1909 2 2 3 2 1859 2 2 3 2 1897 2 2 3 2 1972 2 2 3 2 END

Table C-3 Example OpenBUGS Certification Summary for Component B

mean sd MC_error val25pc median val975pc start sample b 09337 03387 0007504 02695 09371 1592 100001 100000 mu[1] 1764 03101 0007483 1704 1763 1825 100001 100000 mu[2] 1748 008014 4719E-4 1733 1748 1764 100001 100000 mu[3] 1735 02338 0003093 1688 1736 1781 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

62

Certification data for Component C list(Ntot=280) lamb[] asy[] cmp[] typ[] pip[] 163 1 3 1 1 1596 2 3 1 1 1574 3 3 1 1 159 4 3 1 1 161 5 3 1 1 1589 6 3 1 1 1569 7 3 1 1 1565 8 3 1 1 1564 9 3 1 1 1548 10 3 1 1 1706 1 3 1 1 1576 2 3 1 1 1538 3 3 1 1 1555 4 3 1 1 1607 5 3 1 1 1548 6 3 1 1 1572 7 3 1 1 1681 8 3 1 1 1492 9 3 1 1 1655 10 3 1 1 1499 1 3 1 1 1576 2 3 1 1 1545 3 3 1 1 1584 4 3 1 1 1577 5 3 1 1 1599 6 3 1 1 1587 7 3 1 1 152 8 3 1 1 1479 9 3 1 1 1512 10 3 1 1 1497 1 3 1 1 1611 2 3 1 1 156 3 3 1 1 1546 4 3 1 1 1529 5 3 1 1 1478 6 3 1 1 1504 7 3 1 1 1551 8 3 1 1 1438 9 3 1 1 1494 10 3 1 1 1452 1 3 1 1 1347 2 3 1 1 1395 3 3 1 1 1442 4 3 1 1 1448 5 3 1 1 1463 6 3 1 1 1402 7 3 1 1 1474 8 3 1 1 1466 9 3 1 1 1439 10 3 1 1 1424 1 3 1 1 1397 2 3 1 1 135 3 3 1 1 1459 4 3 1 1 1602 5 3 1 1 1435 6 3 1 1 1433 7 3 1 1 1481 8 3 1 1 1482 9 3 1 1

1568 10 3 1 1 147 1 3 1 1 1411 2 3 1 1 1512 3 3 1 1 1452 4 3 1 1 1567 5 3 1 1 148 6 3 1 1 1442 7 3 1 1 143 8 3 1 1 1413 9 3 1 1 1486 10 3 1 1 1444 1 3 1 1 1427 2 3 1 1 1439 3 3 1 1 1468 4 3 1 1 1524 5 3 1 1 1424 6 3 1 1 1459 7 3 1 1 1467 8 3 1 1 1411 9 3 1 1 148 10 3 1 1 1485 1 3 1 1 1422 2 3 1 1 1438 3 3 1 1 1439 4 3 1 1 1515 5 3 1 1 1409 6 3 1 1 1469 7 3 1 1 1453 8 3 1 1 1465 9 3 1 1 1363 10 3 1 1 1458 1 3 1 1 1459 2 3 1 1 1465 3 3 1 1 1468 4 3 1 1 1457 5 3 1 1 1412 6 3 1 1 1474 7 3 1 1 1428 8 3 1 1 1425 9 3 1 1 1393 10 3 1 1 1357 1 3 1 2 1447 2 3 1 2 1448 3 3 1 2 1409 4 3 1 2 1389 5 3 1 2 1415 6 3 1 2 1431 7 3 1 2 1381 8 3 1 2 1403 9 3 1 2 1435 10 3 1 2 1364 1 3 1 2 1361 2 3 1 2 1404 3 3 1 2 1411 4 3 1 2 1429 5 3 1 2 1417 6 3 1 2 138 7 3 1 2 1336 8 3 1 2 1387 9 3 1 2

1372 10 3 1 2 1493 1 3 2 2 1516 1 3 2 2 1451 1 3 2 2 144 1 3 2 2 1523 1 3 2 2 1437 1 3 2 2 1485 1 3 2 2 1433 1 3 2 2 1463 1 3 2 2 1534 1 3 2 2 1533 1 3 2 2 145 1 3 2 2 1537 1 3 2 2 1515 1 3 2 2 1511 1 3 2 2 1449 1 3 2 2 1388 1 3 2 2 1475 1 3 2 2 1468 1 3 2 2 1549 1 3 2 2 153 1 3 2 2 1442 1 3 2 2 1385 1 3 2 2 1513 1 3 2 2 1528 1 3 2 2 1486 1 3 2 2 1472 1 3 2 2 1493 1 3 2 2 1477 1 3 2 2 1405 1 3 2 2 1499 1 3 2 2 1458 1 3 2 2 1514 1 3 2 2 1525 1 3 2 2 1527 1 3 2 2 149 1 3 2 2 1568 1 3 2 2 1452 1 3 2 2 1421 1 3 2 2 1455 1 3 2 2 1525 1 3 2 2 1531 1 3 2 2 144 1 3 2 2 1487 1 3 2 2 151 1 3 2 2 1577 1 3 2 2 1506 1 3 2 2 152 1 3 2 2 1483 1 3 2 2 1522 1 3 2 2 1463 1 3 2 2 1478 1 3 2 2 1505 1 3 2 2 1454 1 3 2 2 1555 1 3 2 2 1497 1 3 2 2 1515 1 3 2 2 1494 1 3 2 2 1594 1 3 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

63

1471 1 3 2 2 1394 1 3 2 2 1445 1 3 2 2 1446 1 3 2 2 1499 1 3 2 2 1484 1 3 2 2 1436 1 3 2 2 1565 2 3 2 2 149 2 3 2 2 1485 2 3 2 2 1529 2 3 2 2 1513 2 3 2 2 1492 2 3 2 2 1475 2 3 2 2 1497 2 3 2 2 1503 2 3 2 2 1448 2 3 2 2 1488 2 3 2 2 1496 2 3 2 2 1499 2 3 2 2 1469 2 3 2 2 1515 2 3 2 2 1543 2 3 2 2 1452 2 3 2 2 1458 2 3 2 2 1435 2 3 2 2 1499 2 3 2 2 1499 2 3 2 2 1589 2 3 2 2 1562 2 3 2 2 1547 2 3 2 2 1505 2 3 2 2 1508 2 3 2 2 152 2 3 2 2

1481 2 3 2 2 1547 2 3 2 2 1501 2 3 2 2 1491 2 3 2 2 1465 2 3 2 2 147 2 3 2 2 1483 2 3 2 2 1479 2 3 2 2 148 2 3 2 2 1543 2 3 2 2 1448 2 3 2 2 1531 2 3 2 2 1478 2 3 2 2 1465 2 3 2 2 1454 2 3 2 2 1591 2 3 2 2 1466 2 3 2 2 157 2 3 2 2 1471 2 3 2 2 1506 2 3 2 2 1491 2 3 2 2 1461 2 3 2 2 1485 2 3 2 2 1495 2 3 2 2 1516 2 3 2 2 1474 2 3 2 2 1483 2 3 2 2 1519 2 3 2 2 1475 2 3 2 2 1466 2 3 2 2 1565 2 3 2 2 1491 2 3 2 2 1435 2 3 2 2 1459 2 3 2 2

1421 2 3 2 2 1468 2 3 2 2 1556 2 3 2 2 1424 1 3 3 2 1427 1 3 3 2 1381 1 3 3 2 1394 1 3 3 2 1473 1 3 3 2 133 1 3 3 2 1446 1 3 3 2 1424 1 3 3 2 133 1 3 3 2 1452 1 3 3 2 1512 1 3 3 2 1314 1 3 3 2 1525 1 3 3 2 1475 1 3 3 2 1542 1 3 3 2 1461 2 3 3 2 1443 2 3 3 2 1547 2 3 3 2 1394 2 3 3 2 1364 2 3 3 2 1395 2 3 3 2 1522 2 3 3 2 1483 2 3 3 2 1452 2 3 3 2 1482 2 3 3 2 1439 2 3 3 2 1458 2 3 3 2 155 2 3 3 2 1502 2 3 3 2 152 2 3 3 2 END

Table C-4 Example OpenBUGS Certification Summary for Component C

mean sd MC_error val25pc median val975pc start sample b 09626 01701 0002428 06285 09633 1293 100001 100000 mu[1] 1399 01566 0002413 1369 1399 143 100001 100000 mu[2] 1491 005701 2386E-4 148 1491 1503 100001 100000 mu[3] 1449 0132 0001181 1423 1449 1475 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

64

34 Results Table C-5 lists the results in CopiesnL for the three components

Table C-5 Parameter Values for Components A B and C

Posterior Distribution Summary Statistics Component Parameter Mean SD 25 50 975

A

β 260 029 203 260 318 μ1 1415 027 1362 1415 1467 μ2 1559 010 1539 1559 158 μ3 1533 016 1501 1533 1565

B

β 095 034 029 095 162 μ1 1762 031 1701 1762 1822 μ2 1748 010 1728 1748 1769 μ3 1735 024 1687 1735 1783

C

β 097 017 064 097 130 μ1 1399 016 1368 1399 1429 μ2 1491 008 1476 1491 1507 μ3 1449 014 1421 1449 1477

Since in each case the three data set types represent independent measurements the three estimates can be combined to obtain a consensus value In this way a between data set type variability is included in the final standard uncertainty Table C-6 lists the results in Copies per Nanoliter obtained using the NIST Consensus Builder (NICOB)C3

Table C-6 Certified Values for Components A B and C

Component Mean u(Mean) Uk=2(Mean) A 151 07 14 B 175 03 06 C 145 05 10

4 Mito-to-nuclear DNA ratios (mtDNAnDNA)

41 Data Measurements were also obtained on the same samples using three mitochondrial detection assays mtND1 mtBatz and mtKav and a single nDNA assay (genomic) to compute the ratios mtDNAnDNA For the complete mtDNAnDNA data see

Table 16 mtDNAnDNA Ratios

42 Model For each component the data was analyzed using the following random effects model to account for between assay variability

119903119903119894119894~119873119873120572120572119894119894 1205901205901198861198861198861198861198861198861198861198861198861198862

where 119894119894 denotes assay (for A and B 1 = mtND1 2 = mtBatz 3 = mtKav for C 1 = mtBatz 2 = mtKav) and 120572120572119894119894~119873119873micro120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119886119886119886119886119886119886119886119886119886119886

2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

65

43 OpenBUGS code inits and data Two versions of the same model are required one for the results from the three mtDNA assays available for components A and B mtND1 mtBatz and mtKav and the other for the results from the two mtDNA assays available for component C mtBatz and mtKav mtDNAnDNA code for components A and B ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 13)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for components A and B list(sigt=c(111)) Data for Component A list(Ndata=102) set[] rep[] asy[] rat[] 1 1 1 1931 1 2 1 187 1 1 2 1845 1 2 2 1764 2 1 1 181 2 2 1 1906 2 1 2 1785 2 2 2 1837 3 1 1 1743 3 2 1 1761 3 1 2 1829 3 2 2 1847 4 1 1 1716 4 2 1 180 4 1 2 174 4 2 2 1813 5 1 1 1923 5 2 1 1835 5 1 2 1935 5 2 2 1851 6 1 1 1757 6 2 1 1724 6 3 1 1768 6 1 2 1743 6 2 2 1671 6 3 2 1766 6 1 3 1692 6 2 3 1662 6 3 3 1767 7 1 1 1924 7 2 1 1944 7 3 1 1893 7 1 2 1786

7 2 2 1752 7 3 2 1743 7 1 3 153 7 2 3 1542 7 3 3 1523 7 1 2 158 7 2 2 1557 7 3 2 1592 7 1 3 2103 7 2 3 210 7 3 3 2049 7 1 1 1749 7 2 1 1769 7 3 1 1777 7 1 2 1965 7 2 2 1889 7 3 2 1947 7 1 3 1813 7 2 3 1864 7 3 3 1864 8 1 1 178 8 2 1 1776 8 3 1 1706 8 1 2 1804 8 2 2 1751 8 3 2 1738 8 1 3 1769 8 2 3 1786 8 3 3 1761 9 1 2 1584 9 2 2 1544 10 1 1 1947 10 2 1 1928 10 3 1 1922 10 1 1 1829

10 2 1 1895 10 3 1 1851 10 1 1 1674 10 2 1 1617 11 1 2 1667 11 2 2 1813 11 3 2 1708 11 4 2 1598 11 1 2 1566 11 2 2 1687 11 3 2 1663 11 4 2 1589 11 1 2 1643 11 2 2 1668 11 3 2 1604 11 4 2 1573 12 1 2 1719 12 2 2 1716 12 3 2 167 12 4 2 1666 12 1 2 1581 12 2 2 1573 12 3 2 1601 12 4 2 1612 12 1 2 1546 12 2 2 1542 12 3 2 1538 12 4 2 1587 13 1 1 1617 13 2 1 1599 13 1 1 1622 13 2 1 1579 13 1 1 1575 13 2 1 1528 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

66

Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A

mean sd MC_error val25pc median val975pc start sample mu 1730 1443 004419 1703 1730 1759 100001 100000

data for Component B list(Ndata=87) set[] rep[] asy[] rat[] 1 1 1 2461 1 2 1 2401 1 1 2 2555 1 2 2 2608 2 1 1 2229 2 2 1 2200 2 1 2 2108 2 2 2 2192 3 1 1 2127 3 2 1 2198 3 1 2 2455 3 2 2 2506 4 1 1 2211 4 2 1 2199 4 1 2 2380 4 2 2 2249 5 1 1 2105 5 2 1 1943 5 1 2 2136 5 2 2 2119 6 1 1 1990 6 2 1 1921 6 3 1 1982 6 1 2 1990 6 2 2 1924 6 3 2 1958 6 1 3 1903 6 2 3 1831

6 3 3 1919 7 1 1 2304 7 2 1 2341 7 3 1 2332 7 1 2 2048 7 2 2 2078 7 3 2 2067 7 1 3 2399 7 2 3 2479 7 3 3 2493 7 1 1 2065 7 2 1 2166 7 3 1 2164 7 1 2 2314 7 2 2 2280 7 3 2 2258 7 1 3 1867 7 2 3 1987 7 3 3 1971 8 1 1 2046 8 2 1 2057 8 3 1 2024 8 1 2 1998 8 2 2 2043 8 3 2 1952 8 1 3 2077 8 2 3 2117 8 3 3 2041 9 1 2 1890 9 2 2 1886

10 1 2 1928 10 2 2 1916 10 3 2 1915 10 1 2 1947 10 2 2 1946 10 3 2 1961 10 4 2 1909 10 1 2 1953 10 2 2 1952 10 3 2 1941 10 4 2 1881 11 1 2 1865 11 2 2 1863 11 3 2 1828 11 4 2 1828 11 1 2 1944 11 2 2 1964 11 3 2 1947 11 4 2 1895 11 1 2 1908 11 2 2 1909 11 3 2 1940 11 4 2 1839 12 1 1 1688 12 2 1 1783 12 1 1 1823 12 2 1 1631 12 1 1 1920 12 2 1 1887 END

Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B

mean sd MC_error val25pc median val975pc start sample mu 2045 2246 0111 1999 2045 2088 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

67

mtDNAnDNA code for component C ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 12)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for component C list(sigt=c(11)) mtDNAnDNA data for component C list(Ndata=66) set[] rep[] asy[] rat[] 1 1 1 3201 1 2 1 3299 2 1 1 3347 2 2 1 3287 3 1 1 3217 3 2 1 3384 4 1 1 3200 4 2 1 3363 5 1 1 3294 5 2 1 3013 6 1 1 2805 6 2 1 2825 6 3 1 2829 6 1 2 2709 6 2 2 2738 6 3 2 2727 7 1 1 2797 7 2 1 2642 7 3 1 2792 7 1 2 2804 7 2 2 2789

7 3 2 2842 7 1 1 2715 7 2 1 2606 7 3 1 2687 7 1 2 2799 7 2 2 2628 7 3 2 2559 7 1 1 2639 7 2 1 2535 7 3 1 2657 7 1 2 2953 7 2 2 2757 7 3 2 2802 8 1 1 2927 8 2 1 2776 8 3 1 2762 8 1 2 3044 8 2 2 2830 8 3 2 2825 9 1 1 2501 9 2 1 2522 10 1 1 2832 10 2 1 2840

10 3 1 2788 10 4 1 2725 10 1 1 2835 10 2 1 2790 10 3 1 2736 10 4 1 2679 10 1 1 2803 10 2 1 2773 10 3 1 2757 10 4 1 2752 11 1 1 2599 11 2 1 2508 11 3 1 2551 11 4 1 2609 11 1 1 2617 11 2 1 2483 11 3 1 2626 11 4 1 2602 11 1 1 2665 11 2 1 2535 11 3 1 2570 11 4 1 2650 END

Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C

mean sd MC_error val25pc median val975pc start sample mu 2774 2367 01176 2729 2774 2821 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

68

Table C-10 Results for mtDNAnDNA Posterior Distribution Summary Statistics

Component Mean Std Dev 25 50 975 A 1730 14 1703 1730 1758 B 2045 23 1994 2047 2087 C 2774 23 2729 2774 2821

5 ReferencesC1 Lunn DJ Spiegelhalter D Thomas A Best N Statistics in Medicine 2009283049ndash3082C2 Polson NG Scott JG Bayesian Analysis 20127887ndash902C3 NIST Consensus Builder Software and userrsquos guide freely available at

httpsconsensusnistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

69

Appendix D Performance in Commercial qPCR Chemistries The performance of the SRM 2372a materials in commercial qPCR chemistries was assessed using manufacturerrsquos recommended protocols Each panel in Figures D-1 to D-4 displays the average cycle threshold (Ct) as a function of DNA concentration for the 5-point dilutions of SRM 2372a components In all cases the external calibrant was SRM 2372 component A Error bars represent approximate 95 confidence intervals Table D-1 lists the estimated [DNA] for the three SRM 2372a components for all commercial assays evaluated This Table also lists the Component BComponent A (BA) and Component CComponent A (CA) ratios calculated from the [DNA] estimates

Figure D-1 Results for the Innogenomics Commercial qPCR assays

20

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantLong(IQ_h)

ABCStd

[A] = (384 plusmn 26) ngμL[B] = (441 plusmn 24) ngμL[C] = (516 plusmn 24) ngμL

Rsup2 = 0999 se(Ct) = 01016

17

18

19

20

21

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantShort(IQ_d)

ABCStd

[A] = (435 plusmn 28) ngμL[B] = (543 plusmn 12) ngμL[C] = (510 plusmn 28) ngμL

Rsup2 = 0999 se(Ct) = 007

24

25

26

27

28

29

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYLong(IQHY_h)

ABCStd

[A] = (387 plusmn 24) ngμL[B] = (448 plusmn 21) ngμL[C] = (559 plusmn 31) ngμL

Rsup2 = 0998 se(Ct) = 01314

16

18

20

22

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYShort(IQHY_d)

ABCStd

[A] = (353 plusmn 23) ngμL[B] = (267 plusmn 15) ngμL[C] = (390 plusmn 18) ngμL

Rsup2 = 1000 se(Ct) = 00420

22

24

26

28

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYY(IQHY_y)

ABCStd

[A] = (572 plusmn 13) ngμL[B] = 0 ngμL[C] = (137 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

70

Figure D-2 Results for the Thermo Fisher Commercial qPCR assays

21

22

23

24

25

26

27

28

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HumanHuman(QFH_h)

ABCStd

[A] = (464 plusmn 13) ngμL[B] = (529 plusmn 19) ngμL[C] = (472 plusmn 22) ngμL

Rsup2 = 1000 se(Ct) = 004

22

23

24

25

26

27

28

29

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoHuman(QFD_h)

ABCStd

[A] = (701 plusmn 53) ngμL[B] = (561 plusmn 11) ngμL[C] = (472 plusmn 11) ngμL

Rsup2 = 0994 se(Ct) = 01922

24

26

28

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoMale(QFD_y)

ABCStd

[A] = (706 plusmn 49) ngμL[B] = 0 ngμL[C] = (122 plusmn 02) ngμL

Rsup2 = 0999 se(Ct) = 007

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPLarge(QFHP_h)

ABCStd

[A] = (340 plusmn 18) ngμL[B] = (432 plusmn 05) ngμL[C] = (342 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 00720

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPSmall(QFHP_d)

ABCStd

[A] = (434 plusmn 15) ngμL[B] = (514 plusmn 10) ngμL[C] = (466 plusmn 07) ngμL

Rsup2 = 1000 se(Ct) = 005

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioLarge(QFT_h)

ABCStd

[A] = (362 plusmn 23) ngμL[B] = (430 plusmn 06) ngμL[C] = (378 plusmn 07) ngμL

Rsup2 = 0999 se(Ct) = 00920

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioSmall(QFT_d)

ABCStd

[A] = (468 plusmn 08) ngμL[B] = (505 plusmn 08) ngμL[C] = (523 plusmn 14) ngμL

Rsup2 = 1000 se(Ct) = 00318

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioY(QFT_y)

ABCStd

[A] = (372 plusmn 11) ngμL[B] = 0 ngμL[C] = (110 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 003

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

71

Figure D-3 Results for the Qiagen Commercial qPCR assays

Figure D-4 Results for the Promega Commercial qPCR assays

19

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresAutosomal(QPH_h)

ABCStd

[A] = (554 plusmn 21) ngμL[B] = (479 plusmn 05) ngμL[C] = (381 plusmn 09) ngμL

Rsup2 = 1000 se(Ct) = 00418

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresY(QPH_y)

ABCStd

[A] = (358 plusmn 10) ngμL[B] = 0 ngμL[C] = (118 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 007

16

17

18

19

20

21

22

23

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProHuman(QPP_h)

ABCStd

[A] = (525 plusmn 49) ngμL[B] = (480 plusmn 42) ngμL[C] = (467 plusmn 45) ngμL

Rsup2 = 0999 se(Ct) = 01019

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProDegradation(QPP_d)

ABCStd

[A] = (567 plusmn 18) ngμL[B] = (558 plusmn 08) ngμL[C] = (484 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 00410

15

20

25

30

35

-50 -40 -30 -20 -10 00 10

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProY(QPP_y)

ABCStd

[A] = (538 plusmn 27) ngμL[B] = (00 plusmn 00) ngμL[C] = (162 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 003

16

18

20

22

24

-05 00 05 10 15 20

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorAutosomal(P_h)

ABCStd

[A] = (759 plusmn 64) ngμL[B] = (775 plusmn 24) ngμL[C] = (1056 plusmn 17) ngμL

Rsup2 = 0995 se(Ct) = 01818

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorY(P_y)

ABCStd

[A] = (596 plusmn 51) ngμL[B] = 0 ngμL[C] = (117 plusmn 01) ngμL

Rsup2 = 0994 se(Ct) = 020

18

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantAutosomal(PQ_h)

ABCStd

[A] = (436 plusmn 13) ngμL[B] = (606 plusmn 20) ngμL[C] = (504 plusmn 10) ngμL

Rsup2 = 1000 se(Ct) = 00218

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantDegradation(PQ_d)

ABCStd

[A] = (420 plusmn 18) ngμL[B] = (529 plusmn 16) ngμL[C] = (530 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 00516

18

20

22

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantY(PQ_y)

ABCStd

[A] = (410 plusmn 20) ngμL[B] = 0 ngμL[C] = (139 plusmn 05) ngμL

Rsup2 = 0999 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

72

Table D-1 Summary [DNA] Results for the Commercial qPCR Assays Estimated [DNA] ngμL Ratios

A B C BA CA Assaya xb u(x)c xb u(x)c xb u(x)c yd u(y)e yd u(y)e

IQd 435 28 543 12 510 28 125 008 117 010 IQh 384 26 441 24 516 24 115 010 134 011

IQHYd 353 23 267 15 390 18 076 006 111 009 IQHYh 387 24 448 21 559 31 116 009 144 012

Ph 759 64 775 24 1056 16 102 009 139 012 PQd 420 18 529 16 530 06 126 007 126 006 PQh 436 13 606 20 504 10 139 006 116 004

QFDh 701 53 561 11 472 11 080 006 067 005 QFHh 464 13 529 19 472 22 114 005 102 006

QFHPd 434 15 514 10 466 07 118 005 107 004 QFHPh 340 18 432 05 342 03 127 007 101 005

QFTd 468 08 505 08 523 14 108 003 112 004 QFTh 362 23 430 06 378 07 118 008 104 007 QPHh 554 21 479 05 380 09 086 003 069 003 QPPd 567 18 558 08 484 11 098 003 085 003 QPPh 525 49 480 42 467 45 092 012 089 012

IQHYy 572 13 NDf 137 06

NRg

024 001 Py 596 51 NDf 117 01 020 002

PQy 410 20 NDf 139 05 034 002 QFDy 706 49 NDf 122 02 017 001 QFTy 372 11 NDf 110 06 030 002 QPHy 358 10 NDf 118 03 033 001 QPPy 538 27 0002 0001 162 11 030 003

a Assay codes are listed in Table 8 b Mean of dilution-adjusted results from five dilutions of the SRM 2372a component c Standard uncertainty of the mean estimated from the stdanrd deviation of the five dilutions d Ratio of the estimated [DNA] of component B or C relative to that of component A e Standard uncertainty of the ratio estimated from the standard uncertainties of the individual components f Not detected Component B was prepared using tissue from only female donors it is not expected to

respond to male-specific assays g Not relevant

  • Abstract
  • Keywords
  • Technical Information Contact for this SRM
  • Table of Contents
  • Table of Tables
  • Table of Figures
  • Purpose and Description
  • Warning SRM 2372a is a Human Source Material
  • Storage and Use
  • History and Background
    • Figure 1 Sales History of SRM 2372
      • Experimental Methods
        • Polymerase Chain Reaction Assays
          • Table 1 NIST-Developed Human Nuclear DNA Assays
          • Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays
          • Table 2 NIST-Optimized Human Mitochondrial DNA Assays
          • Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays
            • Droplet Digital Polymerase Chain Reaction Measurements
              • Nuclear DNA (nDNA)
                • Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays
                  • Mitochondrial DNA (mtDNA)
                    • Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays
                    • Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays
                        • Sample Preparation
                        • Absorbance Measurements
                        • SRM Unit Production
                        • Component Homogeneity
                          • Figure 4 Plate Layout Design for Homogeneity Measurements
                            • Component Stability
                              • Figure 5 Plate Layout Design for Stability Measurements
                                • Component Certification Measurements
                                  • Table 6 Tubes used for Certification Measurements
                                  • Figure 6 Basic Plate Layout Design for Certification Measurements
                                    • Mitochondrial DNA Measurements
                                      • Table 7 Run Parameters for Mitochondrial DNA
                                        • Performance in Commercial qPCR Chemistries
                                          • Table 8 Commercial qPCR Chemistries Tested with SRM 2372a
                                          • Figure 7 Plate Layout Design for Performance Assessments
                                              • Measurement Results
                                                • Absorbance Spectrophotometry
                                                  • Table 9 Absorbance at Selected Wavelengths
                                                  • Figure 8 Absorbance Spectra of SRM 2372a Components A B and C
                                                    • Droplet Volume
                                                    • Certification Measurements
                                                      • Homogeneity Results
                                                        • Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component
                                                        • Table 11 Within- and Between-Tube Relative Standard Deviations
                                                          • Stability Results
                                                            • Figure 9 ddPCR Stability Measurements as Function of Time
                                                            • Table 12 Stability Data as λacute Copy Number per Nanoliter of Component
                                                              • Ten-Assay Certification Results
                                                                • Figure 10 Comparison of Assay Performance
                                                                • Table 13 Certification Data as λacute Copy Number per Nanoliter of Component
                                                                  • Estimation of Copy Number Concentration
                                                                    • Table 14 Measured Copy Number Concentrations
                                                                      • Conversion of Copy Number to Mass Concentration
                                                                        • Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution
                                                                            • Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)
                                                                              • Figure 11 mtDNAnDNA for Components A B and C
                                                                              • Table 16 mtDNAnDNA Ratios
                                                                              • Table 17 mtDNAnDNA Summary Results
                                                                                • Performance in Commercial qPCR Chemistries
                                                                                  • Figure 12 qPCR Autosomal Quantitation Performance Summary
                                                                                  • Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary
                                                                                      • Qualification of ddPCR as Fit-For-Purpose
                                                                                        • Figure 14 Absorbance Spectra of SRM 2372 Components A and B
                                                                                        • Digital Polymerase Chain Reaction Assays
                                                                                        • Heat Treatments (Melting of the DNA into Single Strands)
                                                                                        • Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a
                                                                                          • Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a
                                                                                            • Quantitative Evaluation of ssDNA Content
                                                                                              • ddPCR Measurement of ssDNA Proportion
                                                                                              • Proof of Principle
                                                                                                • Table 19 Comparison of Snap-Cooling to Slow-Cooling
                                                                                                  • Time at Denaturation Temperature
                                                                                                    • Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures
                                                                                                      • Temperature
                                                                                                        • Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a
                                                                                                          • DNA Concentration in the ddPCR Solution
                                                                                                            • Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution
                                                                                                              • Equivalence of Assays
                                                                                                                • Table 23 Comparison of Assays
                                                                                                                    • ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials
                                                                                                                      • Table 24 Evaluation of Limiting Ratio
                                                                                                                          • Conclusions and Recommendations
                                                                                                                            • Recommended Certification Values
                                                                                                                              • Table 25 Recommended Values and 95 Uncertainties for Certification
                                                                                                                                • Use of SRM 2372a for Value-Assigning In-House Reference Materials
                                                                                                                                  • Certificate of Analysis
                                                                                                                                  • References
                                                                                                                                  • Appendix A DNA Extraction Method
                                                                                                                                    • Procedure
                                                                                                                                    • Reagents
                                                                                                                                    • Reference
                                                                                                                                      • Appendix B Spectrophotometric Measurement Protocol
                                                                                                                                        • Spectrophotometer Calibration
                                                                                                                                          • Establish Baseline
                                                                                                                                          • Enable Verification of Wavelength Calibration
                                                                                                                                          • Enable Verification of Absorbance Calibration
                                                                                                                                            • For ldquonativerdquo dsDNA Samples
                                                                                                                                              • Verify Optical Balance of Cuvettes
                                                                                                                                                • For NaOH denatured ssDNA Samples
                                                                                                                                                  • Prepare a 04 molL NaOH Stock Solution
                                                                                                                                                  • Prepare the Sample Solutions
                                                                                                                                                  • Measure Sample Solutions
                                                                                                                                                    • After All Sample Measurements are Complete
                                                                                                                                                      • Enable Verification of Spectrophotometer Stability
                                                                                                                                                      • Clean Up
                                                                                                                                                        • Evaluation
                                                                                                                                                        • Reference
                                                                                                                                                          • Appendix C SRM 2372a Statisticianrsquos report
                                                                                                                                                            • 1 Introduction
                                                                                                                                                            • 11 OpenBUGS Analysis
                                                                                                                                                            • 12 OpenBUGS Exemplar Output
                                                                                                                                                            • 2 Homogeneity assessment
                                                                                                                                                            • 21 Data
                                                                                                                                                            • 22 Model
                                                                                                                                                            • 23 OpenBUGS code and data
                                                                                                                                                              • Table C-1 Example OpenBUGS Homogeneity Summary
                                                                                                                                                                • 24 Results
                                                                                                                                                                  • Figure C-1 Homogeneity Assessment
                                                                                                                                                                    • 3 Certification measurements
                                                                                                                                                                    • 31 Data
                                                                                                                                                                    • 32 Model
                                                                                                                                                                    • 33 OpenBUGS code inits and data
                                                                                                                                                                      • Table C-2 Example OpenBUGS Certification Summary for Component A
                                                                                                                                                                      • Table C-3 Example OpenBUGS Certification Summary for Component B
                                                                                                                                                                      • Table C-4 Example OpenBUGS Certification Summary for Component C
                                                                                                                                                                        • 34 Results
                                                                                                                                                                          • Table C-5 Parameter Values for Components A B and C
                                                                                                                                                                          • Table C-6 Certified Values for Components A B and C
                                                                                                                                                                            • 4 Mito-to-nuclear DNA ratios (mtDNAnDNA)
                                                                                                                                                                            • 41 Data
                                                                                                                                                                            • 42 Model
                                                                                                                                                                            • 43 OpenBUGS code inits and data
                                                                                                                                                                              • Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A
                                                                                                                                                                              • Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B
                                                                                                                                                                              • Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C
                                                                                                                                                                              • Table C-10 Results for mtDNAnDNA
                                                                                                                                                                                • 5 References
                                                                                                                                                                                  • Appendix D Performance in Commercial qPCR Chemistries
                                                                                                                                                                                    • Table D-1 Summary [DNA] Results for the Commercial qPCR Assays
Page 8: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

i

Table of Figures Figure 1 Sales History of SRM 2372 3 Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays 6 Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays 7 Figure 4 Plate Layout Design for Homogeneity Measurements 11 Figure 5 Plate Layout Design for Stability Measurements 11 Figure 6 Basic Plate Layout Design for Certification Measurements 12 Figure 7 Plate Layout Design for Performance Assessments 14 Figure 8 Absorbance Spectra of SRM 2372a Components A B and C 16 Figure 9 ddPCR Stability Measurements as Function of Time 18 Figure 10 Comparison of Assay Performance 20 Figure 11 mtDNAnDNA for Components A B and C 24 Figure 12 qPCR Autosomal Quantitation Performance Summary 26 Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary 26 Figure 14 Absorbance Spectra of SRM 2372 Components A and B 27 Figure 15 Comparison of Snap-Cooling to Slow-Cooling 32 Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures 33 Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a 35 Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution 37 Figure 19 Comparison of Assays 38 Figure 20 Evaluation of Limiting Ratio 40 Figure 21 Degrees of Equivalence 42 Figure C-1 Homogeneity Assessment 56 Figure D-1 Results for the Innogenomics Commercial qPCR assays 69 Figure D-2 Results for the Thermo Fisher Commercial qPCR assays 70 Figure D-3 Results for the Qiagen Commercial qPCR assays 70 Figure D-4 Results for the Promega Commercial qPCR assays 71

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

1

Purpose and Description Standard Reference Material (SRM) 2372a is intended for use in the value assignment of human genomic deoxyribonucleic acid (DNA) forensic quantitation materials A unit of SRM 2372a consists of three well-characterized human genomic DNA materials solubilized in 10 mmolL 2-amino-2-(hydroxymethyl)-13 propanediol hydrochloride (Tris HCl) and 01 mmolL ethylenediaminetetraacetic acid disodium salt (disodium EDTA) pH 80 buffer (TE-4) The three component genomic DNA materials labeled A B and C are respectively derived from a single male donor a single female donor and 13 mixture of a male and a female donor A unit of the SRM consists of one 05 mL tube of each component each tube containing approximately 55 microL of DNA solution Each of these tubes is labeled and is sealed with a color-coded screw cap NIST is guided by and adheres to the ethical principles set forth in the Belmont Report1 SRM 2372a was developed after an appropriate human subject research determination by the NIST Human Subjects Protections Office

Warning SRM 2372a is a Human Source Material The three SRM 2372a components are human genomic DNA extracts prepared at NIST from buffy coat white blood cells from single-source anonymous donors All buffy coats were purchased from Interstate Blood Bank (Memphis TN) The supplier reported that each donor unit used in the preparation of this product was tested by FDA-licensed tests and found to be negative for human immunodeficiency virus (HIV-1 RNA and Anti-HIV 12) hepatitis B (HBV DNA and HBsAg) hepatitis C (HCV RNA and Anti-HCV) West Nile virus (WNV RNA) and syphilis Additionally each donor was tested according to FDA guidelines for Trypanosoma cruzi (Chagas) and found to be negativenon-reactive However no known test method can offer complete assurance that infectious agents are absent from this material Accordingly this human blood-based product should be handled at the Biosafety Level 12 SRM 2372a components and derived solutions should be disposed of in accordance with local state and federal regulations

Storage and Use Until required for use SRM 2372a should be stored in the dark between 2 degC and 8 degC The SRM 2372a component tubes should be mixed briefly and centrifuged (without opening the tube cap) prior to removing sample aliquots for analysis For the certified values to be applicable materials should be withdrawn immediately after opening the tubes and processed without delay Certified values do not apply to any material remaining in recapped tubes The certification only applies to the initial use and the same results are not guaranteed if the remaining material is used later

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

2

History and Background SRM 2372a is the second member of the SRM 2372 Human DNA Quantitation Standard series The supply of the initial material SRM 2372 was exhausted in May 2017 SRM 2372a is derived from commercially obtained buffy coat cells from different donors than those used in SRM 2372 A series of interlaboratory studies coordinated by what was then the Biotechnology Division of the National Institute of Standards and Technology (NIST) during the 1990s demonstrated that the DNA quantitation technologies then in use by the forensic human identification community adequately met the needs of the DNA typing tools then available34 However our 2004 study demonstrated that the increasingly sensitive and informative polymerase chain reaction (PCR)-based multiplex typing tools required more reliable determinations of DNA input quantity for best performance5 Quantitative PCR (qPCR) methods for DNA quantitation require calibration to an external standard of defined DNA concentration ([DNA]) expressed in units of nanograms of DNA per microliter of extract solution The true [DNA] of a commercial secondary standard may vary from its nominal concentration depending on supplier and lot number To enable providers of secondary standards to more reliably evaluate their materials and forensic testing laboratories to validate their use of commercial quantitation methods and supplies in 2007 we delivered the human DNA calibration standard SRM 2372 Human DNA Quantitation Standard6 The quantity of DNA in each of the three components of SRM 2372 was originally assigned based on the absorbance of double-stranded DNA (dsDNA) at 260 nm7 However by 2012 the absorbance properties of the materials had increased beyond the certified uncertainty limits due to slow but progressive changes in DNA tertiary structure Sales of the SRM were restricted until the DNA quantities were reassigned based on measurements of single-stranded DNA (ssDNA) in early 20138-10 Figure 1 displays the sales history of this relatively high-selling SRM While SRM 2372 was fit for the practical purpose of harmonizing DNA quantitation measurements to an accessible and reproducible reference material the conversion of both dsDNA and ssDNA absorbance to mass [DNA] is through conventional proportionality constants that are not metrologically traceable to the modern reference system for measurements the International System of Units (SI) Responding to requests from the forensic and clinical communities for calibration standards that provide traceability to the SI we and others have established droplet digital polymerase chain reaction (ddPCR) as a potentially primary measurement process that can provide traceable results11-18 The copy numbers of human nuclear DNA (nDNA) in the reaction mixture per partition symbolized as λ were determined by ddPCR using ten PCR assays optimized for use with our ddPCR platform These assays target loci on eight chromosomes with three assays on widely separated locations on one chromosome (see Table 1)

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

3

Figure 1 Sales History of SRM 2372

Each small black dot represents the sale of one unit of SRM 2372 The red lines summarize the average rate of sales as originally certified (2007 to 2012) and after re-certification (2013 to 2017) The magenta labels state the average yearly sales rate during these two periods

The nDNA copy number concentration human haploid genome equivalents (HHGE) per nanoliter component solution and symbolized here as λacute is calculated as

copy number [DNA] = λ = λ(VF) HHGEnL

where V is the average droplet volume and F is the total dilution factor the volume fraction of the component solution in the ddPCR reaction mixture The manufacturer of the ddPCR platform used a Bio-Rad QX200 Droplet Digital PCR System (Hercules CA) system does not provide metrologically traceable droplet volume information This lack of information prevented metrologically traceable conversion of the number per partition measurements to the desired number per reaction volume units In consequence NIST developed a measurement method that provides metrologically traceable droplet volumes19 We show elsewhere that the mass [DNA] nanograms of human DNA per microliter component can be traceably estimated as

mass [DNA] = 3031timesλacute ngmicroL

where the constant is the product of the number of nucleotide basepairs (bp) in the human reference genome the average bp molecular weight Avogadrorsquos number of bp per mole and

2372

2372(Recertif ied)

050

010

00

1713 unitsyear

1416 unitsyear

2007 2009 2011 2013 2015 2017

Uni

ts S

old

Calendar Year

Human DNA Quantitation StandardSRM 2372

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

4

unit scaling factors18 The three components of SRM 2372a are certified for both copy number and mass [DNA] Mitochondrial DNA (mtDNA) analysis within the forensic community is useful when the nDNA may be degraded or is in low-to-undetectable quantity The challenges of mtDNA testing include contamination of the standard due to routine laboratory use degradation of plasmids or synthetic oligomers or improper commutability of the standard relative to casework samples Commonly cell line materials are used as standards for qPCR with human genomic DNA which may lead to an incorrect mtDNA-to-nDNA ratio (mtDNAnDNA) for unknown samples Non-certified values of mtDNAnDNA for the three components are provided to bridge the current gap between well characterized mitochondrial DNA quantification assays and the availability of commercial standards With the understanding that the forensic DNA community primarily uses qPCR as a method for DNA quantification the performance of the three SRM 2372a materials was evaluated in eleven commercial qPCR chemistries currently used employed within the forensic DNA community

Experimental Methods Production and value-assignment of any certified reference material (CRM) is a complex process In addition to acquiring processing and packaging the materials included in the CRM the materials must be sufficiently homogenous across the units stable over time within its packaging and well-characterized to deliver results that are fit-for-purpose within the intended measurement community20 The process has been particularly complicated for SRM 2372a because we needed to determine the metrological properties of the required ddPCR measurement systems before we could trust the results they provide The following sections describe the methods and processes used to produce and characterize the SRM 2372a materials including sample preparation SRM unit production PCR assays ddPCR measurement processes spectrophotometric absorbance measurements homogeneity testing stability testing ddPCR nDNA and mtDNAnDNA measurements and performance in commercial qPCR assessments Many of these processes were mutually dependent therefore the sections are presented in a ldquodefinitionalrdquo order that sequentially introduces terminology Polymerase Chain Reaction Assays

Ten PCR assays were developed for characterization of SRM 2372a Each assay was confirmed to target one locus per HHGE using the National Center for Biotechnology Informationrsquos BLASTblastn system21 Primers were purchased from Eurofins Operon (Huntsville AL) TaqMan probes labeled with 6-carboxyfluorescein (FAM) were purchased from Thermo Fisher (Waltham MA) FAM-labeled Blackhole Quencher and Blackhole Quencher+ probes were purchased from LGC Biosearch Technologies (Novato CA) Table 1 details the ten nDNA assays used to certify copy number and mass [DNA] of the SRM 2372a components FAM-labeled TaqMan probes were purchased from Thermo Fisher

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

5

and used in monoplex reactions VIC (a proprietary dye)-labeled probes were purchased from the same source and were used when nDNA and mtDNA were analyzed in the same reaction mixture Figure 2 displays exemplar droplet patterns for the assays All assays were developed at NIST and have been optimized for ddPCR18 Table 2 details three mitochondrial DNA (mtDNA) assays used to estimate mtDNAnDNA in the SRM 2372a components Figure 3 displays exemplar droplet patterns for the assays These assays were optimized for ddPCR from real-time quantitative PCR (qPCR) assays from the literature All primers probes were diluted with TE-4 buffer purchased from Affymetrix-USB Products (Affymetrix Inc Cleveland OH) to a nominal 5 micromol working solution for all single-plex assays (individual nDNA and mtDNA assays) For duplex assays (combined nDNA and mtDNA) primer and probe working solutions were diluted to a nominal 10 micromol

Table 1 NIST-Developed Human Nuclear DNA Assays

Assay Target

Chromosome Band Accession Primers and Probe a

Amplicon Length bp

NEIF Gene EIF5B

Chr 2 p111-q111 NC_00000212

F gccaaacttcagccttctcttc R ctctggcaacatttcacactaca PB+ tcatgcagttgtcagaagctg

67

2PR4 Gene RPS27A

Chr 2 p16 NC_00000212

F cgggtttgggttcaggtctt R tgctacaatgaaaacattcagaagtct PB tttgtctaccacttgcaaagctggccttt

97

POTP STR TPOX

Chr 2 p253 NC_00000212

F ccaccttcctctgcttcacttt R acatgggtttttgcctttgg PT caccaactgaaatatg

60

NR4Q Gene DCK

Chr 4 q133-q211 NC_00000412

F tggtgggaatgttcttcagatga R tcgactgagacaggcatatgtt PB+ tgtatgagaaacctgaacgatggt

83

D5 STR D5S2500

Chr 5 q112 NC_00000510

F ttcatacaggcaagcaatgcat R cttaaagggtaaatgtttgcagtaatagat PT ataatatcagggtaaacaggg

75

ND6 STR D6S474

Chr 6 q21-22 NC_00000612

F gcatggctgagtctaaagttcaaag R gcagcctcagggttctcaa PB+ cccagaaccaaggaagatggt

82

D9 STR D9S2157

Chr 9 q342 NC_00000912

F ggctttgctgggtactgctt R ggaccacagcacatcagtcact PT cagggcacatgaat

60

HBB1 Gene HBB

Chr 11 p155 NC_00001110

F gctgagggtttgaagtccaactc R ggtctaagtgatgacagccgtacct PT agccagtgccagaagagccaagga

76

ND14 STR D14S1434

Chr 14 q3213C NC_0000149

F tccaccactgggttctatagttc R ggctgggaagtcccacaatc PB+ tcagactgaatcacaccatcag

109

22C3 Gene PMM1

Chr 22 q132 NC_00002210

F cccctaagaggtctgttgtgttg R aggtctggtggcttctccaat PB caaatcacctgaggtcaaggccagaaca

78

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

6

a F Forward primer R Reverse primer PB Blackhole quencher probe (FAM labeled) PB+ Blackhole Plus quencher probe (FAM labeled) PT TaqMan MGB probe (FAM labeled)

Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays

Table 2 NIST-Optimized Human Mitochondrial DNA Assays

Assay Location Primers and Probe a Amplicon Length bp Ref

mtND1 34853504 F ccctaaaacccgccacatct

69 22 35333553 R gagcgatggtgagagctaaggt 35063522 PT ccatcaccctctacatc

mtBatz 84468473 F aatattaaacacaaactaccacctacct

79 23 85038524 R tggttctcagggtttgttataa 84758491 PT cctcaccaaagcccata

mtKav 1318813308 F ggcatcaaccaaccacaccta

105 24 1336913392 R attgttaaggttgtggatgatgga 1331013325 PT cattcctgcacatctg

a F Forward primer R Reverse primer PT TaqMan MGB probe (FAM labeled)

2PR4

POTP

NEI

F

R4Q

5

D5

ND6

D9

HBB1

ND1

4

22C3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

7

Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays

Droplet Digital Polymerase Chain Reaction Measurements

Nuclear DNA (nDNA) All nDNA ddPCR measurements used in this study followed the same measurement protocol

bull Droplets were generated on the Auto Droplet Generator using the Bio-Rad ldquoddPCR ldquoSupermix for Probes (No dUTP)rdquo (Bio-Rad cat 186-3024 Control 64079080) and ldquoDroplet Generation Oil for Probesrdquo (Bio-Rad cat 186-4110 Control 64046051) Note desoxyuridine 5-triphosphatase (dUTP) is an enzyme sometimes included in PCR reaction mixtures to minimize PCR carryover contamination

bull Table 3 details the composition of the ddPCR mastermix setup for a given assay and sample where ldquomastermix refers to reaction mixture of manufacturerrsquos proprietary PCR reagent ldquoSupermixrdquo forward and reverse PCR primers for a given assay probe for that assay the sample DNA and sometimes additional water to produce the desired value dilution factor F This setup was used for all assays and samples

bull Non-Template Controls (NTCs) for each assay were included in every analysis bull Droplets were thermal cycled on a ProFlex thermal cycler (Applied Biosystems) for

95 degC for 10 min followed by 60 cycles of 94 degC for 05 min and 61 degC for 1 min then 98 degC for 10 min before a 4 degC hold until the plate was removed

bull Droplets were read on the QX200 Droplet Reader and analysis was performed using the QuantaSoft Analysis Software version 1740917

mtND1 mtBatz mtKav

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

8

bull The numbers of positive and negative droplets at the end of 60 cycles were determined and were exported into a spreadsheet for further analysis Assay-specific intensity thresholds were determined by visual inspection for each assay for each measurement session

Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays

Concentration nDNA Mastermix Microliter

per Reaction 2X ldquoSupermix for Probes (no dUTPs)rdquo 125

5 μmolL Forward Primer 1875 5 μmolL Reverse Primer 1875 5 μmolL Probe (FAM) 125

PCR Grade Water 50 14 Diluted Sample 25 Total Volume 250

Mitochondrial DNA (mtDNA) All the mtDNA measurements used in this study followed the nDNA protocol described above but with different mastermix setups The duplex mtDNA measurements used the setup described in Table 4 The monoplex mtDNA measurements used the setup described in Table 5

Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 125

10 μmolL Forward Primer 09375 10 μmolL Reverse Primer 09375 10 μmolL Probe (VIC) 0625

PCR Grade Water 95 nDNA Mastermix from Table 3 05 Total Volume 250

Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 120

10 μmolL Forward Primer 09 10 μmolL Reverse Primer 09 10 μmolL Probe (FAM) 06

PCR Grade Water 72 Diluted Sample (See Table 7) 24

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

9

Total Volume 240 Sample Preparation

The buffy coat white blood cells used to prepare SRM 2372a materials were purchased from Interstate Blood Bank These materials were from four single-source anonymous individuals supplied to us labeled as Female1 Female2 Male1 and Male2 In our laboratories each donor buffy coat bag was aliquoted (5 mL per aliquot) into sterile 50 mL conical tubes and stored at 4 degC prior to extraction The modified salt-out manual extraction protocol described in Appendix A was performed for each aliquot with rehydration of the DNA in TE-4 buffer From 150 mL to 250 mL of solubilized DNA for each component was accumulated through the extraction of the 5 mL aliquots The DNA was stored in perfluoroalkoxy polymer (PFA) containers that had been bleach sterilized thoroughly rinsed with deionized water rinsed with ethanol and air-dried The minimum volume for production of 2000 units was 120 mL Preliminary estimates of the mass [DNA] of the re-solubilized components were obtained using a Nanodrop microvolume spectrophotometer Dilutions of the component materials were made until the absorbance at 260 nm was approximately 1 corresponding to a nominal mass [DNA] of 50 ngμL of dsDNA7 After addition of the TE-4 buffer the containers were placed on an orbital shaker and gently rotated for several hours prior to refrigerated storage at 4 ordmC The four single-source components were initially screened with the nDNA ddPCR measurement process using four nDNA assays 2PR4 NR4Q ND6 and 22C3 Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate The gravimetric 1rarr4 dilution was prepared by weighing an empty PFA vial and recording the weight then pipetting 30 microL TE-4 into the vial and recording the weight then pipetting 10 microL DNA and recording the final weight The dilution factor was calculated in a spreadsheet These screening measurements verified that the individual component materials were fit-for-purpose The four single-source components were evaluated with the nDNA ddPCR measurement process using the ten nDNA assays Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate Quantitative information was derived from the average of the ten nDNA assays to prepare the component C mixture Thirty milliliters of sample Male2 (estimated at 452 ngμL mass [DNA]) were combined with 80 mL of sample Female2 (estimated at 507 ngμL mass [DNA]) and 10 mL of TE-4 buffer to create a 1 male nDNA to 3 female nDNA producing a 14 malehuman mixture After addition of the TE-4 buffer the component C container was gently rotated on the orbital shaker for several hours prior to refrigerated storage at 4 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

10

Absorbance Measurements

In February 2017 absorbance spectra of the recently prepared fully diluted nominally dsDNA and ssDNA solutions of the three components were acquired with a BioCary 100 spectrophotometer using the measurement protocol described in Appendix B In December 2017 additional spectra were acquired for the remaining bulk solution of components A and B and for SRM 2372 components A and B These solutions had been stored at 4 degC in well-sealed PFA containers There was no remaining bulk component C solution SRM Unit Production

In the morning of the day a component was transferred into the sterile component tubes (Sarstedt Biosphere SC micro tube-05 mL Newton NC) used to package the SRM the container holding the diluted component was removed from the refrigerator and placed on a slow orbital shaker for 2 h The tubes were opened and placed under the laminar flow hood in rows of 5 within 80-hole tube racks for a total of 30 tubes per rack Tubes were placed in every second row of the racks to facilitate filling with an Eppendorf Repeater Xstream pipette (Eppendorf North America Inc Hauppauge NY) fitted with a 1-mL positive displacement tip set to dispense 55 microL per tube The component container was removed from the shaker a magnetic stir-bar was added the container was placed on a magnetic stir plate in the laminar flow hood and the material was stirred gently Filling proceeded until 2200 vials were filled with pipet tip replacement after filling every 200 tubes The filling process consisted of two individuals manually filling each vial within a biosafety cabinet (one physically pipetting each sample the other verifying that each tube was filled) and four individuals tightly closing the lids to each tube on a sterilized and covered bench When all tubes in a rack were filled the rack was moved out of the hood the tubes checked for proper filling their lids closed and the tubes transferred to 100-unit pre-labeled storage boxes As the 100-unit boxes were filled they were transferred to the labeling room where SRM labels were applied by hand and colored inserts were added to the lids red for component A white for component B and blue for component C After all tubes for a given component were labeled the 22 boxes were placed in a locked refrigerator Filling of each of the components took 60 min to 90 min All units were transferred to the refrigerator within 4 hours of beginning the filling process All units of the three materials were allowed to equilibrate for a minimum of two weeks prior to beginning homogeneity analysis Component Homogeneity

The ddPCR measurement protocol described above was used to assess the nDNA homogeneity of the three components One tube from the approximate center of each of the 22 storage boxes was assayed using the NEIF and ND6 assays (see Table 1) A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Three technical replicates were obtained for every tube A total

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

11

of six plates were used two plates per component one with the NEIF assay and one with the ND6 assay Figure 4 details the layout of the samples

Figure 4 Plate Layout Design for Homogeneity Measurements

Component Stability

Stability of the materials was evaluated at 4 degC room temperature (RT asymp 22 degC) and 37 degC Tubes from the homogeneity study were divided among the three temperatures for the stability study Additionally two never-opened tubes (one from box 8 and one from box 12 of each component) were held at each temperature Two tubes at each temperature for each component were opened sampled and resealed 4 times over the course of 8 weeks Measurements were made at week 1 week 3 week 5 and week 8 The measurement on week 8 was from an unopened tube (from box 8) for each of the components at all temperatures The data from the homogeneity study was used as timepoint 0 in the stability study anchored to 4 degC by the storage recommendations A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used for both the NEIF and ND6 assays All temperatures components and both assays fit onto one plate for timepoints 1 to 5 Twelve NTCs placed in rows G and H were included for each assay Figure 5 details the sample layout

Figure 5 Plate Layout Design for Stability Measurements

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

ND6 ND6 ND6

NEIFNEIFNEIFComponent A

Component A

Component B

Component B

Component C

Component C

1 2 3 4 5 6 7 8 9 10 11 12A A-7 A-7 B-7 B-7 C-7 C-7 A-7 A-7 B-7 B-7 C-7 C-7B A-10 A-10 B-10 B-10 C-10 C-10 A-10 A-10 B-10 B-10 C-10 C-10C A-9 A-9 B-9 B-9 C-9 C-9 A-9 A-9 B-9 B-9 C-9 C-9D A-12 A-12 B-12 B-12 C-12 C-12 A-12 A-12 B-12 B-12 C-12 C-12E A-8 A-8 B-8 B-8 C-8 C-8 A-8 A-8 B-8 B-8 C-8 C-8F A-11 A-11 B-11 B-11 C-11 C-11 A-11 A-11 B-11 B-11 C-11 C-11G NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTCH NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

NEIF

4C

RT

37C

ND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

12

Component Certification Measurements

After establishing homogeneity and stability measurements intended to be used to certify nDNA values and inform mtDNA values were started using randomly selected SRM tubes from the boxes listed in Table 7 The ten nDNA assays described in Table 1 and two of the three mtDNA assays described in Table 2 were run simultaneously on the same sample plate A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used in each of the six measurement campaigns

Table 6 Tubes used for Certification Measurements

Component Plate A B C

dd20170711 Box 1 Box 1 Box 1 dd20170712a Box 10 tube a Box 10 tube a Box 10 tube a dd20170712b Box 10 tube b Box 10 tube b Box 10 tube b dd20170713 Box 20 Box 20 Box 20 dd20170714 Box 5 Box 5 Box 5 dd20170801 Box 5 Box 5 Box 5

Every sample was analyzed in duplicate with every assay in all six plates Component A was always located in rows B and C component B in rows D and E component C in rows F and G and NTCs in rows A and H Figure 6 displays the basic layout of the ten nDNA and two mtDNA assays In addition to the plate column assignment scheme shown the three assay groupings were also ordered as Group 2 Group 3 Group 1 and Group 3 Group 1 Group 2 to address potential positional bias

Group 1 Group 2 Group 3 2PR4 POTP NEIF NR4Q D5 ND6 D9 HBB1 ND14 22C3 mtND1 mtBatz 1 2 3 4 5 6 7 8 9 10 11 12 A NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Comp A B Comp A C Comp B D Comp B E Comp C F Comp C G

H NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Figure 6 Basic Plate Layout Design for Certification Measurements

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

13

Mitochondrial DNA Measurements

A ldquoduplexed reactionrdquo was used to determine mtDNAnDNA where 1 μL of one of the nDNA mastermix was added to 49 μL of the mtDNA mastermix resulting in a 1rarr50 volumetric dilution of the nDNA The nDNA was targeted with a FAM-labeled probe while the mtDNA was targeted with a VIC labeled probe The primer and probe concentrations were increased from 5 micromol to 10 micromol in the duplexed reactions Additional testing of the mtDNA assays were performed in monoplex with FAM-labeled probes to determine the additional sources of bias between sampling assay setup and dilution factors Table 7 outlines the different methods examined to determine mtDNAnDNA for the three components

Table 7 Run Parameters for Mitochondrial DNA Droplet

Generation nDNA mtDNA

Plate Component Assay Dilution Reps Assay Dilution Reps dd20170824 A manualDG HBB1 1rarr4 3 mtND1 1rarr200 9 dd20170825 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170829 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170831 AB autoDG HBB1 1rarr3 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr6 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr12 2 mtND1 1rarr10rarr100 3 dd20170907 ABC manualDG HBB1 1rarr3 2 mtBatz 1rarr10rarr100 3 dd20170913 AB manualDG NEIF 1rarr4 3 mtND1 mtBatz mtKav 1rarr10rarr100 3 dd20170913 C manualDG NEIF 1rarr4 3 mtBatz mtKav 1rarr10rarr100 3 Performance in Commercial qPCR Chemistries

Table 8 lists the 11 qPCR chemistries that were commercially available in late 2017 All three of the SRM 2372a components were evaluated with these chemistries using component A of SRM 2372 as the reference standard For each component of SRM 2372a four tubes were pulled from box 22 and combined into one tube A 1rarr5 dilution of all samples was prepared followed by three serial 1rarr2 dilutions A five-point serial dilution of each SRM 2372a component was amplified in triplicate for each chemistry This consisted of the neat material the 1rarr5 dilution then the three serial 1rarr2 dilutions A six-point serial dilution of SRM 2372 Component A was also created for the standard curve for the commercial qPCR chemistries This consisted of the neat material followed by a 1rarr5 dilution then followed by four 1rarr2 dilutions serially The six-point serial dilution series of SRM 2372 Component A was also amplified in triplicate These commercial chemistries were amplified in individual plates on an Applied Biosystems 7500 HID Real-Time PCR Instrument (Foster City CA) following the manufacturersrsquo recommended protocols for reaction mix setup and cycling conditions Figure 7 details the plate layout for the qPCR plates Data was analyzed with the HID Real-Time PCR Analysis Software v12 and further analyzed in a spreadsheet

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

14

Table 8 Commercial qPCR Chemistries Tested with SRM 2372a

Manufacturer Commercial Kit Target Codea

Innogenomics

InnoQuant Long IQh Short IQd

InnoQuantHY Long IQHYh Short IQHYd Y IQHYy

Thermo Fisher Quantifiler Duo Human QFDh Male QFDy

Quantifiler HP Large QFHPh Small QFHPd

Quantifiler Human Human QFHh

Quantifiler Trio Large QFTh Small QFTd Y QFTy

Qiagen

Quantiplex Autosomal QPh

Quantiplex Hyres Autosomal QPHh Y QPHy

Quantiplex Pro Human QPPh Degradation QPPd Y QPPy

Promega

Plexor Autosomal Ph Y Py

PowerQuant Autosomal PQh Degradation PQd Y PQy

a Assays are coded with a short acronym derived from the kit name followed by the following codes for the type of target ldquohrdquo autosomal non-degradation described as ldquoAutosomalrdquo ldquoHumanrdquo or ldquoLongrdquo ldquodrdquo autosomal degradation described as ldquoDegradationrdquo or ldquoShortrdquo ldquoyrdquo Y-chromosome described as ldquoMalerdquo or ldquoYrdquo

1 2 3 4 5 6 7 8 9 10 11 12

A

B 2372-A_57 2372-A_57 2372-A_57 2372a-neat_A

2372a-neat_A

2372a-neat_A

2372a-neat_B

2372a-neat_B

2372a-neat_B

2372a-neat_C

2372a-neat_C

2372a-neat_C

C 2372-A_11

4 2372-A_11

4 2372-A_11

4 2372a-15_A

2372a-15_A

2372a-15_A

2372a-15_B

2372a-15_B

2372a-15_B

2372a-15_C

2372a-15_C

2372a-15_C

D 2372-A_5

7 2372-A_5

7 2372-A_5

7 2372a-110_A

2372a-110_A

2372a-110_A

2372a-110_B

2372a-110_B

2372a-110_B

2372a-110_C

2372a-110_C

2372a-110_C

E 2372-A_2

85 2372-A_2

85 2372-A_2

85 2372a-120_A

2372a-120_A

2372a-120_A

2372a-120_B

2372a-120_B

2372a-120_B

2372a-120_C

2372a-120_C

2372a-120_C

F 2372-A_1

425 2372-A_1

425 2372-A_1

425 2372a-140_A

2372a-140_A

2372a-140_A

2372a-140_B

2372a-140_B

2372a-140_B

2372a-140_C

2372a-140_C

2372a-140_C

G 2372-A_0

7125 2372-A_0

7125 2372-A_0

7125

H NTC NTC NTC

Figure 7 Plate Layout Design for Performance Assessments

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

15

Measurement Results After preparing the SRM 2372a bulk solutions their approximate 50 ngmicroL mass [DNA] were confirmed spectrophotometrically The bulk solutions were then packaged into individual tubes The materials contained in the tubes were characterized for homogeneity stability nDNA copy number [DNA] and mtDNAnDNA ratio The following sections detail the results of these measurements Absorbance Spectrophotometry

By widely accepted convention an aqueous DNA solution having an absorbance of 10 at 260 nm corresponds to a mass [DNA] of 50 ngmicroL for dsDNA and (37 to 40) ngmicroL for ssDNA7 Absorbance at (230 270 280 and 330) nm are traditionally used in the assessment of DNA quality725 Due to minor baseline shifts during acquisition of some spectra it was necessary to bias-adjusted all spectra to have absorbance of 00 at 330 nm Table 9 lists the resulting bias-adjusted values for selected wavelengths and the conventional mass [DNA] conversion of the absorbance at 260 nm The close agreement between the mass concentration values for the nominally dsDNA and the converted-to-ssDNA spectra confirm that the bulk solutions were correctly prepared to be asymp 50 ngμL dsDNA The slight increase in absorbance over the 10 months between acquiring the spectra suggests that like SRM 2372 the tertiary structure of the dsDNA in the SRM 2372a materials may be changing slowly with time Figure 8 displays the spectra

Table 9 Absorbance at Selected Wavelengths

Wavelength nm Form Date Component 230 260 270 280 ngμL dsDNA 2142017 A 0394 0931 0758 0488 466

B 0480 1112 0905 0583 556 C 0451 1006 0815 0524 503

ssDNA 2142017 A 0390 0614 0540 0316 454 B 0457 0723 0634 0372 535 C 0411 0641 0563 0330 474

dsDNA 12142017 A 0402 0956 0775 0503 478 B 0484 1128 0914 0589 564

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

16

Figure 8 Absorbance Spectra of SRM 2372a Components A B and C

dsDNA is denatured to ssDNA by diluting with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to DNA mass concentration

Droplet Volume

Droplet volumes for the two lots of ldquoSupermixrdquo used in the studies described in this report were determined using NISTrsquos Special Test Method 11050S-D as described in Dagata et al19 The lot used for most of the ddPCR analyses produced (07349 plusmn 00085) nL droplets where the ldquoplusmnrdquo value is the combined standard uncertainty The lot used for analyses performed after October 11 2017 produced (07376 plusmn 00085) nL droplets NIST measurements have demonstrated that droplet volumes are not influenced by the presence or absence of DNA in the sample solution Measurements repeated over time on an earlier lot of the Supermix product used in these studies suggest that droplet volumes remain constant over the shelf-life of the lot However droplet volumes for different lots of this product as well as for different Supermix products have been observed to differ by several percent We intend to document these measurements in a future report Other researchers have demonstrated that non-constant droplet volume distributions may contribute to ddPCR measurement imprecision and bias2627 However these effects become significant only with distributions wider than observed using Special Test Method 11050S-D19 and at copydroplet values larger than the λ le 05 copydroplet typically used in our measurements Certification Measurements

Homogeneity Results Table 10 lists the λacute nDNA copy number values adjusted for both droplet volume and sample dilution from the homogeneity measurements The data was analyzed using the three-level Gaussian hierarchical ANOVA model described in Appendix C Table 11 lists the values of the within- and between-tube variance components expressed as percent relative standard deviations At the 95 confidence level all components are homogeneous

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372a-A2141721417-NaOH121417

220 260 300 340

Wavelength nm

2372a-B2141721417-NaOH121517

220 260 300 340

Wavelength nm

2372a-C21417

21417-NaOH

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

17

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 NEIF A-1-01 1632 1529 1511 B-1-01 1790 1909 1885 C-1-01 1493 1385 1510 ND6 1616 1640 1582 NAe 1870 1860 1565 1562 1466

NEIF A-1-02 1531 1525 1549 B-1-02 1712 1719 1704 C-1-02 1516 1513 1577 ND6 1615 1589 1601 1775 1718 1756 1490 1547 1570

NEIF A-1-03 1524 1540 1533 B-1-03 1693 1755 1749 C-1-03 1451 1528 1506 ND6 1582 1152 1555 1824 1483 1896 1485 1505 1471

NEIF A-1-04 1454 1458 1427 B-1-04 1754 1745 1739 C-1-04 1440 1486 1520 ND6 1627 1605 1607 1670 1632 1814 1529 1508 1506

NEIF A-1-05 1465 1540 1504 B-1-05 1757 1687 1735 C-1-05 1523 1472 1483 ND6 1607 1633 1586 1683 1759 1703 1513 1520 1491

NEIF A-1-06 1512 1491 1491 B-1-06 1697 1709 1714 C-1-06 1437 1493 1522 ND6 1511 1483 1524 1693 1723 1722 1492 1481 1461

NEIF A-1-07 1418 1480 1455 B-1-07 1722 1625 1703 C-1-07 1485 1477 1463 ND6 1479 1557 1656 1690 1684 1688 1475 1547 1485

NEIF A-1-08 1496 1560 1501 B-1-08 1884 1802 1772 C-1-08 1433 1405 1478 ND6 1514 1615 1540 1864 1887 1821 1497 1501 1495

NEIF A-1-09 1588 1581 1637 B-1-09 1713 1734 1727 C-1-09 1463 1499 1505 ND6 1474 1438 1454 1709 1756 1758 1503 1491 1516

NEIF A-1-10 1574 1513 1570 B-1-10 1759 1809 1738 C-1-10 1534 1458 1454 ND6 1594 1713 1669 1764 1816 1787 1448 1465 1474

NEIF A-1-11 1557 1508 1633 B-1-11 1749 1782 1740 C-1-11 1533 1514 1555 ND6 1571 1611 1631 1741 1792 1784 1488 1470 1483

NEIF A-1-12 1500 1476 1515 B-1-12 1748 1712 1718 C-1-12 1450 1525 1497 ND6 1626 1586 1608 1797 1614 1790 1496 1483 1519

NEIF A-1-13 1672 1634 1804 B-1-13 1632 1668 1701 C-1-13 1537 1527 1515 ND6 1490 1516 1497 1626 1566 1681 1499 1479 1475

NEIF A-1-14 1551 1568 1538 B-1-14 1752 1535 1673 C-1-14 1515 1490 1494 ND6 1617 1614 1639 1724 1741 1685 1469 1480 1466

NEIF A-1-15b 1793 1722 1733 B-1-15 1871 1835 1887 C-1-15 1511 1568 1594 ND6 1818 1753 1762 1770 1752 1798 1515 1543 1565

NEIF A-1-16 1573 1520 1617 B-1-16 1688 1725 1743 C-1-16 1449 1452 1471 ND6 1634 1627 1575 1744 1697 1710 1543 1448 1491

NEIF A-1-17 1516 1516 1628 B-1-17 1789 1769 1825 C-1-17 1388 1421 1394 ND6 1613 1567 1567 1716 1762 1638 1452 NAe 1435

NEIF A-1-18 1612 1535 1672 B-1-18 1826 1780 1849 C-1-18 1475 1455 1445 ND6 1529 1495 1528 1805 1801 1931 1458 1531 1459

NEIF A-1-19 1598 1514 1596 B-1-19 1751 1759 1880 C-1-19 1468 1525 1446 ND6 1571 1535 1511 1827 1766 1786 1435 1478 1421

NEIF A-1-20 1575 1568 1586 B-1-20 1738 1678 1768 C-1-20 1549 1531 1499 ND6 1652 1545 1627 1752 1747 1768 1499 1465 1468

NEIF A-1-21 1658 1611 1578 B-1-21 1742 1836 1713 C-1-21 1530 1440 1484 ND6 1647 1607 1612 1746 1744 1764 1499 1454 NAe

NEIF A-1-22 1547 1528 1532 B-1-22 1761 1743 1660 C-1-22 1442 1487 1436 ND6 1537 1513 1536 1666 1692 1819 1589 1591 1556

NEIF A-1-15c 1456 1448 1523 ND6 1525 1402 1436

NEIF A-2-15d 1679 1599 1674 ND6 1581 1550 1584

a Tube coded as Component-Tube-Box b These results were identified as a probable error in the preparation or volumetric delivery of the sample

solution into the microfluidic device and were not used in the homogeneity evaluations c Repeat assessment of the solution in sample A-1-15 d Results for a second tube from box 15 analyzed at the same time as the repeat assessment of A-1-15 e Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

18

Table 11 Within- and Between-Tube Relative Standard Deviations

Within-Tube Component NEIF ND6 Between-Tube

A 28 40 33 B 27 38 25 C 23 18 19

Stability Results Table 12 lists the λacute volume- and dilution-adjusted nDNA copy number values from the stability measurements Figure 9 summarizes these results The results for the tubes held at 4 degC room temperature (RT asymp 22 degC) and 37 degC are very similar for all three components The absence of between-temperature differences identifies the between-week differences as attributable to variation in the measurement processes rather than sample instability These results indicate that the SRM 2372a genomic copy number is thermally stable from 4 degC to 37 degC over an extended period

Figure 9 ddPCR Stability Measurements as Function of Time

Each dot-and-bar symbol represents the mean λacute and the approximate 95 expanded uncertainty for all results for a given component at a given storage temperature The thick black horizontal lines represent an approximate 95 confidence interval on the λacute values estimated from the homogeneity data

13

14

15

16

17

18

19

0 1 2 3 4 5 6 7 8

λacute H

aplo

id n

DNA

Copi

es μL

Weeks

A

0 1 2 3 4 5 6 7 8

Weeks

B

4 degC 22 degC 37 degC

0 1 2 3 4 5 6 7 8

Weeks

C

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

19

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Week Temp Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3

1 4 degC NEIF A-10 1568 1441 B-10 1666 1653 C-10 1424 1394 ND6 A-10 1474 1494 B-10 1655 1705 C-10 1461 1394

NEIF A-07 1500 1463 B-07 1718 1640 C-07 1446 1452 ND6 A-07 1479 1479 B-07 1708 1686 C-07 1522 1482

RT (22 degC) NEIF A-12 1674 1608 B-12 1805 1777 C-12 1506 1495 ND6 A-12 1522 1507 B-12 1759 1788 C-12 1478 1461 NEIF A-09 1635 1539 B-09 1791 1793 C-09 1411 NAb ND6 A-09 1543 1519 B-09 1661 1682 C-09 1482 1496

37 degC NEIF A-11 1614 1637 B-11 1748 1744 C-11 1515 1464 ND6 A-11 1494 1502 B-11 1740 1697 C-11 1495 1535 NEIF A-08 1592 1614 B-08 1691 1725 C-08 1528 1500 ND6 A-08 1581 1595 B-08 1727 1699 C-08 1458 1483 3 4 degC NEIF A-10 1520 1706 B-10 NAb 1733 C-10 1427 1473 ND6 A-10 1543 1624 B-10 1670 1755 C-10 1443 1364 NEIF A-07 1537 1570 B-07 1549 1565 C-07 1424 1512 ND6 A-07 1517 1479 B-07 1705 1723 C-07 1483 1439

RT (22 degC) NEIF A-12 1521 1607 B-12 1817 1842 C-12 1566 1471 ND6 A-12 1644 1597 B-12 1845 1786 C-12 1504 1511 NEIF A-09 1432 1626 B-09 1825 1856 C-09 1392 1387 ND6 A-09 1539 1528 B-09 1734 1737 C-09 1535 1465

37 degC NEIF A-11 1651 1685 B-11 1759 1779 C-11 1437 1440 ND6 A-11 1680 1594 B-11 1806 1806 C-11 1471 1509 NEIF A-08 1580 1613 B-08 1740 1791 C-08 1582 1590 ND6 A-08 1592 1524 B-08 1769 1694 C-08 1468 1520 5 4 degC NEIF A-10 1486 1440 B-10 1626 1551 C-10 1381 1330 ND6 A-10 1555 1577 B-10 1673 1625 C-10 1547 1395 NEIF A-07 1387 1429 B-07 1723 1787 C-07 1330 1314 ND6 A-07 1538 1522 B-07 1862 1909 C-07 1452 1458

RT (22 degC) NEIF A-12 1601 1631 B-12 1718 1752 C-12 1546 1541 ND6 A-12 1699 1677 B-12 1772 1721 C-12 1490 1471 NEIF A-09 1381 1484 B-09 1675 1694 C-09 1371 1376 ND6 A-09 1625 1615 B-09 1904 1815 C-09 1468 1472

37 degC NEIF A-11 1456 1517 B-11 1720 1704 C-11 1505 1493 ND6 A-11 1650 1658 B-11 1148 1152 C-11 1527 1600 NEIF A-08 1590 1666 B-08 1745 1723 C-08 1469 1540 ND6 A-08 1235 1244 B-08 1809 1749 C-08 1508 1529 8 4 degC NEIF A-08a 1667 1632 1589 B-08a 1880 1884 1940 C-08a 1525 1475 1542 ND6 A-08a 1599 1593 1590 B-08a 1859 1897 1972 C-08a 1550 1502 1520

RT (22 degC) NEIF A-08b 1693 1666 1634 B-08b 1769 1884 1771 C-08b 1595 1557 1545 ND6 A-08b 1644 1620 1647 B-08b 1862 1738 1811 C-08b 1611 1585 1535

37 degC NEIF A-08c 1620 1576 1608 B-08c 1934 1952 1956 C-08c 1595 1600 1569 ND6 A-08c 1576 1553 1532 B-08c 1889 1844 1889 C-08c 1569 1514 1581

a Tube coded as Component-Tube-Box b Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

20

Ten-Assay Certification Results The original intention was to certify the copy number [DNA] of the three components with results from five independently prepared sets of samples analyzed over four days Each sample of the three components was from a new tube from different boxes and was evaluated with 10 nDNA assays The assay positions were alternated across these five ldquo10-assay certificationrdquo plates The λ results for these five plates are systematically somewhat larger than those for the homogeneity and stability data On review we realized that a different workspace and its suite of pipettes had been used to prepare the 1rarr10 dilution of the 1rarr4 diluted components into the reaction mastermix The ldquohighrdquo workstation was used for stability and homogeneity measurements while the ldquolowrdquo workstation was used for certification measurements Gravimetric testing of the pipettes established that the volumes delivered by these alternate pipettes was sufficiently biased to account for the observed λ differences A sixth ldquocertificationrdquo plate was prepared using the original (and now re-verified) pipettes The λ results or this plate agreed well with the homogeneity and stability data Table 13 lists the λacute volume- and dilution-adjusted nDNA copy number values for all six of the ten-assay datasets Figure 10 displays the extent of agreement among the ten assays While there are small-but-consistent differences between the assays there is only a 4 difference between the assays providing smallest and largest λ

Figure 10 Comparison of Assay Performance

The blue dot-and-bars represent the mean and its combined standard uncertainty for the A B and C components for each assay relative to the over-all grand mean for each component The black dot-and-bars represent combination of three component results for each assay The solid horizontal line represents the median difference of the assay means from the grand mean the dotted horizontal lines bound an approximate 95 confidence region about the median

2PR4POTPNEIFR4Q5 D5ND6 D9 HBB1ND14 22C3-10

-8

-6

-4

-2

0

2

4

6

8

10

Rela

tive

Diffe

renc

e fr

om M

ean

Assays

Component MeansAssay MeansMedianU95

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

21

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

7112017 2PR4 1667 1698 1869 1960 1630 1706 POTP 1655 1685 1850 1781 1596 1576 NEIF 1647 1665 1709 1707 1574 1538 R4Q5 1594 1658 1856 1787 1590 1555 D5 1655 1683 1848 1829 1610 1607 ND6 1629 1647 1828 1826 1589 1548 D9 1655 1652 1861 1858 1569 1572 HBB1 1644 1611 1825 1869 1565 1681 ND14 1645 1707 1871 1773 1564 1492 22C3 1545 1603 1776 1902 1548 1655

7122017 2PR4 1668 1739 2174 2231 1499 1497 POTP 1587 1702 2202 2292 1576 1611 NEIF 1678 1614 2172 2216 1545 1560 R4Q5 1628 1762 NA 2078 1584 1546 D5 1738 1649 2230 2192 1577 1529 ND6 1667 1678 2094 2034 1599 1478 D9 1695 1653 1949 2110 1587 1504 HBB1 1757 1750 1971 2123 1520 1551 ND14 1771 1764 2123 2066 1479 1438 22C3 1762 1759 2107 2046 1512 1494

7122017 2PR4 1638 1667 1810 1806 1452 1424 POTP 1676 1638 1865 1801 1347 1397 NEIF 1616 1624 1806 1786 1395 1350 R4Q5 1559 1587 1878 1768 1442 1459 D5 1608 1652 1824 1790 1448 1602 ND6 1611 1589 1768 1670 1463 1435 D9 1517 1607 1811 1749 1402 1433 HBB1 1557 1582 1763 1741 1474 1481 ND14 1604 1625 1865 1781 1466 1482 22C3 1675 1616 1815 1866 1439 1568

7132017 2PR4 1888 1939 1784 1821 1470 1444 POTP 1868 1815 1735 1813 1411 1427 NEIF 1953 1902 1778 1813 1512 1439 R4Q5 1779 1871 1835 1749 1452 1468 D5 1924 2124 1794 1848 1567 1524 ND6 1861 1787 1700 1711 1480 1424 D9 1824 1844 1799 1779 1442 1459 HBB1 1809 1799 1765 1769 1430 1467 ND14 1864 1857 1754 1725 1413 1411 22C3 1930 1942 1792 1814 1486 1480

7142017 2PR4 1597 1590 1718 1768 1485 1458 POTP 1533 1404 1753 1698 1422 1459 NEIF 1462 1533 1773 1792 1438 1465 R4Q5 1493 1519 1777 1740 1439 1468 D5 1538 1586 1803 1795 1515 1457 ND6 1492 1516 1718 1693 1409 1412 D9 1528 1538 1802 1743 1469 1474 HBB1 1548 1581 1834 1745 1453 1428 ND14 1513 1514 1731 1714 1465 1425 22C3 1562 NA 1842 NA 1363 1393

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

22

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

101217 2PR4 1431 1411 1791 1770 1357 1364 POTP 1407 1395 1746 1820 1447 1361 NEIF 1487 1398 1803 1803 1448 1404 R4Q5 1450 1439 1767 1739 1409 1411 D5 1394 1422 1745 1773 1389 1429 ND6 1396 1393 1824 1729 1415 1417 D9 1414 1390 1752 1744 1431 1380 HBB1 1401 1382 1813 1754 1381 1336 ND14 1412 1460 1648 1671 1403 1387 22C3 1435 1410 1818 1741 1435 1372

Estimation of Copy Number Concentration It was decided that all three sets of results (homogeneity stability and certification) would be used to derive the certified values with the λ of the first five ldquoten-assayrdquo plates adjusted by a data-driven estimate of the volumetric bias This bias was incorporated into the statistical model providing better estimates of the λ means while capturing all of the variance components (homogeneity stability between-assay and between data sets) as random effects using Bayesian MCMC programmed in OpenBUGS28 Table 14 lists the estimated values and their uncertainties See Appendix C for details

Table 14 Measured Copy Number Concentrations

Genomic Copies per Nanoliter of Solution Component Mean a u(Mean) b U95(Mean) c CV95

d A 151 07 14 93 B 175 03 06 34 C 145 05 10 69

a Mean the experimentally determined most probable estimate of the true value b u(Mean) the standard uncertainty of the Mean c Mean plusmnU95(Mean) the interval which is believed with about 95 confidence to contain the true value d CV95 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the counting unit one29 2) the validity of the Poisson endpoint transformation for digital PCR endpoint assays when applied to samples providing le 05 copies per droplet 3) the determination that the copies are dispersed as dsDNA and 4) calibrated mean droplet volume measurements made at NIST during sample dilution and mastermix preparation Conversion of Copy Number to Mass Concentration As described in Reference 18 metrologically valid conversion from copies per nanoliter to nanograms per microliter requires quantitative estimates for the number of nucleotide bases per reference human genome and the average nucleotide mass Using the values estimated in that report and asserting that each target copy corresponds to one HHGE the required transformation is

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

23

[DNA] ngmicroL

= [DNA] HHGE

nLtimes

3301 ngHHGE

The combined standard uncertainty is

119906119906 [DNA] ng

microL = 1199061199062 [DNA] HHGE

nL + 0531100

times[DNA] HHGE

nL 2

Both the measured and the conversion factor uncertainties are associated with ldquolargerdquo degrees of freedom Therefor the approximate 95 confidence uncertainty is estimated

11988011988095 [DNA] ng

microL = 2 times 119906119906

[DNA] ngmicroL

Table 15 lists the transformation of the Table 14 copy number values These values agree well with the conventional 260 nm absorbance values averaging (4 plusmn 5) larger

Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution

ngmicroL Component Value a u(Measurement)b u(Conversion)c U95(Value)d CV95

e A 498 23 03 47 93 B 578 10 03 21 36 C 479 17 03 33 70

a Estimated true value b Standard uncertainty component from the ddPCR measurements c Standard uncertainty estimated from the relative uncertainty of the conversion factor d Value plusmnU95(Value) is believed with about 95 confidence to contain the true value e 100U95(Value)Value the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the measured estimates of copies per nanoliter and 2) determinations made by us from internationally recognized literature resources of the average number of nucleotide bases per HHGE and the mean molecular weights of the bases18 Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)

While not intended to provide certified values mtDNAnDNA was measured as part of 13 experiments over 5 months Measurements were made using dilution-adjusted λ for the mtND1 mtBatz andor mtKav mitochondrial assays referenced against the dilution-adjusted λ of either the NEIF or HBB1 nDNA assays By chance both DNAs in the component C mixture have a binding site mutation that silences the mtND1 probe Figure 11 graphically summarizes the results by assay and component Table 16 lists the data Table 17 summarizes results from a random effects model that accounts for between-assay variability See Appendix C for details

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

24

Due to the limited number of mtDNA assays used in these measurements these results are metrologically traceable only to the assays used

Figure 11 mtDNAnDNA for Components A B and C

Each dot-and-bar symbol represents a mean mtDNAnDNA and its approximate 95 expanded uncertainty The dotted horizontal lines represent approximate 95 confidence intervals on the ratios

Table 16 mtDNAnDNA Ratios

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

20170711a 1 mtND1 1931 1870 2461 2401 1 mtBatz 1845 1764 2555 2608 3201 3299

20170711b 1 mtND1 1810 1906 2229 2200 1 mtBatz 1785 1837 2108 2192 3347 3287

20170712 1 mtND1 1743 1761 2127 2198 1 mtBatz 1829 1847 2455 2506 3217 3384

20170713 1 mtND1 1716 1800 2211 2199 1 mtBatz 1740 1813 2380 2249 3200 3363

20170714 1 mtND1 1923 1835 2105 1943 1 mtBatz 1935 1851 2136 2119 3294 3013

20170814 1 mtND1 1780 1776 1706 2046 2057 2024 1 mtBatz 1804 1751 1738 1998 2043 1952 2927 2776 2762 1 mtKav 1769 1786 1761 2077 2117 2041 3044 2830 2825

20170824 1 mtND1 1947 1928 1922 2 mtND1 1829 1895 1851 3 mtND1 1674 1617

20170825 1 mtBatz 1667 1813 1708 1598 1928 1916 1915 2832 2840 2788 2725 2 mtBatz 1566 1687 1663 1589 1947 1946 1961 1909 2835 2790 2736 2679 3 mtBatz 1643 1668 1604 1573 1953 1952 1941 1881 2803 2773 2757 2752

20170829 1 mtBatz 1719 1716 1670 1666 1865 1863 1828 1828 2599 2508 2551 2609 2 mtBatz 1581 1573 1601 1612 1944 1964 1947 1895 2617 2483 2626 2602 3 mtBatz 1546 1542 1538 1587 1908 1909 1940 1839 2665 2535 2570 2650

20170831 1 mtND1 1617 1599 1688 1783

ND1

ND1

ND1Ka

v

Kav

Kav

Batz

Batz

Batz

A

B

C

160

210

260

mtD

NAn

DNA

Ratio

Assay

Bind

ing s

ite m

utat

ion

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

25

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

2 mtND1 1622 1579 1823 1631 3 mtND1 1575 1528 1920 1887

20170907 1 mtBatz 1584 1544 1890 1886 2501 2522 20170913 1 mtND1 1757 1724 1768 1990 1921 1982

1 mtBatz 1743 1671 1766 1990 1924 1958 2805 2825 2829 1 mtKav 1692 1662 1767 1903 1831 1919 2709 2738 2727

20171016 1 mtND1 1924 1944 1893 2304 2341 2332 1 mtBatz 1786 1752 1743 2048 2078 2067 2797 2642 2792 1 mtKav 1530 1542 1523 2399 2479 2493 2804 2789 2842 2 mtBatz 1580 1557 1592 2715 2606 2687 2 mtKav 2103 2100 2049 2799 2628 2559 3 mtND1 1749 1769 1777 2065 2166 2164 3 mtBatz 1965 1889 1947 2314 2280 2258 2639 2535 2657 3 mtKav 1813 1864 1864 1867 1987 1971 2953 2757 2802

Table 17 mtDNAnDNA Summary Results

mtDNAnDNA Component Value a u(Value)b U95(Value)c CV95

d A 1737 16 38 17 B 2058 22 44 21 C 2791 23 47 17

a Estimated true value b Standard uncertainty c Value plusmnU95(Value) is believed with about 95 confidence to contain the true value d 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) Performance in Commercial qPCR Chemistries

Traditional qPCR assays only quantify relative to a reference standard Thus qPCR assays do not provide absolute [DNA] for the individual materials but they can provide [DNA] of one component relative to another Figure 12 summarizes the BA and CA ratios for the commercial autosomal (Human) targets from the eleven assays evaluated (see Table 8) The Y-chromosome (Male) assays cannot be summarized in this manner as the (all female) component B does not include a Y chromosome Figure 13 summaries the mass [DNA] results for the (all male) component A and (male-female mixture) component C Since this material was prepared as a nominal 13 malefemale mixture the expected malehuman (total) ratio is 14 sjc As displayed in these figures there is intrinsic variability between the SRM 2372a components and between the commercial qPCR chemistries examined This reflects the difference in the targets and amplification chemistries for each of the commercial kits Many of the commercial chemistries on the market to date contain multi-copy targets for both autosomal and Y-chromosomal quantification to increase the sensitivity of the assays

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

26

Figure 12 qPCR Autosomal Quantitation Performance Summary

Open circles denote the BA CA ratios the 17 commercial qPCR assays evaluated each labeled with the code listed in Table 8 The solid black circle represents ratios as estimated by ddPCR The red lines bound the approximate 95 confidence intervals on the ddPCR-estimated ratios

Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary

Open squares denote the ngμL values for SRM 2372a Components A (male) and C (mixture of male and female) for the seven commercial Y-chromosome qPCR assays evaluated each labeled with the code listed in Table 8 The dotted red lines represent malehuman ratios of 13 14 and 15

Appendix D provides a graphical analysis of all the data generated in the performance assessments for each of the targets amplified in the eleven commercial qPCR chemistries and the [DNA] values for all of the assays

dd

IQ_h

IQ_d

IQHY_h

IQHY_d

QFD_h

QFH_h QFHP_h

QFHP_dQFT_h

QFT_d

QPH_h

QPP_hQPP_d

P_h

PQ_h

PQ_d

06

08

10

12

14

07 09 11 13

CA

BA

IQHY_y

QFD_y

QFT_y

QPH_y

QPP_y

P_y

PQ_y

1413

15

10

12

14

16

18

30 40 50 60 70

[DNA

] C n

gμL

[DNA]A nguL

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

27

Qualification of ddPCR as Fit-For-Purpose We have documented the requirements for establishing metrological traceability of ddPCR human genomic assay results expressed as copies per nanoliter of sample in a series of three peer-reviewed papers151617 Through the experimental evidence they are based upon these papers demonstrate that ddPCR can accurately enumerate the number of independently migrating genomic copies in an aqueous extract Our most recent paper establishes the metrological traceability of these ddPCR results when transformed into the nanogram per microliter units used by the forensic and clinical communities18 However our previous studies have not addressed a fundamental requirement for correctly interpreting digital PCR measurements In broad simplification end-point assay systems count the number of independently partitioning entities regardless of the number of copies the entities contain With assays that target just one site per HHGE entities that migrate as ssDNA will yield counts that are twice the value had they migrated as dsDNA30 Should some originally dsDNA be converted over time to ssDNA ddPCR measurements of the number of independently migrating entities would increase although the number of HHGE would be unchanged Measurements of mass [DNA] provided by absorbance spectrophotometry and ddPCR reflect different properties absorbance at 260 nm is sensitive to the spatial configuration of the nucleotides as well as their number while ddPCR counts independently migrating entities The SRM 2372a materials have absorbance spectra shown in Figure 8 consistent with being predominantly dsDNA The SRM 2372 materials now have the absorbance spectra shown in Figure 14 which are consistent with them being entirely ssDNA

Figure 14 Absorbance Spectra of SRM 2372 Components A and B dsDNA is denatured to ssDNA by diluting the source material with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to mass [DNA]

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372-A200720122012-NaOH121417

0

4

6

0

4

6

220 260 300 340

Wavelength nm

2372-B200720122012-NaOH121517

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

28

To ensure the validity of ddPCR-based value assignments it is thus necessary to evaluate the nature of the partitioning entities Should a solution of what is thought to be dsDNA contain an appreciable fraction of ssDNA then the ddPCR results would require adjustment31 To avoid further delay in making SRM 2372a available to the forensic community the following sections detail a series of experiments that have enabled us to conclude with high confidence that the SRM 2372a components have a dsDNA conformation and that for this material ldquoentitiesrdquo can be considered dsDNA ldquocopiesrdquo The remaining stock of the SRM 2372 components also have a dsDNA conformation however the strands in the old materials are much less robustly connected than they are in the new materials Digital Polymerase Chain Reaction Assays

Three nDNA assays were used HBB1 NEIF and ND6 The HBB1 and NEIF assays reproducibly yield very similar results for DNA extracts from many different donors The HBB1 target is at the tip of the short (p) arm of chromosome 11 The NEIF target is in the centromere region of chromosome 2 The ND6 assay was selected for use because it typically yields the lowest results of the ten nDNA assays used to certify the SRM 2372a components (see Figure 10 below) The ND6 target is about halfway down the long (q) arm of chromosome 6 Heat Treatments (Melting of the DNA into Single Strands)

In all experiments the non-heated control and the heat-treated diluted solutions were stored in the same type of PCR reaction tube The starting point for the heat treatment was to determine if there was a difference between snap-cooling on ice and slower cooling at 25 degC Two tubes of the diluted material were placed in a GeneAmp 9700 thermal cycler (Applied Biosystems) and cycled up to 98 degC and held for 15 min After 15 min one tube was removed and placed on ice while the remaining tube was cooled at 1 degCs in the thermal cycler to 25 degC then placed on ice until prepared for ddPCR Additional heat treatments included cycling samples on a GeneAmp 9700 thermal cycler

a) from 98 degC for 5 min then cooling to 25 degC for 2 min During the 25 degC hold a randomly selected tube was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 45 min at 98 degC

b) from 94 degC for 1 min then cooling to 25 degC for 1 min During the 1 min hold at 25 degC a random sample was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 9 min at 94 degC

c) from 90 degC for 1 min then cooling to 25 degC for 1 min for a total of 20 cycles Samples were pulled at 25 degC after 1 min 3 min 5 min 10 min and 20 min and placed on ice

After the above experiments heat treatments were performed using a ProFlex thermal cycler This thermal cycler has six temperature zones that can be le 5 degC apart This enables heating

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

29

tubes at the same time but at different temperatures After treatment tubes were cooled at 5 degCs to 4 degC in the thermal cycler While designed for use with PCR plates a MicroAmp 96-well tray for the VeriFlex block enabled its use with single PCR reaction tubes Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a

As a first step in evaluating whether the DNA in the SRM materials disperse as dsDNA or ssDNA a direct comparison of the five materials was made using the HBB1 and NEIF assays Table 18 reports summary statistics for the ddPCR measurements reported as λ copiesdroplet and the estimated mass [DNA] values from the ddPCR results18 from the conventional dsDNA relationship7 and from the conventional ssDNA relationship after denaturing with 04 molL NaOH89 While the mass [DNA] values differ somewhat between the three methods there is no evidence for the SRM materials segregating exclusively or even mostly as ssDNA The extent of the hypochromic interactions among the nucleotides does not accurately reflect the extent to which the two strands of dsDNA have separated

Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a

λ copies per droplet λ copies per droplet ngmicroL Sample Assay Nrep

a Mean CVb ncmbc Mean CVb ddPCRd dsDNAe ssDNAf

2372-A HBB1 4 03645 19 8 03644 20 566g 524 57 2372-A NEIF 4 03644 23 2372-B HBB1 4 03513 23 8 03487 35 580h 536 61 2372-B NEIF 4 03461 47 2372a-A HBB1 4 02603 27 8 02649 30 474 465 454 2372a-A NEIF 4 02694 24 2372a-B HBB1 4 02908 34 8 02891 37 517 556 534 2372a-B NEIF 4 02873 45 2372a-C HBB1 4 02757 28 8 02711 28 485 502 474 2372a-C NEIF 4 02665 16

a Number of technical replicates for the given assay b CV = 100(standard deviation)mean c Combined number of technical replicates ignoring potential differences between the assays d Estimated from the ddPCR λ measurements listed in this Table18 e Estimated from the conventional relationship between the mass [DNA] of dsDNA and absorbance at

260 nm7 using absorbance spectra acquired shortly after solutions were produced f Estimated from the relationship between the mass [DNA] of ssDNA and absorbance at 260 nm after 11

dilution with 04 molL NaOH89 g Values adjusted to measured (132 plusmn19) volume loss of archived units due to evaporation h Values adjusted to measured (71 plusmn25) volume loss of archived due to evaporation

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

30

Quantitative Evaluation of ssDNA Content

Absorbance spectrophotometry is routinely used to quantitatively assess DNA tertiary structure3233 On storage ldquoat physiological pH the intrastrand repulsion between the negatively charged phosphate groups forces the double helix into more rigid rodlike conformationrdquo34 suggesting a mechanism for the observed slow changes in the SRM 2372 materials As demonstrated in Table 18 absorbance at 260 nm does not necessarily correlate with ddPCR behavior Yet ldquofurthermore the repulsion between phosphate groups on opposite strands tends to separate the complementary strandsrdquo34 Therefore a method that is independent of hypochromism is needed to adequately evaluate the proportion (if any) of ldquotruerdquo (completely disjoint) ssDNA in the SRM materials Circular dichroism spectroscopy is sensitive to the conformational structures of DNA in solution35 and different DNA configurations migrate differentially in agarose gel electrophoresis7 Both techniques have been used to qualitatively discriminate ssDNA from dsDNA in ldquopurerdquo single-sequence materials but neither technique appears to be able to quantify small proportions of ssDNA in a complex multi-chromosome genomic material Wilson and Ellison have proposed a real-time chamber digital PCR (cdPCR) technique that can estimate excess proportions of ssDNA copies from the crossing-threshold distribution31 However their method is insufficiently sensitive for low ssDNA proportions It is well known that heating DNA beyond the melting temperature separates mammalian dsDNA into ssDNA36 albeit we have shown using cdPCR that boiling dsDNA for some minutes also renders some target sites non-amplifiable15 In the following section we document a ddPCR-based approach for estimating the proportion if any of ssDNA in a sample that involves less destructive heating While the method is not yet optimized the results from these preliminary studies demonstrate that the SRM 2372 and 2372a components disperse into the ddPCR droplets as dsDNA ddPCR Measurement of ssDNA Proportion Given two otherwise identical aliquots of a given DNA material one subjected to some treatment and the other untreated the ratio between the average number of copies per droplet of treated material and that of its untreated sibling λtλu quantitatively estimates the change in the number of independently dispersing entities Should the treatment reduce the number of accessible amplifiable targets (AAT) λtλu will be less than 1 Reduction in AAT may result from damage to the target itself andor conformational changes to the fragment containing the target that render the target inaccessible or non-amplifiable16 Assuming no denaturation of dsDNA to ssDNA either because of an ineffective treatment or the material having started as ldquopurerdquo ssDNA if the treatment does not itself damage targets then λtλu will equal 1 Should treatment convert at least some dsDNA to ssDNA without damaging targets the number of AAT will increase and λtλu will be greater than 1 Assuming complete conversion of ldquopurerdquo dsDNA to ldquopurerdquo ssDNA without other damage then λtλu will equal 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

31

Values greater than 1 but less than 2 indicate some conversion of dsDNA to ssDNA but also some combination of (a) incomplete denaturation of dsDNA to ssDNA (b) partial renaturation of ssDNA to dsDNA after treatment but before ddPCR analysis (c) damage to targets during the process or ndash of greatest interest ndash (d) the original sample material was an impure mixture of dsDNA and ssDNA The following discussions graphically summarize sets of λtλu followed by tabular listings of the relevant experimental conditions ddPCR measurements and the numerical λtλu values The information in these tables includes

Plate Index Experiment identifier indicates date performed in YYYYMMDD format Material Sample material SRM 2372-A 2372-B 2372a-A 2372a-B or 2372a-C Assay ddPCR assay HBB1 NEIF or ND6 degC Denaturation temperature in degC Min Length of time the treated material was held at denaturation temperature in

minutes Cool Method used to cool treated material from denaturation temperature Snap

Slow R1 or R5 where ldquoSnaprdquo indicates cooling on ice ldquoSlowrdquo cooling at room temperature to 25 degC ldquoR1rdquo cooling in a thermal cycler to 25 degC at 1 degCs and ldquoR5rdquo cooling in a thermal cycler to 4 degC at 5 degCs

F Dilution factor expressed as (volume Material) per (final volume of diluted solution) 12 14 or 18

n Number of technical replicates Mean Mean of technical replicates CV Coefficient of Variation expressed as a percentage 100 Standard Deviation

Mean

λtλu the copies per droplet result of the treated solution λt divided by the copies per droplet result of its untreated sibling λu

u(λtλu) Standard uncertainty of λtλu 119906119906(120582120582t 120582120582ufrasl ) = 120582120582t 120582120582ufrasl CV(120582120582t)100

2

+ CV(120582120582u)100

2

Proof of Principle Figure 15 summarizes two replicate sets of measurements performed to assess whether the SRM 2372 and 2372a materials respond similarly whether the denatured materials renature if allowed to cool slowly and whether results are (reasonably) repeatable Table 19 details the experimental conditions The λtλu values for the four materials are similar all much greater than 1 but not approaching 2 Most of the ratios are on or close to the ldquoslow cooling is the same as snap coolingrdquo line suggesting that renaturation is not a significant concern While the replicate results for the SRM 2372-A and 2372a-A materials do not overlap as closely as those for SRM 2372-B and 2372a-B similar treatment conditions produce similar results

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

32

Figure 15 Comparison of Snap-Cooling to Slow-Cooling

Symbols mark the means of replicate measurements made using the same dilutions on successive days with bars representing plusmn one standard combined uncertainty Results are labeled by material ldquoArdquo and ldquoBrdquo for SRM 2372 components and ldquoaArdquo ldquoaBrdquo and ldquoaCrdquo for SRM 2372a components The diagonal black line represents equality between the snap- and slow-cooling treatments

Table 19 Comparison of Snap-Cooling to Slow-Cooling

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171218 2372-A HBB1 14 4 03830 29

dd20171218 2372-A HBB1 98 15 Slow 14 2 05644 18 1474 0051 dd20171218 2372-A HBB1 98 15 Snap 14 4 05852 15 1528 0051

dd20171218 2372a-A HBB1 14 4 03029 18

dd20171218 2372a-A HBB1 98 15 Slow 14 4 04430 21 1463 0040 dd20171218 2372a-A HBB1 98 15 Snap 14 4 04454 33 1471 0055

dd20171218 2372-B HBB1 14 4 03594 31

dd20171218 2372-B HBB1 98 15 Slow 14 4 05806 18 1615 0058 dd20171218 2372-B HBB1 98 15 Snap 14 4 05825 13 1620 0055

dd20171218 2372a-B HBB1 14 4 03383 15

dd20171218 2372a-B HBB1 98 15 Slow 14 4 05043 38 1491 0061 dd20171218 2372a-B HBB1 98 15 Snap 14 4 05254 20 1553 0039

dd20171219 2372-A HBB1 14 4 03357 23

dd20171219 2372-A HBB1 98 15 Slow 14 4 05340 16 1591 0046 dd20171219 2372-A HBB1 98 15 Snap 14 4 05340 12 1591 0042

A

aA

B

aB

A

aA

B

aB

14

15

16

17

14 15 16 17

λ tλ

u Sn

ap C

oole

d

λtλu Slow Cooled

Rep 1Rep 2Snap = Slow

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

33

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171219 2372a-A HBB1 14 4 02735 24

dd20171219 2372a-A HBB1 98 15 Slow 14 4 04151 03 1518 0037 dd20171219 2372a-A HBB1 98 15 Snap 14 4 03918 21 1433 0045 dd20171219 2372-B HBB1 14 4 03346 22

dd20171219 2372-B HBB1 98 15 Slow 14 4 05476 17 1637 0045 dd20171219 2372-B HBB1 98 15 Snap 14 4 05480 29 1638 0059 dd20171219 2372a-B HBB1 14 4 03158 19

dd20171219 2372a-B HBB1 98 15 Slow 14 4 04798 16 1519 0038 dd20171219 2372a-B HBB1 98 15 Snap 14 4 04856 17 1538 0039

Time at Denaturation Temperature Figure 16 summarizes λtλu for the SRM 2372-B material as a function of denaturation temperature and the cumulative time the solution is held at that temperature Table 20 details the experimental conditions

Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures

The symbols represent the ratio of the mean result of three technical replicates of the untreated and each of the treated solutions the bars represent combined standard uncertainties The lines denote regression fits to the function λtλu = atimese-bt where t is the cumulative time Values and the standard uncertainties of the coefficients are listed in the legend

12

14

16

18

20

0 5 10 15 20

λ t λ

u

Cummulative Treatment Time min

degC a u (a ) b u (b )90 1804 0039 00198 0000494 1788 0029 00099 0002998 1855 0019 00067 00019

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

34

For all three of the denaturation temperatures studied the relationship between λtλu and cumulative time t can be modeled as

λtλu = ae-bt

where a is the extrapolated value of λtλu at t = 0 and b is the rate of change in AAT per minute As expected the rate of loss due to the treatment increases with increasing treatment temperature and cumulative time at that temperature37 Given the much longer treatment time for the 98 degC series the agreement among the three t = 0 parameter (a) values is remarkable This suggests that the first 1 min of treatment efficiently separates dsDNA into ssDNA The pooled relative standard uncertainty of these regression parameter is 17 somewhat smaller than the 25 pooled relative standard uncertainties of the individual λtλu However the plusmnstandard uncertainty interval of the 1-min λtλu results for the 94 degC and 90 degC both overlap the extrapolated t = 0 value suggesting that reliable results can be obtained using a single treatment of short duration

Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171228 2372-B HBB1 14 3 03650 17 dd20171228 2372-B HBB1 98 5 R1 14 3 06157 12 1687 0035 dd20171228 2372-B HBB1 98 10 R1 14 3 05493 08 1505 0028 dd20171228 2372-B HBB1 98 15 R1 14 3 05019 13 1375 0029 dd20171228 2372-B HBB1 98 20 R1 14 3 04583 21 1256 0034 dd20171228 2372-B HBB1 98 25 R1 14 3 04154 32 1138 0041 dd20171228 2372-B HBB1 98 30 R1 14 3 03784 14 1037 0023 dd20171228 2372-B HBB1 98 35 R1 14 3 03408 10 0934 0019 dd20171228 2372-B HBB1 98 40 R1 14 3 02980 01 0816 0014 dd20171228 2372-B HBB1 98 40 R1 14 3 02815 08 0771 0015

dd20171229 2372-B HBB1 14 3 03708 27

dd20171229 2372-B HBB1 94 1 R1 14 3 06598 26 1779 0056 dd20171229 2372-B HBB1 94 2 R1 14 3 06553 32 1767 0063 dd20171229 2372-B HBB1 94 3 R1 14 3 06301 20 1699 0045 dd20171229 2372-B HBB1 94 4 R1 14 3 06252 14 1686 0037 dd20171229 2372-B HBB1 94 5 R1 14 3 06577 14 1773 0039 dd20171229 2372-B HBB1 94 6 R1 14 3 06156 07 1660 0030 dd20171229 2372-B HBB1 94 7 R1 14 3 06118 10 1650 0033 dd20171229 2372-B HBB1 94 8 R1 14 3 06256 12 1687 0035 dd20171229 2372-B HBB1 94 9 R1 14 3 05999 10 1618 0032

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

35

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180104 2372-B HBB1 14 4 03713 25 dd20180104 2372-B HBB1 90 1 R1 14 3 06916 22 1863 0052 dd20180104 2372-B HBB1 90 3 R1 14 3 06465 06 1741 0032 dd20180104 2372-B HBB1 90 5 R1 14 3 06431 20 1732 0045 dd20180104 2372-B HBB1 90 10 R1 14 3 06014 25 1620 0049 dd20180104 2372-B HBB1 90 15 R1 14 3 06024 10 1623 0032 dd20180104 2372-B HBB1 90 20 R1 14 3 06003 14 1617 0035

Temperature Figure 17 summarizes the effect of denaturation temperature using a single treatment of 05 min on the 2372-B and 2372a-B materials Table 21 details the experimental conditions

Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

The symbols represent the ratio of the mean result of four technical replicates of the untreated and three replicates of the treated solutions the bars represent combined standard uncertainties In all cases the duration of the treatment was 05 min The lines denote regression fits to the linear function λtλu = a +bT where Tis the denaturation temperature

The SRM 2372 and 2372a materials respond very differently to changes in denaturation temperature The older spectrophotometrically ssDNA material suffers somewhat greater loss in λtλu as temperature increases The new spectrophotometrically dsDNA material suffers much greater loss in λtλu as temperature decreases

14

15

16

17

18

19

20

82 86 90 94 98

λ tλ

u

Temperature degC

2372-B

2372a-B

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

36

Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180109 2372-B HBB1 14 4 03577 25 dd20180109 2372-B HBB1 100 05 R5 14 3 06314 16 1765 0051

dd20180104a 2372-B HBB1 14 4 03577 05

dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032 dd20180104a 2372-B HBB1 94 05 R5 14 3 06459 25 1806 0046 dd20180104a 2372-B HBB1 90 05 R5 14 2 06591 08 1842 0017 dd20180104a 2372-B HBB1 88 05 R5 14 3 06589 05 1842 0013 dd20180104a 2372-B HBB1 84 05 R5 14 3 06729 10 1881 0022 dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

We hypothesize that in the older SRM 2372 material the between-strand bond is relatively fragile and can be completely severed at relatively low temperature while avoiding much temperature-dependent AAT loss In contrast the between-strand bond in the SRM 2372a material requires higher temperature to completely disassociate the strands This suggests that there is not a sample-independent ldquooptimumrdquo combination of denaturation temperature and treatment duration However 98 degC for 05 min appears to provide comparable results for these two materials DNA Concentration in the ddPCR Solution Figure 18 summarizes the effect of [DNA] in the ddPCR solution with SRM 2372-B at a denaturation temperature of 98 degC and treatment duration of 05 min Table 22 details the experimental conditions

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

37

Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution

The solid black symbols represent the ratio of the mean result of four technical replicates of the untreated and treated solutions with the λtλu values shown on the left-hand vertical axis the bars represent combined standard uncertainties The solid blue circles represent λu and the open red circles λt with values shown on the right-hand vertical axis at this graphical scale the standard deviations of the technical replicates are covered by the symbol The diagonal blue and red lines represent the log-linear relationships between just the 18 and 14 λ values

The primary effect of [DNA] appears to be the loss of linearity in λt with copiesdroplet values greater than about 1 copiesdroplet There may be some advantage to use somewhat greater dilutions than the routine 14 however the small increase in λtλu may be offset by the increased variability with small λu This suggests the use of dilutions that give λu values of (01 to 04) copiesdroplet

Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180117 2372-B HBB1 12 4 07127 11 dd20180117 2372-B HBB1 98 05 R5 12 4 12331 14 1730 0031

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 3 06151 15 1843 0040

dd20180117 2372-B HBB1 18 4 01549 18 dd20180117 2372-B HBB1 98 05 R5 18 4 02897 20 1870 0051

12 14 18

170

175

180

185

190

02

04

06

08

101214

λ T

arge

t D

ropl

et

λ tλ

u

Dilution Factor

λ₊λᵤλᵤλ₊

λtλu

λu

λt

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

38

Equivalence of Assays Figure 19 summarizes the agreement between the HBB1 NEIF and ND6 assays at a denaturation temperature of 98 degC and treatment duration of 05 min Table 23 details the experimental conditions

Figure 19 Comparison of Assays

The symbols compared the λtλu results from the HBB1 and either the NEIF or ND6 assays Each λtλu value is the ratio of the mean result of four technical replicates of an untreated and three or four replicates of the treated solutions The bars represent combined standard uncertainties The diagonal line represents equality between the assays

There is little or no difference between the λtλu results from the HBB1 NEIF and ND6 assays suggesting that the chromosomal location of the assay target has little influence Since the NEIFHBB1 comparison includes results for both the SRM 2372 and 2372a materials this equivalence is not material-specific

14

16

18

20

14 16 18 20

λ tλ

u NE

IF o

r ND6

λtλu HBB1

NEIFHBB1

ND6HBB1

2nd Assay = HBB1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

39

Table 23 Comparison of Assays

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu)

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045 dd20180105 2372a-B ND6 94 05 R5 14 3 05526 36 1711 0074 dd20180105 2372a-B ND6 90 05 R5 14 3 05326 18 1649 0051 dd20180105 2372a-B ND6 88 05 R5 14 3 05077 22 1572 0052 dd20180105 2372a-B ND6 86 05 R5 14 3 05038 23 1560 0053 dd20180105 2372a-B ND6 84 05 R5 14 3 04906 10 1519 0041

dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

40

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials

Figure 20 summarizes the currently available λtλu results obtained by denaturing at 98 degC for 05 min rapidly cooling in the thermal cycler with dilutions that provide asymp 03 copiesdroplet The figure also displays the consensus result and its 95 confidence region estimated using the Hierarchical Bayes method provided in the NIST Consensus Builder (NICOB) system3839 Table 24 details the experimental conditions

Figure 20 Evaluation of Limiting Ratio

The symbols represent the ratio of the mean result of three or four technical replicates of the untreated and treated solutions the bars represent combined standard uncertainties The thick horizontal line represents a consensus value covering all displayed results the dotted horizontal lines bracket the 95 confidence interval about the consensus value

The Hierarchical Bayes consensus result for these values is 183 with a 95 confidence interval of plusmn003 All but three of the 16 λtλu values are within one standard uncertainty of this consensus value with most of the λtλu means lying within the 95 confidence interval This suggests that there are no statistically significant differences between the SRM 2372 and 2372a materials with regard to ssDNA proportion

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

17

18

19

20

λ tλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

41

Table 24 Evaluation of Limiting Ratio

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180104a 2372-B HBB1 14 4 03577 05 dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032

dd20180108 2372-B HBB1 14 3 03463 09 dd20180108 2372-B HBB1 98 05 R5 14 3 06562 14 1895 0031

dd20180109 2372-B HBB1 14 3 03577 25 dd20180109 2372-B HBB1 98 05 R5 14 3 06495 21 1816 0059

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 4 06151 15 1843 0040

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

42

The agreement among the measured λtλu is more directly visualized using estimated unilateral ldquoDegrees of Equivalencerdquo d Each d is the absolute difference between the measured and consensus values

119889119889 = |120582120582119905119905 120582120582119906119906frasl minus consensus|

and is associated with a 95 confidence interval U95(d) that combines the uncertainty estimates of the measured and consensus values The U95(d) calculation details are specific to the consensus estimation method39 Values of d plusmn U95(d) that include 0 indicate that the λtλu result does not differ from the majority at about a 95 level of confidence Figure 21 displays the d for the λtλu results for the Hierarchical Bayes consensus value39 By this figure of merit all the λtλu can be considered as members of a single population

Figure 21 Degrees of Equivalence

The symbols represent the Degree of Equivalence d for the λtλu results relative to the Hierarchical Bayes consensus value The bars represent 95 level of confidence uncertainty intervals d plusmn U95(d) The horizontal line marks the d = 0 ldquono differencerdquo value Results with d plusmn U95(d) that include this line are not significantly different from the majority

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

-024

-018

-012

-006

000

006

012

018

024

Degr

ees o

f Equ

ival

ence

λtλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

43

Conclusions and Recommendations With high confidence all tubes of the SRM 2372a solution have the same nDNA copy number content within measurement repeatability Additionally within measurement repeatability all tubes of the SRM 2372a have the same mtDNAnDNA With high confidence the nDNA copy number content of the SRM 2372a solution is thermally stable from 4 degC to 37 degC over an extended period The solution should not be shipped or stored below 4 degC With high confidence the ddPCR measurements used to establish the entity and mass concentrations and the mtDNAnDNA ratios are adequately precise and unbiased With high confidence the DNA in the SRM 2372a solutions is now dsDNA and can be expected to remain dsDNA for at least ten years when stored at 4 degC Recommended Certification Values

Since the component materials are believed to be homogeneous are of very similar composition and were prepared and processed similarly the observed differences in the measured entity concentration uncertainties are likely artifacts of the measurement process To limit over-interpretation of these differences we recommend that the copy per nanoliter values of the three components be assigned same 95 relative uncertainty interval (CV95) based on the largest of the measured CV95 values (93 ) rounded up to 10 Since the ngμL transformation does not appreciably inflate this relative uncertainty we recommend that the mass concentration values also be assigned CV95 of 10 Similarly we recommend that the mtDNAnDNA ratios be rounded to integer values all three components be assigned the same CV95 and that this uncertainty be based on the largest (21 ) of the measured CV95 rounded up to 25 Table 25 lists the values that we recommend be used in the Certificate of Analysis

Table 25 Recommended Values and 95 Uncertainties for Certification

Component Units Value U95(Value) a A Copies per nanoliter 151 15 B Copies per nanoliter 175 18 C Copies per nanoliter 145 15 A ngμL 498 50

B ngμL 578 58 C ngμL 479 48 A mtDNAnDNA 174 4

B mtDNAnDNA 206 5 C mtDNAnDNA 279 7

a The interval Value plusmn U95(Value) is believed with about 95 confidence to contain the true value of the measurand

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

44

Use of SRM 2372a for Value-Assigning In-House Reference Materials

The within-component and between-assay variability observed for the commercial qPCR assays is not unexpected since different qPCR methods exploit qualitatively different molecular targets Given the intrinsic variability of the human genome the prospect of developing a human DNA source material that responds equally to all current and future qPCR assays is improbable All three components should be used to elucidate potential bias when using these materials to value assign in-house DNA reference and control materials for qPCR assays

Certificate of Analysis In accordance with ISO Guide 31 2000 a NIST SRM certificate is a document containing the name description and intended purpose of the material the logo of the US Department of Commerce the name of NIST as a certifying body instructions for proper use and storage of the material certified property value(s) with associated uncertainty(ies) method(s) used to obtain property values the period of validity if appropriate and any other technical information deemed necessary for its proper use A Certificate is issued for an SRM certified for one or more specific physical or engineering performance properties and may contain NIST reference information or both values in addition to certified values A Certificate of Analysis is issued for an SRM certified for one or more specific chemical properties Note ISO Guide 31 is updated periodically check with ISO for the latest version [httpswwwnistgovsrmsrm-definitions] For the most current version of the Certificate of Analysis for NIST SRM 2372a Human DNA Quantification Standard please visit httpswww-snistgovsrmorsview_detailcfmsrm=2372a

References

1 National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research Department of Health Education and Welfare (1978) Belmont Report Ethical Principles and Guidelines for the Protection of Human Subjects of Research DHEW Publication No (OS) 78-0012 httpvideocastnihgovpdfohrp_belmont_reportpdf

2 CDCNIH Biosafety in Microbiological and Biomedical Laboratories 5th ed HHS publication No (CDC) 21-1112 Chosewood LC Wilson DE Eds US Government Printing Office Washington DC (2009) httpwwwcdcgovbiosafetypublicationsbmbl5indexhtm

3 Duewer DL Kline MC Redman JW Newall PJ Reeder DJ NIST Mixed Stain Studies 1 and 2 Interlaboratory Comparison of DNA Quantification Practice and Short Tandem Repeat Multiplex Performance with Multiple-Source Samples J Forensic Sci 200146(5)1199-1210 PMID 11569565

4 Kline MC Duewer DL Redman JW Butler JM NIST Mixed Stain Study 3 DNA quantitation accuracy and its influence on short tandem repeat multiplex signal intensity Anal Chem 200375(10)2463-2469 httpsdoiorg101021ac026410i

5 Kline MC Duewer DL Redman JW Butler JM Results from the NIST 2004 DNA Quantitation Study J Forensic Sci 200550(3)571-578 PMID 15932088

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

45

6 Kline MC Duewer DL Travis JC Smith MV Redman JW Vallone PM Decker AE Butler JM Production and Certification of Standard Reference Material 2372 Human DNA Quantitation Standard Anal Bioanal Chem 2009394(4)1183-1192 httpsdoiorg101007s00216-009-2782-0

7 Sambrook J and Russell DW (2001) Molecular Cloning a Laboratory Manual Cold Spring Harbor Laboratory Press Cold Spring Harbor New York

8 EU-JRC Protocol NK603 ndash Community Reference Laboratory Event-specific method for the quantification of maize line NK603 using real-time PCR Section 33 Spectrophotometric Measurement of DNA Concentration httpgmo-crljrceceuropaeusummariesNK603-WEB-ProtocolValidationpdf

9 ISO 215712005(E) Foodstuffs ndash Methods of analysis for the detection of genetically modified organisms and derived products ndash Nucleic acid extraction Annex B Methods for quantitation of extracted DNA pp 34-36 International Organization for Standardization Geneva (2005)

10 Certificate of Analysis Standard Reference Material 2372 Human DNA Quantitation Standard (2013) Gaithersburg MD USA Standard Reference Materials Program National Institute of Standards and Technology httpswww-snistgovsrmorsview_certcfmsrm=2372

11 Vogelstein B Kinzler KW Digital PCR Proc Natl Acad Sci USA 1999969236-9241 12 Hindson BJ Ness KD Masquelier DA Belgrader P Heredia NJ Makarewicz AJ Bright IJ Lucero

MY Hiddessen AL Legler TC Kitano TK Hodel MR Petersen JF Wyatt PW Steenblock ER Shah PH Bousse LJ Troup CB Mellen JC Wittmann DK Erndt NG Cauley TH Koehler RT So AP Dube S Rose KA Montesclaros L Wang S Stumbo DP Hodges SP Romine S Milanovich FP White HE Regan JF Karlin-Neumann GA Hindson CM Saxonov S Colston BW High-Throughput Droplet Digital PCR System for Absolute Quantitation of DNA Copy Number Anal Chem 2011838604-8610 httpsdoiorg101021ac202028g

13 Pinheiro LB Coleman VA Hindson CM Herrmann J Hindson BJ Bhat S Emslie KR Evaluation of a Droplet Digital Polymerase Chain Reaction Format for DNA Copy Number Quantification AnalChem 2012841003-1011 httpsdoiorg101021ac202578x

14 Corbisier P Pinheiro L Mazoua S Kortekaas AM Chung PY Gerganova T Roebben G Emons H Emslie K DNA copy number concentration measured by digital and droplet digital quantitative PCR using certified reference materials Anal Bioanal Chem 2015407(7)1831-1840 httpsdoiorg101007s00216-015-8458-z

15 Duewer DL Kline MC Romsos EL Real-time cdPCR opens a window into events occurring in the first few PCR amplification cycles Anal Bioanal Chem 2015407(30)9061-9069 httpsdoiorg101007s00216-015-9073-8

16 Kline MC Romsos EL Duewer DL Evaluating Digital PCR for the Quantification of Human Genomic DNA Accessible Amplifiable Targets Anal Chem 201688(4)2132-2139 httpsdoiorg101021acsanalchem5b03692

17 Kline MC Duewer DL Evaluating Droplet Digital Polymerase Chain Reaction for the Quantification of Human Genomic DNA Lifting the Traceability Fog Anal Chem 201789(8)4648-4654 httpsdoiorg101021acsanalchem7b00240

18 Duewer DL Kline MC Romsos EL Toman B Evaluating Droplet Digital PCR for the quantification of human genomic DNA Conversion from Copies per Nanoliter to Nanogram Genomic DNA per Microliter Anal Bioanal Chem 2018in press httpsdoiorg101007s00216-018-0982-1

19 Dagata JA Farkas N Kramer JA Method for measuring the volume of nominally 100 μm diameter spherical water in oil emulsion droplets NIST Special Publication 260-184 2016 httpsdoiorg106028NISTSP260-184

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

20 ISO 170342016 General requirements for the competence of reference material producers International Organization for Standardization Geneva (2016)

21 NCBI Standard Nucleotide BLAST httpsblastncbinlmnihgovBlastcgiPAGE_TYPE=BlastSearch

22 Timken MD Swango KL Orrego C Buoncristiani MR A Duplex Real-Time qPCR Assay for the Quantification of Human Nuclear and Mitochondrial DNA in Forensic Samples Implications for Quantifying DNA in Degraded Samples J Forensic Sci 200550(5)1044-60 PMID 16225209

23 Walker JA Hedges DJ Perodeau BP Landry KE Stoilova N Laborde ME Shewale J Sinha SK Batzer MA Multiplex polymerase chain reaction for simultaneous quantitation of human nuclear mitochondrial and male Y-chromosome DNA application in human identification Anal Biochem 200533789ndash97 httpsdoiorg101016jab200409036

24 Kavlick MF Lawrence HS Merritt RT Fisher C Isenberg A Robertson JM Budowle B Quantification of Human Mitochondrial DNA Using Synthesized DNA Standards J Forensic Sci 201156(6) 1457ndash1463 httpsdoiorg101111j1556-4029201101871x

25 Stulnig TM Amberger A Exposing contaminating phenol in nucleic acid preparations BioTechniques 199416(3) 403-404

26 Majumdar N Wessel T Marks J Digital PCR Modeling for Maximal Sensitivity Dynamic Range and Measurement Precision PLoS ONE 201510(3)e0118833 httpsdoi101371journalpone0118833

27 Tellinghuisen J Partition Volume Variability in Digital Polymerase Chain Reaction Methods Polydispersity Causes Bias but Can Improve Precision Anal Chem 2016881283-12187 httpsdoiorg101021acsanalchem6b03139

28 Lunn DJ Spiegelhalter D Thomas A Best N (2009) Statistics in Medicine 283049ndash3082 29 De Biegravevre P Dybkaer R Fajgelj A Hibbert DB Metrological traceability of measurement results in

chemistry Concepts and implementation Pure Appl Chem 201183(10)1873-1935 httpsdoiorg101007s00769-011-0805-y

30 Huggett JF Foy CA Benes V Emslie K Garson JA Haynes R Hellemans J Kubista M Mueller RD Nolan T Pfaffl MW Shipley GL Vandesompele J Wittwer CT Bustin SA The Digital MIQE Guidelines Minimum Information for Publication of Quantitative Digital PCR Experiments Clin Chem 201359(6)892-902 httpsdoiorg101373clinchem2013206375

31 Wilson PJ Ellison SLR Extending digital PCR analysis by modelling quantification cycle data BMC Bioinfo 201617421 httpsdoiorg101186s12859-016-1275-3

32 Wang X Son A Effects of pretreatment on the denaturation and fragmentation of genomic DNA for DNA hybridization Environ Sci Processes Impacts 2013152204 httpsdoiorg101039c3em00457k

33 Nwokeoji AO Kilby PM Portwood DE Dickman MJ Accurate Quantification of Nucleic Acids Using Hypochromicity Measurements in Conjunction with UV Spectrophotometry Anal Chem 201789(24)13567-13574 https101021acsanalchem7b04000

34 Farkas DH Kaul KL Wiedbrauk DL Kiechle FL Specimen collection and storage for diagnostic molecular pathology investigation Arch Pathol Lab Med 1996120(6)591-6

35 Kypr J Kejnovskaacute I Renčiuk D Vorličkovaacute M Circular dichroism and conformational polymorphism of DNA Nucleic Acids Res 200937(6)1713-1725 httpsdoiorg101093nargkp026

36 Strachan T Read AP (1999) Human Molecular Genetics 2nd Ed John Wiley amp Sons NY

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

37 Bhat S McLaughlin JL Emslie KR Effect of sustained elevated temperature prior to amplification on template copy number estimation using digital polymerase chain reaction Analyst 2011136724ndash32 httpsdoiorg101039c0an00484g

38 NIST Consensus Builder (NICOB) httpsconsensusnistgov 39 Koepke A Lafarge T Toman B Possolo A NIST Consensus Builder Userrsquos Manual

httpsconsensusnistgovNISTConsensusBuilder-UserManualpdf

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

Appendix A DNA Extraction Method The following method has been adapted from Miller et al (1988)A1 Procedure

1 Transfer approximately 5 mL of buffy coat into a 50 mL sterile polypropylene centrifuge tube and add 30 mL Red Blood Cell Lysis Buffer

2 Mix by gentle inversion and let stand at room temperature for 15 min

3 Centrifuge at approximately 209 rads (2000 rpm) for 20 min

4 Discard supernatant

5 Add 9 mL of Nucleic Lysis Buffer and re-suspend the pelleted cells by agitating with a plastic transfer pipette

6 Add 300 microL of 20 sodium dodecyl sulfate (SDS) and 18 mL of Protease K Solution (freshly prepared 2 mg to 5 mg Protease K per 1 mL Protease K Buffer)

7 Mix by inversion followed by incubation overnight or longer at 37 ordmC to 40 ordmC in a shaking water bath The length of incubation is determined by the time required to solubilize the cells present in the tube

8 Add 4 mL saturated ammonium acetate solution (in deionized water stored at 2 ordmC to 8 ordmC) Shake vigorously for 15 s

9 Centrifuge at approximately 471 rads (4500 rpm) for 10 min

10 Pipette supernatant into a 50 mL sterile polypropylene centrifuge tube and add two volumes of room temperature absolute ethanol Mix by gentle inversion until the DNA aggregates and floats to the top of the tube

11 Spool the aggregated DNA from the current tube into a fresh tube of 80 ethanol using a sterile inoculation loop At this step multiple tubes of precipitated DNA may be pooled to make a large lot of the purified material Mix by gentle inversion

12 Transfer the aggregated DNA to a fresh dry 50 mL polypropylene centrifuge tube Remove as much ethanol as possible by squeezing the DNA against the wall of the tube with a sterile inoculation loop Air-dry the material in a laminar flow hood

13 Resolubilize DNA in TE-4 buffer and store material at 4 ordmC Reagents

Red Blood Cell Lysis Buffer 144 mmolL ammonium chloride 1 mmolL sodium bicarbonate in deionized (DI) water Store at 2 ordmC to 8 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

Nucleic Lysis Buffer 10 mmolL Tris-HCl 400 mmolL sodium chloride 20 mmolL disodium ethylenediaminetetraacetic acid in DI Water adjust pH to 82 Store at 2 ordmC to 8 ordmC

Protease K Buffer

20 mmolL disodium ethylenediaminetetraacetic acid 10 (mass per volume) Sodium Dodecyl Sulfate (SDS) in DI water Store at 2 ordmC to 8 ordmC

Protease K

Sigma-Aldrich cat P6556 Tritirachium album lyophilized powder Store at ndash20 ordmC

Ammonium Acetate

192 molL ammonium acetate (saturated) in DI water Store at 2 ordmC to 8 ordmC

Absolute Ethanol (200 proof)

Store at room temperature Reference

A1 Miller SA Dykes DD Polesky HF A simple salting out procedure for extracting DNA from human nucleated cells Nucleic Acids Res 198816(3)1215 httpsdoiorg101093nar1631215

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

48

Appendix B Spectrophotometric Measurement Protocol The following is for use with a scanning recording spectrophotometer The procedures include the spectrophotometer calibration measurement of the ldquonativerdquo dsDNA and measurement of the NaOH denatured ssDNA The procedure for evaluating solutions after treatment with NaOH is adapted from references89 Spectrophotometer Calibration

Establish Baseline Set the spectrophotometer to acquire data from 220 nm to 345 nm in steps of 01 nm Zero the instrument with nothing (air) in the sample and reference cells Acquire an air-air baseline of the instrument with nothing (air) in the sample and reference cells If the trace has significant structure or excessive noise take corrective action If the trace is acceptable proceed with data collection Enable Verification of Wavelength Calibration SRM 2034 is a solution of holmium oxide in 10 perchloric acid contained in a sealed quartz cuvette delivering an intrinsic UVVis wavelength standard for absorption spectrophotometryB1 Evaluation of the location of peak maxima in the spectra of this solution enables verification of the spectrophotometers wavelength calibration Place SRM 2034 in the sample cell Create and name a file to hold the verification spectra Perform three replicate SRM 2034-to-air scans saving each scan to the file Remove SRM 2034 from the spectrophotometer and store appropriately Enable Verification of Absorbance Calibration SRM 2031 consists of three metal-on-fused silica filters held in cuvette-sized filter holders plus an empty filter holder The three filters are certified for transmittance and transmittance density (-log10(transmittance) effectively absorbance) at several wavelengths with nominal wavelength-independent transmittances of 10 30 and 90 Evaluation of the absorbance of each of these three filters at the certified 250 nm 280 nm and 340 nm wavelengths enables verification of the spectrophotometers absorbance calibration Place the empty SRM 2031 filter holder in the reference cell Remove the protective covers from the 10 transmittance filter and place in the sample cell Perform three replicate scans saving each scan to file Remove the filter from the spectrophotometer replace the protective covers and store appropriately Repeat for the 30 and 90 filters Close the file containing the verification spectra For ldquonativerdquo dsDNA Samples

Verify Optical Balance of Cuvettes The measurements of the DNA-containing samples are made using a 100 microL quartz cuvette in the sample cell compared to an optically matched cuvette in the reference cell The optical equivalence of these two cuvettes can be verified by filling both cuvettes with appropriate volumes of a sample solution matrix (TE-4 buffer) that does not contain DNA Fill two dry

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

49

clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes followed by a methanol rinse in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquonativerdquo dsDNA remove and clean both the reference cuvette and the sample cuvette For NaOH denatured ssDNA Samples

Prepare a 04 molL NaOH Stock Solution To a 15 mL disposable centrifuge tube add 016 g NaOH pellets add deionized (DI) water to the 10 mL line and mix until all the NaOH pellets have dissolved Centrifuge at 524 rads (5000 rpm) for 5 min For every 9 samples to be analyzed aliquot 1 mL of this 04 molL NaOH stock solution into a 15 mL screw top tube Use these tubes to prepare the NaOH blanks and DNA samples In a 15 mL screw cap vial mix 200 microL TE-4 buffer with 200 microL 04 molL sodium hydroxide solution to achieve a final NaOH concentration of 02 molL Vortex 5 s then centrifuge at 1257 rads (12 000 rpm) for 1 min Fill two dry clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Prepare the Sample Solutions For each unit of material to be measured (Tube or vial)

a) Vortex the unit for 5 s b) Centrifuge at 1257 rads (12 000 rpm) for 1 min c) Remove a 100 microL aliquot and place in a new labeled vial d) Add 100 microL of the freshly prepared 04 molL NaOH solution (with the same pipet) e) Vortex 5 s f) Centrifuge at 1257 rads (12 000 rpm) for 1 min

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

50

Measure Sample Solutions Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes rinse with methanol in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the NaOH denatured DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquodenaturedrdquo ssDNA close the file holding the spectra for this series Remove and clean both the reference cuvette and the sample cuvette After All Sample Measurements are Complete

Enable Verification of Spectrophotometer Stability The stability of the spectrophotometer over the course of the data collection can be evaluated by comparing spectra for SRM 2034 and 2031 taken at the start of data collection with spectra taken at the end of data collection Repeat the wavelength and absorbance verification measurements described above collecting the spectra in a suitably named file Clean Up Transfer all relevant data files from the spectrophotometers storage to a transfer device Clean the spectrophotometer work area as necessary and restore the SRMs and cuvettes to their appropriate storage locations Evaluation

Export the files to a spreadsheet Average the replicate spectra Subtract the average absorbance at A320 from the absorbance at A260 resulting in the corrected absorbance at A260 The conventional spectrophotometric estimate of the mass concentration of the dsDNA in the sample is achieved by multiplying the corrected absorbance by 50 The conventional estimate for the NaOH-treated ssDNA is achieved by multiplying the corrected absorbance by 37times2 where the ldquotimes2rdquo adjusts for the 1rarr2 volumetric dilution of the sample Note reliable spectrophotometric measurements require corrected A260 to be greater than 005 Reference

B1 Travis JC Acosta JC Andor G Bastie J Blattner P Chunnilall CJ Crosson SC Duewer DL Early EA Hengstberger F Kim C-S Liedquist L Manoocheri F Mercader F Metzdorf J Mito A Monardh LAG Nevask S Nilsson M Noeumll M Rodriguez AC Ruiacutez A Schirmacher A Smith MV Valencia G van Tonder N Zwinkels J Intrinsic Wavelength Standard Absorption Bands in Holmium Oxide Solution for UVvisible Molecular Absorption Spectrophotometry J Phys Chem Ref Data 200534(1)41-56

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

51

Appendix C SRM 2372a Statisticianrsquos report Extracted from Blaza Toman

November 30 2017 1 Introduction Candidate material for SRM 2372a is composed of three components labeled A B and C All components are human genomic DNA and prepared at NIST from buffy coat white blood cells from single-source anonymous donors Component A and B are single source (malefemale respectively) and Component C is a 1 male to 3 female mixture of two single-source donors Droplet digital PCR (ddPCR) assays were used to obtain measurements of the concentration of copies per microliter of sample Ten NIST-developed human nuclear DNA and three literature-derived human mitochondrial DNA PCR assays were used in this study

11 OpenBUGS Analysis All parameters in the following sections were estimated using Markov Chain Monte Carlo (MCMC) implemented in the OpenBUGS freewareC1 OpenBUGS is a software application for the Bayesian analysis of complex statistical models

12 OpenBUGS Exemplar Output All following sections list the OpenBUGs model (ldquocoderdquo) initialization values (ldquoinitsrdquo) if they are needed and the data used to evaluate the various data sets discussed in the body of this report homogeneity stability nDNA certification and mtDNAnDNA value assignment The provided information in each of the following sections is sufficient to produce estimates but the estimates may not exactly reproduce the exemplar results provided The MCMC paradigm does not produce ldquoanalyticalrdquo values but approaches them asymptotically as the number o model updates increases All exemplar results presented here are based upon the following bull 10-fold ldquothinningrdquo to minimize autocorrelation there were 10 MCMC estimates for every

model update used in parameter estimation bull a ldquoburn inrdquo of 5 000 updates before beginning parameter estimation and bull 10 000 model updates

The following statistics are provided in the tables listing the exemplar results

mean arithmetic mean of the parameterrsquos model update values sd standard deviation of the parameterrsquos model update values MC_error Monte Carlo standard deviation of the mean meanradic(number of model updates) val25pc 25th percentile of the model update values median 50th percentile of the model update values val975pc 975th percentile of the model update values startnumber of ldquoburn inrdquo updates sample number of model update values

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

52

2 Homogeneity assessment 21 Data

Samples of the material of all three components were drawn from 22 different boxes Two assays (NEIF ND6) were used to obtain 3 replicates of the measurement in copies per nanoliter for each sample and component For the complete homogeneity data see

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

22 Model The data were analyzed using a three-level Gaussian hierarchical ANOVA model

119910119910119894119894119894119894119894119894~119873119873120572120572119894119894119894119894119894119894 120590120590119908119908119894119894119905119905ℎ119894119894119894119894 119887119887119887119887119887119887 1198941198941198941198942

where ~N() denotes ldquodistributed as a normal distribution of some mean μ and variancerdquo 119894119894 denotes component (1 for A 2 for B 3 for C) 119895119895 denotes assay (1 for NEIF 2 for ND6) 119896119896 denotes box (1hellip22) and 120572120572119894119894119894119894119894119894~119873119873micro119894119894119894119894120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119887119887119887119887119887119887119894119894

2

All variance components were given a Half Cauchy prior as described in Reference C2 Note The Half Cauchy prior is specified by ldquo~dnorm(0 00016)I(0001)rdquo The ldquoIrdquo notation is for censoring It cuts off any random draws from the Normal distribution that are lower than 0001 An informative prior is needed when there are less than four data per distribution and the half Cauchy is a good general-purpose prior

23 OpenBUGS code and data Homogeneity code ModelBegin for(i in 13)xiNsb[i] ~ dnorm(0 00016)I(0001) chSqNsb[i] ~ dgamma(0505) sigb[i] lt- xiNsb[i]sqrt(chSqNsb[i]) for(i in 13)for(j in 12)mu[ij]~dnorm(010E-5) for(i in 13) muA[i]lt-mean(mu[i])sigbP[i]lt-100(muA[i]sqrt(sigb[i])) for(i in 13)for(j in 12)xiNs[ij] ~ dnorm(0 00016)I(0001) chSqNs[ij]~dgamma(0505) sigt[ij]lt-xiNs[ij]sqrt(chSqNs[ij]) sigtP[ij]lt-100(mu[ij]sqrt(sigt[ij])) for(i in 13)for(j in 12)for(k in 122)alpha[ijk]~dnorm(mu[ij]sigb[i]) for(idx in 1393)lamb[idx]~dnorm(alpha[cmp[idx]asy[idx]box[idx]]sigt[cmp[idx]asy[idx]]) ModelEnd Inits None required

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

53

Homogeneity Data lamb[] asy[] box[] cmp[] rep[] 1632 1 1 1 1 1529 1 1 1 2 1511 1 1 1 3 1616 2 1 1 1 1640 2 1 1 2 1582 2 1 1 3 1531 1 2 1 1 1525 1 2 1 2 1549 1 2 1 3 1615 2 2 1 1 1589 2 2 1 2 1601 2 2 1 3 1524 1 3 1 1 1540 1 3 1 2 1533 1 3 1 3 1582 2 3 1 1 1152 2 3 1 2 1555 2 3 1 3 1454 1 4 1 1 1458 1 4 1 2 1427 1 4 1 3 1627 2 4 1 1 1605 2 4 1 2 1607 2 4 1 3 1465 1 5 1 1 1540 1 5 1 2 1504 1 5 1 3 1607 2 5 1 1 1633 2 5 1 2 1586 2 5 1 3 1512 1 6 1 1 1491 1 6 1 2 1491 1 6 1 3 1511 2 6 1 1 1483 2 6 1 2 1524 2 6 1 3 1418 1 7 1 1 1480 1 7 1 2 1455 1 7 1 3 1479 2 7 1 1 1557 2 7 1 2 1656 2 7 1 3 1496 1 8 1 1 1560 1 8 1 2 1501 1 8 1 3 1514 2 8 1 1 1615 2 8 1 2 1540 2 8 1 3 1588 1 9 1 1 1581 1 9 1 2 1637 1 9 1 3 1474 2 9 1 1 1438 2 9 1 2 1454 2 9 1 3 1574 1 10 1 1 1513 1 10 1 2 1570 1 10 1 3 1594 2 10 1 1 1713 2 10 1 2 1669 2 10 1 3

1557 1 11 1 1 1508 1 11 1 2 1633 1 11 1 3 1571 2 11 1 1 1611 2 11 1 2 1631 2 11 1 3 1500 1 12 1 1 1476 1 12 1 2 1515 1 12 1 3 1626 2 12 1 1 1586 2 12 1 2 1608 2 12 1 3 1672 1 13 1 1 1634 1 13 1 2 1804 1 13 1 3 1490 2 13 1 1 1516 2 13 1 2 1497 2 13 1 3 1551 1 14 1 1 1568 1 14 1 2 1538 1 14 1 3 1617 2 14 1 1 1614 2 14 1 2 1639 2 14 1 3 1456 1 15 1 1 1448 1 15 1 2 1523 1 15 1 3 1525 2 15 1 1 1402 2 15 1 2 1436 2 15 1 3 1573 1 16 1 1 1520 1 16 1 2 1617 1 16 1 3 1634 2 16 1 1 1627 2 16 1 2 1575 2 16 1 3 1516 1 17 1 1 1516 1 17 1 2 1628 1 17 1 3 1613 2 17 1 1 1567 2 17 1 2 1567 2 17 1 3 1612 1 18 1 1 1535 1 18 1 2 1672 1 18 1 3 1529 2 18 1 1 1495 2 18 1 2 1528 2 18 1 3 1598 1 19 1 1 1514 1 19 1 2 1596 1 19 1 3 1571 2 19 1 1 1535 2 19 1 2 1511 2 19 1 3 1575 1 20 1 1 1568 1 20 1 2 1586 1 20 1 3 1652 2 20 1 1 1545 2 20 1 2 1627 2 20 1 3 1658 1 21 1 1

1611 1 21 1 2 1578 1 21 1 3 1647 2 21 1 1 1607 2 21 1 2 1612 2 21 1 3 1547 1 22 1 1 1528 1 22 1 2 1532 1 22 1 3 1537 2 22 1 1 1513 2 22 1 2 1536 2 22 1 3 1790 1 1 2 1 1909 1 1 2 2 1885 1 1 2 3 1870 2 1 2 2 1860 2 1 2 3 1712 1 2 2 1 1719 1 2 2 2 1704 1 2 2 3 1775 2 2 2 1 1718 2 2 2 2 1756 2 2 2 3 1693 1 3 2 1 1755 1 3 2 2 1749 1 3 2 3 1824 2 3 2 1 1483 2 3 2 2 1896 2 3 2 3 1754 1 4 2 1 1745 1 4 2 2 1739 1 4 2 3 1670 2 4 2 1 1632 2 4 2 2 1814 2 4 2 3 1757 1 5 2 1 1687 1 5 2 2 1735 1 5 2 3 1683 2 5 2 1 1759 2 5 2 2 1703 2 5 2 3 1697 1 6 2 1 1709 1 6 2 2 1714 1 6 2 3 1693 2 6 2 1 1723 2 6 2 2 1722 2 6 2 3 1722 1 7 2 1 1625 1 7 2 2 1703 1 7 2 3 1690 2 7 2 1 1684 2 7 2 2 1688 2 7 2 3 1884 1 8 2 1 1802 1 8 2 2 1772 1 8 2 3 1864 2 8 2 1 1887 2 8 2 2 1821 2 8 2 3 1713 1 9 2 1 1734 1 9 2 2 1727 1 9 2 3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

54

1709 2 9 2 1 1756 2 9 2 2 1758 2 9 2 3 1759 1 10 2 1 1809 1 10 2 2 1738 1 10 2 3 1764 2 10 2 1 1816 2 10 2 2 1787 2 10 2 3 1749 1 11 2 1 1782 1 11 2 2 1740 1 11 2 3 1741 2 11 2 1 1792 2 11 2 2 1784 2 11 2 3 1748 1 12 2 1 1712 1 12 2 2 1718 1 12 2 3 1797 2 12 2 1 1614 2 12 2 2 1790 2 12 2 3 1632 1 13 2 1 1668 1 13 2 2 1701 1 13 2 3 1626 2 13 2 1 1566 2 13 2 2 1681 2 13 2 3 1752 1 14 2 1 1535 1 14 2 2 1673 1 14 2 3 1724 2 14 2 1 1741 2 14 2 2 1685 2 14 2 3 1871 1 15 2 1 1835 1 15 2 2 1887 1 15 2 3 1770 2 15 2 1 1752 2 15 2 2 1798 2 15 2 3 1688 1 16 2 1 1725 1 16 2 2 1743 1 16 2 3 1744 2 16 2 1 1697 2 16 2 2 1710 2 16 2 3 1789 1 17 2 1 1769 1 17 2 2 1825 1 17 2 3 1716 2 17 2 1 1762 2 17 2 2 1638 2 17 2 3 1826 1 18 2 1 1780 1 18 2 2 1849 1 18 2 3 1805 2 18 2 1 1801 2 18 2 2 1931 2 18 2 3 1751 1 19 2 1 1759 1 19 2 2 1880 1 19 2 3 1827 2 19 2 1 1766 2 19 2 2

1786 2 19 2 3 1738 1 20 2 1 1678 1 20 2 2 1768 1 20 2 3 1752 2 20 2 1 1747 2 20 2 2 1768 2 20 2 3 1742 1 21 2 1 1836 1 21 2 2 1713 1 21 2 3 1746 2 21 2 1 1744 2 21 2 2 1764 2 21 2 3 1761 1 22 2 1 1743 1 22 2 2 1660 1 22 2 3 1666 2 22 2 1 1692 2 22 2 2 1819 2 22 2 3 1493 1 1 3 1 1385 1 1 3 2 1510 1 1 3 3 1565 2 1 3 1 1562 2 1 3 2 1466 2 1 3 3 1516 1 2 3 1 1513 1 2 3 2 1577 1 2 3 3 1490 2 2 3 1 1547 2 2 3 2 1570 2 2 3 3 1451 1 3 3 1 1528 1 3 3 2 1506 1 3 3 3 1485 2 3 3 1 1505 2 3 3 2 1471 2 3 3 3 1440 1 4 3 1 1486 1 4 3 2 1520 1 4 3 3 1529 2 4 3 1 1508 2 4 3 2 1506 2 4 3 3 1523 1 5 3 1 1472 1 5 3 2 1483 1 5 3 3 1513 2 5 3 1 1520 2 5 3 2 1491 2 5 3 3 1437 1 6 3 1 1493 1 6 3 2 1522 1 6 3 3 1492 2 6 3 1 1481 2 6 3 2 1461 2 6 3 3 1485 1 7 3 1 1477 1 7 3 2 1463 1 7 3 3 1475 2 7 3 1 1547 2 7 3 2 1485 2 7 3 3 1433 1 8 3 1

1405 1 8 3 2 1478 1 8 3 3 1497 2 8 3 1 1501 2 8 3 2 1495 2 8 3 3 1463 1 9 3 1 1499 1 9 3 2 1505 1 9 3 3 1503 2 9 3 1 1491 2 9 3 2 1516 2 9 3 3 1534 1 10 3 1 1458 1 10 3 2 1454 1 10 3 3 1448 2 10 3 1 1465 2 10 3 2 1474 2 10 3 3 1533 1 11 3 1 1514 1 11 3 2 1555 1 11 3 3 1488 2 11 3 1 1470 2 11 3 2 1483 2 11 3 3 1450 1 12 3 1 1525 1 12 3 2 1497 1 12 3 3 1496 2 12 3 1 1483 2 12 3 2 1519 2 12 3 3 1537 1 13 3 1 1527 1 13 3 2 1515 1 13 3 3 1499 2 13 3 1 1479 2 13 3 2 1475 2 13 3 3 1515 1 14 3 1 1490 1 14 3 2 1494 1 14 3 3 1469 2 14 3 1 1480 2 14 3 2 1466 2 14 3 3 1511 1 15 3 1 1568 1 15 3 2 1594 1 15 3 3 1515 2 15 3 1 1543 2 15 3 2 1565 2 15 3 3 1449 1 16 3 1 1452 1 16 3 2 1471 1 16 3 3 1543 2 16 3 1 1448 2 16 3 2 1491 2 16 3 3 1388 1 17 3 1 1421 1 17 3 2 1394 1 17 3 3 1452 2 17 3 1 1435 2 17 3 3 1475 1 18 3 1 1455 1 18 3 2 1445 1 18 3 3 1458 2 18 3 1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

55

1531 2 18 3 2 1459 2 18 3 3 1468 1 19 3 1 1525 1 19 3 2 1446 1 19 3 3 1435 2 19 3 1 1478 2 19 3 2 1421 2 19 3 3 1549 1 20 3 1

1531 1 20 3 2 1499 1 20 3 3 1499 2 20 3 1 1465 2 20 3 2 1468 2 20 3 3 1530 1 21 3 1 1440 1 21 3 2 1484 1 21 3 3 1499 2 21 3 1

1454 2 21 3 2 1442 1 22 3 1 1487 1 22 3 2 1436 1 22 3 3 1589 2 22 3 1 1591 2 22 3 2 1556 2 22 3 3 END

Table C-1 Example OpenBUGS Homogeneity Summary

mean sd MC_error val25pc median val975pc start sample mu[11] 15460 0122 000123 1522 1546 1570 5001 10000 mu[12] 15640 0137 000141 1537 1564 1591 5001 10000 mu[21] 17490 0113 000118 1727 1749 1772 5001 10000 mu[22] 17480 0129 000125 1722 1748 1774 5001 10000 mu[31] 14870 0073 000073 1473 1487 1501 5001 10000 mu[32] 14960 0069 000063 1482 1496 1509 5001 10000 muA[1] 15550 0092 000092 1536 1555 1573 5001 10000 muA[2] 17490 0085 000086 1732 1749 1765 5001 10000 muA[3] 14910 0050 000046 1481 1491 1501 5001 10000 sigbP[1] 3269 04985 000514 2376 3244 4324 5001 10000 sigbP[2] 2548 04385 000461 1733 2533 3472 5001 10000 sigbP[3] 1850 0292 000260 1329 1832 2477 5001 10000

sigtP[11] 2774 0298 000260 2268 2750 3442 5001 10000 sigtP[12] 4036 0429 000438 3290 4005 4986 5001 10000 sigtP[21] 2664 0301 000321 2155 2640 3313 5001 10000 sigtP[22] 3807 0395 000377 3116 3772 4662 5001 10000 sigtP[31] 2313 0250 000209 1880 2291 2873 5001 10000 sigtP[32] 1812 0195 000170 1473 1797 2243 5001 10000

24 Results The material appears to be homogeneous as can be seen in Figure C-1 Based on these results the between-tube variability is of the same order as within-tube variability

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

56

Component A Assay NEIF Component A Assay ND6 box plot gamma[11]

[111][112][113]

[114]

[115][116]

[117]

[118]

[119]

[1110][1111]

[1112]

[1113]

[1114]

[1115]

[1116][1117]

[1118]

[1119][1120]

[1121]

[1122]

114

box plot gamma[12]

[121][122]

[123]

[124][125]

[126]

[127][128]

[129]

[1210]

[1211][1212]

[1213]

[1214]

[1215]

[1216]

[1217]

[1218][1219]

[1220][1221]

[1222]

115

Component B Assay NEIF Component B Assay ND6

box plot gamma[21]

[211]

[212]

[213][214]

[215][216]

[217]

[218]

[219]

[2110][2111]

[2112]

[2113][2114]

[2115]

[2116]

[2117]

[2118]

[2119]

[2120]

[2121]

[2122]

129

box plot gamma[22]

[221]

[222][223]

[224][225][226][227]

[228]

[229]

[2210][2211]

[2212]

[2213]

[2214]

[2215]

[2216][2217]

[2218]

[2219]

[2220][2221][2222]

128

Component C Assay NEIF Component C Assay ND6

box plot gamma[31]

[311]

[312]

[313][314]

[315][316]

[317]

[318]

[319][3110]

[3111]

[3112]

[3113]

[3114]

[3115]

[3116]

[3117]

[3118]

[3119]

[3120]

[3121]

[3122]109

box plot gamma[32]

[321][322]

[323]

[324][325]

[326]

[327][328][329]

[3210]

[3211]

[3212]

[3213][3214]

[3215]

[3216]

[3217]

[3218]

[3219]

[3220][3221]

[3222]

11

Figure C-1 Homogeneity Assessment

Note These results were obtained using the 1-4 gravimetric dilution-adjusted λ but uncorrected for droplet volume

or the 1-10 volumetric dilution That is the vertical axis in each panel should be multiplied by 07349times10 = 7349 This constant difference in scale does not impact the homogeneity analysis Error bars span approximate 95 confidence intervals

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

57

3 Certification measurements 31 Data

Originally the experiment consisted of 5 plates containing the 3 components (A B C) and 12 assays Ten of the assays were for genomic DNA identification (for the certified value) and two of the assays were for mitochondrial detection which is used for informational values (and are not included in this data set) Each of the plates contained samples from a new vial from different boxes (noted in the ldquovial usedrdquo portion the number references the box) The assay positions were alternated across the five plates as well For the plate design and which tubes were used see

Figure 6 Basic Plate Layout Design for Certification Measurements It was later determined that the homogeneity and stability data sets were obtained using a different set of pipettes than the certification data set and since all data was to be combined for certification another plate was obtained using the original pipette For the complete certification data see

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component Stability data was measured at different temperatures over several weeks using assays NEIF and ND6 Only the 4degC data was also used for certification For the complete stability data see

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

32 Model It was decided that all three sets of data should be used to derive the certified values In this way the uncertainty of the certified values would better reflect any potential heterogeneity in the samples as well as any additional unknown factors that may have effected the results Because two different pipettes were used to obtain different sets of data there was a significant bias in some of the measurements Specifically the ldquocertificationrdquo data from the first five plates was obtained using pipette set ldquo1rdquo and the remaining measurements were obtained using pipette set ldquo2rdquo This bias was introduced into the statistical model as part of the mean Possible effects of different assays and type of data (heterogeneity certification and stability) were accounted for in the statistical model in terms of a random effect Specifically the response for each component (A B C) was represented as

119910119910119894119894119894119894119894119894~1198731198731205741205741198941198941198941198941198941198941205901205901198941198942 where ~N() denotes ldquodistributed as a normal distribution of some mean and variancerdquo i = 123 denotes type of measurement heterogeneity certification and stability j = 1hellip10 denotes assay 2PR4 POTP NEIF R4Q5 D5 ND6 D9 HBB1 ND14 22C3 k = 1 2 denotes pipette set 120574120574119894119894119894119894119894119894 = 120572120572119894119894119894119894 + 120573120573119894119894 120572120572119894119894119894119894~119873119873(micro1198941198941205901205901205721205722) 120573120573119894119894 = 120573120573 for data type 1 and 120573120573119894119894 = 0 for data types 2 and 3 For a given 2372a component the variance 1205901205901205721205722 measures the between assay variability and 1205901205901198941198942 measures the repeatability variability for each data type The data was analyzed using Bayesian MCMC programmed in OpenBUGSC2 with non-informative Gaussian priors for the means and Gamma priors for the variance components

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

58

33 OpenBUGS code inits and data Certification code ModelBegin for(i in 13)mu[i]~dnorm(010E-5) sig[i]~dgamma(10E-510E-5) sigalph~dgamma(10E-510E-5) for(i in 110)alpha[1i]~dnorm(mu[1] sigalph) for(i in 12)alpha[2i]~dnorm(mu[2] sigalph) alpha[3i]~dnorm(mu[3]sigalph) b~dnorm(010E-5) beta[1]lt-b beta[2]lt-00 for(i in 1Ntot)mean[i]lt-alpha[typ[i]asy[i]]+beta[pip[i]] for(i in 1Ntot)lamb[i]~dnorm(mean[i]sig[typ[i]]) ModelEnd Certification Inits list(sig=c(111) sigalph=1) Certification Data for Component A list(Ntot=275) lamb[] asy[] cmp[] typ[] pip[] 1667 1 1 1 1 1655 2 1 1 1 1647 3 1 1 1 1594 4 1 1 1 1655 5 1 1 1 1629 6 1 1 1 1655 7 1 1 1 1644 8 1 1 1 1645 9 1 1 1 1545 10 1 1 1 1698 1 1 1 1 1685 2 1 1 1 1665 3 1 1 1 1658 4 1 1 1 1683 5 1 1 1 1647 6 1 1 1 1652 7 1 1 1 1611 8 1 1 1 1707 9 1 1 1 1603 10 1 1 1 1668 1 1 1 1 1587 2 1 1 1 1678 3 1 1 1 1628 4 1 1 1 1738 5 1 1 1 1667 6 1 1 1 1695 7 1 1 1 1757 8 1 1 1 1771 9 1 1 1 1762 10 1 1 1 1739 1 1 1 1 1702 2 1 1 1 1614 3 1 1 1 1762 4 1 1 1 1649 5 1 1 1 1678 6 1 1 1 1653 7 1 1 1

175 8 1 1 1 1764 9 1 1 1 1759 10 1 1 1 1638 1 1 1 1 1676 2 1 1 1 1616 3 1 1 1 1559 4 1 1 1 1608 5 1 1 1 1611 6 1 1 1 1517 7 1 1 1 1557 8 1 1 1 1604 9 1 1 1 1675 10 1 1 1 1667 1 1 1 1 1638 2 1 1 1 1624 3 1 1 1 1587 4 1 1 1 1652 5 1 1 1 1589 6 1 1 1 1607 7 1 1 1 1582 8 1 1 1 1625 9 1 1 1 1616 10 1 1 1 1888 1 1 1 1 1868 2 1 1 1 1953 3 1 1 1 1779 4 1 1 1 1924 5 1 1 1 1861 6 1 1 1 1824 7 1 1 1 1809 8 1 1 1 1864 9 1 1 1 193 10 1 1 1 1939 1 1 1 1 1815 2 1 1 1 1902 3 1 1 1 1871 4 1 1 1 2124 5 1 1 1

1787 6 1 1 1 1844 7 1 1 1 1799 8 1 1 1 1857 9 1 1 1 1942 10 1 1 1 1597 1 1 1 1 1533 2 1 1 1 1462 3 1 1 1 1493 4 1 1 1 1538 5 1 1 1 1492 6 1 1 1 1528 7 1 1 1 1548 8 1 1 1 1513 9 1 1 1 1562 10 1 1 1 159 1 1 1 1 1404 2 1 1 1 1533 3 1 1 1 1519 4 1 1 1 1586 5 1 1 1 1516 6 1 1 1 1538 7 1 1 1 1581 8 1 1 1 1514 9 1 1 1 1431 1 1 1 2 1407 2 1 1 2 1487 3 1 1 2 145 4 1 1 2 1394 5 1 1 2 1396 6 1 1 2 1414 7 1 1 2 1401 8 1 1 2 1412 9 1 1 2 1435 10 1 1 2 1411 1 1 1 2 1395 2 1 1 2 1398 3 1 1 2 1439 4 1 1 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

59

1422 5 1 1 2 1393 6 1 1 2 139 7 1 1 2 1382 8 1 1 2 146 9 1 1 2 141 10 1 1 2 1632 1 1 2 2 1531 1 1 2 2 1524 1 1 2 2 1454 1 1 2 2 1465 1 1 2 2 1512 1 1 2 2 1418 1 1 2 2 1496 1 1 2 2 1588 1 1 2 2 1574 1 1 2 2 1557 1 1 2 2 15 1 1 2 2 1672 1 1 2 2 1551 1 1 2 2 1573 1 1 2 2 1516 1 1 2 2 1612 1 1 2 2 1598 1 1 2 2 1575 1 1 2 2 1658 1 1 2 2 1547 1 1 2 2 1529 1 1 2 2 1525 1 1 2 2 154 1 1 2 2 1458 1 1 2 2 154 1 1 2 2 1491 1 1 2 2 148 1 1 2 2 156 1 1 2 2 1581 1 1 2 2 1513 1 1 2 2 1508 1 1 2 2 1476 1 1 2 2 1634 1 1 2 2 1568 1 1 2 2 152 1 1 2 2 1516 1 1 2 2 1535 1 1 2 2 1514 1 1 2 2 1568 1 1 2 2 1611 1 1 2 2 1528 1 1 2 2 1511 1 1 2 2 1549 1 1 2 2 1533 1 1 2 2 1427 1 1 2 2 1504 1 1 2 2 1491 1 1 2 2 1455 1 1 2 2

1501 1 1 2 2 1637 1 1 2 2 157 1 1 2 2 1633 1 1 2 2 1515 1 1 2 2 1804 1 1 2 2 1538 1 1 2 2 1617 1 1 2 2 1628 1 1 2 2 1672 1 1 2 2 1596 1 1 2 2 1586 1 1 2 2 1578 1 1 2 2 1532 1 1 2 2 1616 2 1 2 2 1615 2 1 2 2 1582 2 1 2 2 1627 2 1 2 2 1607 2 1 2 2 1511 2 1 2 2 1479 2 1 2 2 1514 2 1 2 2 1474 2 1 2 2 1594 2 1 2 2 1571 2 1 2 2 1626 2 1 2 2 149 2 1 2 2 1617 2 1 2 2 1634 2 1 2 2 1613 2 1 2 2 1529 2 1 2 2 1571 2 1 2 2 1652 2 1 2 2 1647 2 1 2 2 1537 2 1 2 2 164 2 1 2 2 1589 2 1 2 2 1152 2 1 2 2 1605 2 1 2 2 1633 2 1 2 2 1483 2 1 2 2 1557 2 1 2 2 1615 2 1 2 2 1438 2 1 2 2 1713 2 1 2 2 1611 2 1 2 2 1586 2 1 2 2 1516 2 1 2 2 1614 2 1 2 2 1627 2 1 2 2 1567 2 1 2 2 1495 2 1 2 2 1535 2 1 2 2 1545 2 1 2 2 1607 2 1 2 2

1513 2 1 2 2 1582 2 1 2 2 1601 2 1 2 2 1555 2 1 2 2 1607 2 1 2 2 1586 2 1 2 2 1524 2 1 2 2 1656 2 1 2 2 154 2 1 2 2 1454 2 1 2 2 1669 2 1 2 2 1631 2 1 2 2 1608 2 1 2 2 1497 2 1 2 2 1639 2 1 2 2 1575 2 1 2 2 1567 2 1 2 2 1528 2 1 2 2 1511 2 1 2 2 1627 2 1 2 2 1612 2 1 2 2 1536 2 1 2 2 1568 1 1 3 2 152 1 1 3 2 1486 1 1 3 2 1441 1 1 3 2 1706 1 1 3 2 144 1 1 3 2 15 1 1 3 2 1537 1 1 3 2 1387 1 1 3 2 1463 1 1 3 2 157 1 1 3 2 1429 1 1 3 2 1667 1 1 3 2 1632 1 1 3 2 1589 1 1 3 2 1474 2 1 3 2 1543 2 1 3 2 1555 2 1 3 2 1599 2 1 3 2 1494 2 1 3 2 1624 2 1 3 2 1577 2 1 3 2 1593 2 1 3 2 1479 2 1 3 2 1517 2 1 3 2 1538 2 1 3 2 159 2 1 3 2 1479 2 1 3 2 1479 2 1 3 2 1522 2 1 3 2 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

60

Table C-2 Example OpenBUGS Certification Summary for Component A

mean sd MC_error val25pc median val975pc start sample b 2584 02948 0005376 2001 2588 3158 100001 100000 mu[1] 1417 02706 0005366 1364 1416 147 100001 100000 mu[2] 1559 008368 4776E-4 1543 1559 1576 100001 100000 mu[3] 1533 01504 0001419 1504 1533 1563 100001 100000

Certification Data for Component B list(Ntot=278) lamb[] asy[] cmp[] typ[] pip[] 1869 1 2 1 1 185 2 2 1 1 1709 3 2 1 1 1856 4 2 1 1 1848 5 2 1 1 1828 6 2 1 1 1861 7 2 1 1 1825 8 2 1 1 1871 9 2 1 1 1776 10 2 1 1 196 1 2 1 1 1781 2 2 1 1 1707 3 2 1 1 1787 4 2 1 1 1829 5 2 1 1 1826 6 2 1 1 1858 7 2 1 1 1869 8 2 1 1 1773 9 2 1 1 1902 10 2 1 1 2174 1 2 1 1 2202 2 2 1 1 2172 3 2 1 1 223 5 2 1 1 2094 6 2 1 1 1949 7 2 1 1 1971 8 2 1 1 2123 9 2 1 1 2107 10 2 1 1 2231 1 2 1 1 2292 2 2 1 1 2216 3 2 1 1 2078 4 2 1 1 2192 5 2 1 1 2034 6 2 1 1 211 7 2 1 1 2123 8 2 1 1 2066 9 2 1 1 2046 10 2 1 1 181 1 2 1 1 1865 2 2 1 1 1806 3 2 1 1 1878 4 2 1 1 1824 5 2 1 1 1768 6 2 1 1 1811 7 2 1 1 1763 8 2 1 1 1865 9 2 1 1 1815 10 2 1 1

1806 1 2 1 1 1801 2 2 1 1 1786 3 2 1 1 1768 4 2 1 1 179 5 2 1 1 167 6 2 1 1 1749 7 2 1 1 1741 8 2 1 1 1781 9 2 1 1 1866 10 2 1 1 1784 1 2 1 1 1735 2 2 1 1 1778 3 2 1 1 1835 4 2 1 1 1794 5 2 1 1 17 6 2 1 1 1799 7 2 1 1 1765 8 2 1 1 1754 9 2 1 1 1792 10 2 1 1 1821 1 2 1 1 1813 2 2 1 1 1813 3 2 1 1 1749 4 2 1 1 1848 5 2 1 1 1711 6 2 1 1 1779 7 2 1 1 1769 8 2 1 1 1725 9 2 1 1 1814 10 2 1 1 1718 1 2 1 1 1753 2 2 1 1 1773 3 2 1 1 1777 4 2 1 1 1803 5 2 1 1 1718 6 2 1 1 1802 7 2 1 1 1834 8 2 1 1 1731 9 2 1 1 1842 10 2 1 1 1768 1 2 1 1 1698 2 2 1 1 1792 3 2 1 1 174 4 2 1 1 1795 5 2 1 1 1693 6 2 1 1 1743 7 2 1 1 1745 8 2 1 1 1714 9 2 1 1 1791 1 2 1 2

1746 2 2 1 2 1803 3 2 1 2 1767 4 2 1 2 1745 5 2 1 2 1824 6 2 1 2 1752 7 2 1 2 1813 8 2 1 2 1648 9 2 1 2 1818 10 2 1 2 177 1 2 1 2 182 2 2 1 2 1803 3 2 1 2 1739 4 2 1 2 1773 5 2 1 2 1729 6 2 1 2 1744 7 2 1 2 1754 8 2 1 2 1671 9 2 1 2 1741 10 2 1 2 179 1 2 2 2 1712 1 2 2 2 1693 1 2 2 2 1754 1 2 2 2 1757 1 2 2 2 1697 1 2 2 2 1722 1 2 2 2 1884 1 2 2 2 1713 1 2 2 2 1759 1 2 2 2 1749 1 2 2 2 1748 1 2 2 2 1632 1 2 2 2 1752 1 2 2 2 1871 1 2 2 2 1688 1 2 2 2 1789 1 2 2 2 1826 1 2 2 2 1751 1 2 2 2 1738 1 2 2 2 1742 1 2 2 2 1761 1 2 2 2 1909 1 2 2 2 1719 1 2 2 2 1755 1 2 2 2 1745 1 2 2 2 1687 1 2 2 2 1709 1 2 2 2 1625 1 2 2 2 1802 1 2 2 2 1734 1 2 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

61

1809 1 2 2 2 1782 1 2 2 2 1712 1 2 2 2 1668 1 2 2 2 1535 1 2 2 2 1835 1 2 2 2 1725 1 2 2 2 1769 1 2 2 2 178 1 2 2 2 1759 1 2 2 2 1678 1 2 2 2 1836 1 2 2 2 1743 1 2 2 2 1885 1 2 2 2 1704 1 2 2 2 1749 1 2 2 2 1739 1 2 2 2 1735 1 2 2 2 1714 1 2 2 2 1703 1 2 2 2 1772 1 2 2 2 1727 1 2 2 2 1738 1 2 2 2 174 1 2 2 2 1718 1 2 2 2 1701 1 2 2 2 1673 1 2 2 2 1887 1 2 2 2 1743 1 2 2 2 1825 1 2 2 2 1849 1 2 2 2 188 1 2 2 2 1768 1 2 2 2 1713 1 2 2 2 166 1 2 2 2 1775 2 2 2 2 1824 2 2 2 2 167 2 2 2 2 1683 2 2 2 2 1693 2 2 2 2 169 2 2 2 2 1864 2 2 2 2 1709 2 2 2 2 1764 2 2 2 2

1741 2 2 2 2 1797 2 2 2 2 1626 2 2 2 2 1724 2 2 2 2 177 2 2 2 2 1744 2 2 2 2 1716 2 2 2 2 1805 2 2 2 2 1827 2 2 2 2 1752 2 2 2 2 1746 2 2 2 2 1666 2 2 2 2 187 2 2 2 2 1718 2 2 2 2 1483 2 2 2 2 1632 2 2 2 2 1759 2 2 2 2 1723 2 2 2 2 1684 2 2 2 2 1887 2 2 2 2 1756 2 2 2 2 1816 2 2 2 2 1792 2 2 2 2 1614 2 2 2 2 1566 2 2 2 2 1741 2 2 2 2 1752 2 2 2 2 1697 2 2 2 2 1762 2 2 2 2 1801 2 2 2 2 1766 2 2 2 2 1747 2 2 2 2 1744 2 2 2 2 1692 2 2 2 2 186 2 2 2 2 1756 2 2 2 2 1896 2 2 2 2 1814 2 2 2 2 1703 2 2 2 2 1722 2 2 2 2 1688 2 2 2 2 1821 2 2 2 2 1758 2 2 2 2 1787 2 2 2 2

1784 2 2 2 2 179 2 2 2 2 1681 2 2 2 2 1685 2 2 2 2 1798 2 2 2 2 171 2 2 2 2 1638 2 2 2 2 1931 2 2 2 2 1786 2 2 2 2 1768 2 2 2 2 1764 2 2 2 2 1819 2 2 2 2 1666 1 2 3 2 1626 1 2 3 2 188 1 2 3 2 1653 1 2 3 2 1733 1 2 3 2 1551 1 2 3 2 1884 1 2 3 2 1718 1 2 3 2 1549 1 2 3 2 1723 1 2 3 2 194 1 2 3 2 164 1 2 3 2 1565 1 2 3 2 1787 1 2 3 2 1655 2 2 3 2 167 2 2 3 2 1673 2 2 3 2 1705 2 2 3 2 1755 2 2 3 2 1625 2 2 3 2 1708 2 2 3 2 1705 2 2 3 2 1862 2 2 3 2 1686 2 2 3 2 1723 2 2 3 2 1909 2 2 3 2 1859 2 2 3 2 1897 2 2 3 2 1972 2 2 3 2 END

Table C-3 Example OpenBUGS Certification Summary for Component B

mean sd MC_error val25pc median val975pc start sample b 09337 03387 0007504 02695 09371 1592 100001 100000 mu[1] 1764 03101 0007483 1704 1763 1825 100001 100000 mu[2] 1748 008014 4719E-4 1733 1748 1764 100001 100000 mu[3] 1735 02338 0003093 1688 1736 1781 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

62

Certification data for Component C list(Ntot=280) lamb[] asy[] cmp[] typ[] pip[] 163 1 3 1 1 1596 2 3 1 1 1574 3 3 1 1 159 4 3 1 1 161 5 3 1 1 1589 6 3 1 1 1569 7 3 1 1 1565 8 3 1 1 1564 9 3 1 1 1548 10 3 1 1 1706 1 3 1 1 1576 2 3 1 1 1538 3 3 1 1 1555 4 3 1 1 1607 5 3 1 1 1548 6 3 1 1 1572 7 3 1 1 1681 8 3 1 1 1492 9 3 1 1 1655 10 3 1 1 1499 1 3 1 1 1576 2 3 1 1 1545 3 3 1 1 1584 4 3 1 1 1577 5 3 1 1 1599 6 3 1 1 1587 7 3 1 1 152 8 3 1 1 1479 9 3 1 1 1512 10 3 1 1 1497 1 3 1 1 1611 2 3 1 1 156 3 3 1 1 1546 4 3 1 1 1529 5 3 1 1 1478 6 3 1 1 1504 7 3 1 1 1551 8 3 1 1 1438 9 3 1 1 1494 10 3 1 1 1452 1 3 1 1 1347 2 3 1 1 1395 3 3 1 1 1442 4 3 1 1 1448 5 3 1 1 1463 6 3 1 1 1402 7 3 1 1 1474 8 3 1 1 1466 9 3 1 1 1439 10 3 1 1 1424 1 3 1 1 1397 2 3 1 1 135 3 3 1 1 1459 4 3 1 1 1602 5 3 1 1 1435 6 3 1 1 1433 7 3 1 1 1481 8 3 1 1 1482 9 3 1 1

1568 10 3 1 1 147 1 3 1 1 1411 2 3 1 1 1512 3 3 1 1 1452 4 3 1 1 1567 5 3 1 1 148 6 3 1 1 1442 7 3 1 1 143 8 3 1 1 1413 9 3 1 1 1486 10 3 1 1 1444 1 3 1 1 1427 2 3 1 1 1439 3 3 1 1 1468 4 3 1 1 1524 5 3 1 1 1424 6 3 1 1 1459 7 3 1 1 1467 8 3 1 1 1411 9 3 1 1 148 10 3 1 1 1485 1 3 1 1 1422 2 3 1 1 1438 3 3 1 1 1439 4 3 1 1 1515 5 3 1 1 1409 6 3 1 1 1469 7 3 1 1 1453 8 3 1 1 1465 9 3 1 1 1363 10 3 1 1 1458 1 3 1 1 1459 2 3 1 1 1465 3 3 1 1 1468 4 3 1 1 1457 5 3 1 1 1412 6 3 1 1 1474 7 3 1 1 1428 8 3 1 1 1425 9 3 1 1 1393 10 3 1 1 1357 1 3 1 2 1447 2 3 1 2 1448 3 3 1 2 1409 4 3 1 2 1389 5 3 1 2 1415 6 3 1 2 1431 7 3 1 2 1381 8 3 1 2 1403 9 3 1 2 1435 10 3 1 2 1364 1 3 1 2 1361 2 3 1 2 1404 3 3 1 2 1411 4 3 1 2 1429 5 3 1 2 1417 6 3 1 2 138 7 3 1 2 1336 8 3 1 2 1387 9 3 1 2

1372 10 3 1 2 1493 1 3 2 2 1516 1 3 2 2 1451 1 3 2 2 144 1 3 2 2 1523 1 3 2 2 1437 1 3 2 2 1485 1 3 2 2 1433 1 3 2 2 1463 1 3 2 2 1534 1 3 2 2 1533 1 3 2 2 145 1 3 2 2 1537 1 3 2 2 1515 1 3 2 2 1511 1 3 2 2 1449 1 3 2 2 1388 1 3 2 2 1475 1 3 2 2 1468 1 3 2 2 1549 1 3 2 2 153 1 3 2 2 1442 1 3 2 2 1385 1 3 2 2 1513 1 3 2 2 1528 1 3 2 2 1486 1 3 2 2 1472 1 3 2 2 1493 1 3 2 2 1477 1 3 2 2 1405 1 3 2 2 1499 1 3 2 2 1458 1 3 2 2 1514 1 3 2 2 1525 1 3 2 2 1527 1 3 2 2 149 1 3 2 2 1568 1 3 2 2 1452 1 3 2 2 1421 1 3 2 2 1455 1 3 2 2 1525 1 3 2 2 1531 1 3 2 2 144 1 3 2 2 1487 1 3 2 2 151 1 3 2 2 1577 1 3 2 2 1506 1 3 2 2 152 1 3 2 2 1483 1 3 2 2 1522 1 3 2 2 1463 1 3 2 2 1478 1 3 2 2 1505 1 3 2 2 1454 1 3 2 2 1555 1 3 2 2 1497 1 3 2 2 1515 1 3 2 2 1494 1 3 2 2 1594 1 3 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

63

1471 1 3 2 2 1394 1 3 2 2 1445 1 3 2 2 1446 1 3 2 2 1499 1 3 2 2 1484 1 3 2 2 1436 1 3 2 2 1565 2 3 2 2 149 2 3 2 2 1485 2 3 2 2 1529 2 3 2 2 1513 2 3 2 2 1492 2 3 2 2 1475 2 3 2 2 1497 2 3 2 2 1503 2 3 2 2 1448 2 3 2 2 1488 2 3 2 2 1496 2 3 2 2 1499 2 3 2 2 1469 2 3 2 2 1515 2 3 2 2 1543 2 3 2 2 1452 2 3 2 2 1458 2 3 2 2 1435 2 3 2 2 1499 2 3 2 2 1499 2 3 2 2 1589 2 3 2 2 1562 2 3 2 2 1547 2 3 2 2 1505 2 3 2 2 1508 2 3 2 2 152 2 3 2 2

1481 2 3 2 2 1547 2 3 2 2 1501 2 3 2 2 1491 2 3 2 2 1465 2 3 2 2 147 2 3 2 2 1483 2 3 2 2 1479 2 3 2 2 148 2 3 2 2 1543 2 3 2 2 1448 2 3 2 2 1531 2 3 2 2 1478 2 3 2 2 1465 2 3 2 2 1454 2 3 2 2 1591 2 3 2 2 1466 2 3 2 2 157 2 3 2 2 1471 2 3 2 2 1506 2 3 2 2 1491 2 3 2 2 1461 2 3 2 2 1485 2 3 2 2 1495 2 3 2 2 1516 2 3 2 2 1474 2 3 2 2 1483 2 3 2 2 1519 2 3 2 2 1475 2 3 2 2 1466 2 3 2 2 1565 2 3 2 2 1491 2 3 2 2 1435 2 3 2 2 1459 2 3 2 2

1421 2 3 2 2 1468 2 3 2 2 1556 2 3 2 2 1424 1 3 3 2 1427 1 3 3 2 1381 1 3 3 2 1394 1 3 3 2 1473 1 3 3 2 133 1 3 3 2 1446 1 3 3 2 1424 1 3 3 2 133 1 3 3 2 1452 1 3 3 2 1512 1 3 3 2 1314 1 3 3 2 1525 1 3 3 2 1475 1 3 3 2 1542 1 3 3 2 1461 2 3 3 2 1443 2 3 3 2 1547 2 3 3 2 1394 2 3 3 2 1364 2 3 3 2 1395 2 3 3 2 1522 2 3 3 2 1483 2 3 3 2 1452 2 3 3 2 1482 2 3 3 2 1439 2 3 3 2 1458 2 3 3 2 155 2 3 3 2 1502 2 3 3 2 152 2 3 3 2 END

Table C-4 Example OpenBUGS Certification Summary for Component C

mean sd MC_error val25pc median val975pc start sample b 09626 01701 0002428 06285 09633 1293 100001 100000 mu[1] 1399 01566 0002413 1369 1399 143 100001 100000 mu[2] 1491 005701 2386E-4 148 1491 1503 100001 100000 mu[3] 1449 0132 0001181 1423 1449 1475 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

64

34 Results Table C-5 lists the results in CopiesnL for the three components

Table C-5 Parameter Values for Components A B and C

Posterior Distribution Summary Statistics Component Parameter Mean SD 25 50 975

A

β 260 029 203 260 318 μ1 1415 027 1362 1415 1467 μ2 1559 010 1539 1559 158 μ3 1533 016 1501 1533 1565

B

β 095 034 029 095 162 μ1 1762 031 1701 1762 1822 μ2 1748 010 1728 1748 1769 μ3 1735 024 1687 1735 1783

C

β 097 017 064 097 130 μ1 1399 016 1368 1399 1429 μ2 1491 008 1476 1491 1507 μ3 1449 014 1421 1449 1477

Since in each case the three data set types represent independent measurements the three estimates can be combined to obtain a consensus value In this way a between data set type variability is included in the final standard uncertainty Table C-6 lists the results in Copies per Nanoliter obtained using the NIST Consensus Builder (NICOB)C3

Table C-6 Certified Values for Components A B and C

Component Mean u(Mean) Uk=2(Mean) A 151 07 14 B 175 03 06 C 145 05 10

4 Mito-to-nuclear DNA ratios (mtDNAnDNA)

41 Data Measurements were also obtained on the same samples using three mitochondrial detection assays mtND1 mtBatz and mtKav and a single nDNA assay (genomic) to compute the ratios mtDNAnDNA For the complete mtDNAnDNA data see

Table 16 mtDNAnDNA Ratios

42 Model For each component the data was analyzed using the following random effects model to account for between assay variability

119903119903119894119894~119873119873120572120572119894119894 1205901205901198861198861198861198861198861198861198861198861198861198862

where 119894119894 denotes assay (for A and B 1 = mtND1 2 = mtBatz 3 = mtKav for C 1 = mtBatz 2 = mtKav) and 120572120572119894119894~119873119873micro120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119886119886119886119886119886119886119886119886119886119886

2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

65

43 OpenBUGS code inits and data Two versions of the same model are required one for the results from the three mtDNA assays available for components A and B mtND1 mtBatz and mtKav and the other for the results from the two mtDNA assays available for component C mtBatz and mtKav mtDNAnDNA code for components A and B ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 13)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for components A and B list(sigt=c(111)) Data for Component A list(Ndata=102) set[] rep[] asy[] rat[] 1 1 1 1931 1 2 1 187 1 1 2 1845 1 2 2 1764 2 1 1 181 2 2 1 1906 2 1 2 1785 2 2 2 1837 3 1 1 1743 3 2 1 1761 3 1 2 1829 3 2 2 1847 4 1 1 1716 4 2 1 180 4 1 2 174 4 2 2 1813 5 1 1 1923 5 2 1 1835 5 1 2 1935 5 2 2 1851 6 1 1 1757 6 2 1 1724 6 3 1 1768 6 1 2 1743 6 2 2 1671 6 3 2 1766 6 1 3 1692 6 2 3 1662 6 3 3 1767 7 1 1 1924 7 2 1 1944 7 3 1 1893 7 1 2 1786

7 2 2 1752 7 3 2 1743 7 1 3 153 7 2 3 1542 7 3 3 1523 7 1 2 158 7 2 2 1557 7 3 2 1592 7 1 3 2103 7 2 3 210 7 3 3 2049 7 1 1 1749 7 2 1 1769 7 3 1 1777 7 1 2 1965 7 2 2 1889 7 3 2 1947 7 1 3 1813 7 2 3 1864 7 3 3 1864 8 1 1 178 8 2 1 1776 8 3 1 1706 8 1 2 1804 8 2 2 1751 8 3 2 1738 8 1 3 1769 8 2 3 1786 8 3 3 1761 9 1 2 1584 9 2 2 1544 10 1 1 1947 10 2 1 1928 10 3 1 1922 10 1 1 1829

10 2 1 1895 10 3 1 1851 10 1 1 1674 10 2 1 1617 11 1 2 1667 11 2 2 1813 11 3 2 1708 11 4 2 1598 11 1 2 1566 11 2 2 1687 11 3 2 1663 11 4 2 1589 11 1 2 1643 11 2 2 1668 11 3 2 1604 11 4 2 1573 12 1 2 1719 12 2 2 1716 12 3 2 167 12 4 2 1666 12 1 2 1581 12 2 2 1573 12 3 2 1601 12 4 2 1612 12 1 2 1546 12 2 2 1542 12 3 2 1538 12 4 2 1587 13 1 1 1617 13 2 1 1599 13 1 1 1622 13 2 1 1579 13 1 1 1575 13 2 1 1528 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

66

Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A

mean sd MC_error val25pc median val975pc start sample mu 1730 1443 004419 1703 1730 1759 100001 100000

data for Component B list(Ndata=87) set[] rep[] asy[] rat[] 1 1 1 2461 1 2 1 2401 1 1 2 2555 1 2 2 2608 2 1 1 2229 2 2 1 2200 2 1 2 2108 2 2 2 2192 3 1 1 2127 3 2 1 2198 3 1 2 2455 3 2 2 2506 4 1 1 2211 4 2 1 2199 4 1 2 2380 4 2 2 2249 5 1 1 2105 5 2 1 1943 5 1 2 2136 5 2 2 2119 6 1 1 1990 6 2 1 1921 6 3 1 1982 6 1 2 1990 6 2 2 1924 6 3 2 1958 6 1 3 1903 6 2 3 1831

6 3 3 1919 7 1 1 2304 7 2 1 2341 7 3 1 2332 7 1 2 2048 7 2 2 2078 7 3 2 2067 7 1 3 2399 7 2 3 2479 7 3 3 2493 7 1 1 2065 7 2 1 2166 7 3 1 2164 7 1 2 2314 7 2 2 2280 7 3 2 2258 7 1 3 1867 7 2 3 1987 7 3 3 1971 8 1 1 2046 8 2 1 2057 8 3 1 2024 8 1 2 1998 8 2 2 2043 8 3 2 1952 8 1 3 2077 8 2 3 2117 8 3 3 2041 9 1 2 1890 9 2 2 1886

10 1 2 1928 10 2 2 1916 10 3 2 1915 10 1 2 1947 10 2 2 1946 10 3 2 1961 10 4 2 1909 10 1 2 1953 10 2 2 1952 10 3 2 1941 10 4 2 1881 11 1 2 1865 11 2 2 1863 11 3 2 1828 11 4 2 1828 11 1 2 1944 11 2 2 1964 11 3 2 1947 11 4 2 1895 11 1 2 1908 11 2 2 1909 11 3 2 1940 11 4 2 1839 12 1 1 1688 12 2 1 1783 12 1 1 1823 12 2 1 1631 12 1 1 1920 12 2 1 1887 END

Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B

mean sd MC_error val25pc median val975pc start sample mu 2045 2246 0111 1999 2045 2088 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

67

mtDNAnDNA code for component C ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 12)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for component C list(sigt=c(11)) mtDNAnDNA data for component C list(Ndata=66) set[] rep[] asy[] rat[] 1 1 1 3201 1 2 1 3299 2 1 1 3347 2 2 1 3287 3 1 1 3217 3 2 1 3384 4 1 1 3200 4 2 1 3363 5 1 1 3294 5 2 1 3013 6 1 1 2805 6 2 1 2825 6 3 1 2829 6 1 2 2709 6 2 2 2738 6 3 2 2727 7 1 1 2797 7 2 1 2642 7 3 1 2792 7 1 2 2804 7 2 2 2789

7 3 2 2842 7 1 1 2715 7 2 1 2606 7 3 1 2687 7 1 2 2799 7 2 2 2628 7 3 2 2559 7 1 1 2639 7 2 1 2535 7 3 1 2657 7 1 2 2953 7 2 2 2757 7 3 2 2802 8 1 1 2927 8 2 1 2776 8 3 1 2762 8 1 2 3044 8 2 2 2830 8 3 2 2825 9 1 1 2501 9 2 1 2522 10 1 1 2832 10 2 1 2840

10 3 1 2788 10 4 1 2725 10 1 1 2835 10 2 1 2790 10 3 1 2736 10 4 1 2679 10 1 1 2803 10 2 1 2773 10 3 1 2757 10 4 1 2752 11 1 1 2599 11 2 1 2508 11 3 1 2551 11 4 1 2609 11 1 1 2617 11 2 1 2483 11 3 1 2626 11 4 1 2602 11 1 1 2665 11 2 1 2535 11 3 1 2570 11 4 1 2650 END

Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C

mean sd MC_error val25pc median val975pc start sample mu 2774 2367 01176 2729 2774 2821 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

68

Table C-10 Results for mtDNAnDNA Posterior Distribution Summary Statistics

Component Mean Std Dev 25 50 975 A 1730 14 1703 1730 1758 B 2045 23 1994 2047 2087 C 2774 23 2729 2774 2821

5 ReferencesC1 Lunn DJ Spiegelhalter D Thomas A Best N Statistics in Medicine 2009283049ndash3082C2 Polson NG Scott JG Bayesian Analysis 20127887ndash902C3 NIST Consensus Builder Software and userrsquos guide freely available at

httpsconsensusnistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

69

Appendix D Performance in Commercial qPCR Chemistries The performance of the SRM 2372a materials in commercial qPCR chemistries was assessed using manufacturerrsquos recommended protocols Each panel in Figures D-1 to D-4 displays the average cycle threshold (Ct) as a function of DNA concentration for the 5-point dilutions of SRM 2372a components In all cases the external calibrant was SRM 2372 component A Error bars represent approximate 95 confidence intervals Table D-1 lists the estimated [DNA] for the three SRM 2372a components for all commercial assays evaluated This Table also lists the Component BComponent A (BA) and Component CComponent A (CA) ratios calculated from the [DNA] estimates

Figure D-1 Results for the Innogenomics Commercial qPCR assays

20

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantLong(IQ_h)

ABCStd

[A] = (384 plusmn 26) ngμL[B] = (441 plusmn 24) ngμL[C] = (516 plusmn 24) ngμL

Rsup2 = 0999 se(Ct) = 01016

17

18

19

20

21

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantShort(IQ_d)

ABCStd

[A] = (435 plusmn 28) ngμL[B] = (543 plusmn 12) ngμL[C] = (510 plusmn 28) ngμL

Rsup2 = 0999 se(Ct) = 007

24

25

26

27

28

29

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYLong(IQHY_h)

ABCStd

[A] = (387 plusmn 24) ngμL[B] = (448 plusmn 21) ngμL[C] = (559 plusmn 31) ngμL

Rsup2 = 0998 se(Ct) = 01314

16

18

20

22

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYShort(IQHY_d)

ABCStd

[A] = (353 plusmn 23) ngμL[B] = (267 plusmn 15) ngμL[C] = (390 plusmn 18) ngμL

Rsup2 = 1000 se(Ct) = 00420

22

24

26

28

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYY(IQHY_y)

ABCStd

[A] = (572 plusmn 13) ngμL[B] = 0 ngμL[C] = (137 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

70

Figure D-2 Results for the Thermo Fisher Commercial qPCR assays

21

22

23

24

25

26

27

28

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HumanHuman(QFH_h)

ABCStd

[A] = (464 plusmn 13) ngμL[B] = (529 plusmn 19) ngμL[C] = (472 plusmn 22) ngμL

Rsup2 = 1000 se(Ct) = 004

22

23

24

25

26

27

28

29

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoHuman(QFD_h)

ABCStd

[A] = (701 plusmn 53) ngμL[B] = (561 plusmn 11) ngμL[C] = (472 plusmn 11) ngμL

Rsup2 = 0994 se(Ct) = 01922

24

26

28

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoMale(QFD_y)

ABCStd

[A] = (706 plusmn 49) ngμL[B] = 0 ngμL[C] = (122 plusmn 02) ngμL

Rsup2 = 0999 se(Ct) = 007

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPLarge(QFHP_h)

ABCStd

[A] = (340 plusmn 18) ngμL[B] = (432 plusmn 05) ngμL[C] = (342 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 00720

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPSmall(QFHP_d)

ABCStd

[A] = (434 plusmn 15) ngμL[B] = (514 plusmn 10) ngμL[C] = (466 plusmn 07) ngμL

Rsup2 = 1000 se(Ct) = 005

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioLarge(QFT_h)

ABCStd

[A] = (362 plusmn 23) ngμL[B] = (430 plusmn 06) ngμL[C] = (378 plusmn 07) ngμL

Rsup2 = 0999 se(Ct) = 00920

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioSmall(QFT_d)

ABCStd

[A] = (468 plusmn 08) ngμL[B] = (505 plusmn 08) ngμL[C] = (523 plusmn 14) ngμL

Rsup2 = 1000 se(Ct) = 00318

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioY(QFT_y)

ABCStd

[A] = (372 plusmn 11) ngμL[B] = 0 ngμL[C] = (110 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 003

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

71

Figure D-3 Results for the Qiagen Commercial qPCR assays

Figure D-4 Results for the Promega Commercial qPCR assays

19

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresAutosomal(QPH_h)

ABCStd

[A] = (554 plusmn 21) ngμL[B] = (479 plusmn 05) ngμL[C] = (381 plusmn 09) ngμL

Rsup2 = 1000 se(Ct) = 00418

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresY(QPH_y)

ABCStd

[A] = (358 plusmn 10) ngμL[B] = 0 ngμL[C] = (118 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 007

16

17

18

19

20

21

22

23

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProHuman(QPP_h)

ABCStd

[A] = (525 plusmn 49) ngμL[B] = (480 plusmn 42) ngμL[C] = (467 plusmn 45) ngμL

Rsup2 = 0999 se(Ct) = 01019

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProDegradation(QPP_d)

ABCStd

[A] = (567 plusmn 18) ngμL[B] = (558 plusmn 08) ngμL[C] = (484 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 00410

15

20

25

30

35

-50 -40 -30 -20 -10 00 10

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProY(QPP_y)

ABCStd

[A] = (538 plusmn 27) ngμL[B] = (00 plusmn 00) ngμL[C] = (162 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 003

16

18

20

22

24

-05 00 05 10 15 20

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorAutosomal(P_h)

ABCStd

[A] = (759 plusmn 64) ngμL[B] = (775 plusmn 24) ngμL[C] = (1056 plusmn 17) ngμL

Rsup2 = 0995 se(Ct) = 01818

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorY(P_y)

ABCStd

[A] = (596 plusmn 51) ngμL[B] = 0 ngμL[C] = (117 plusmn 01) ngμL

Rsup2 = 0994 se(Ct) = 020

18

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantAutosomal(PQ_h)

ABCStd

[A] = (436 plusmn 13) ngμL[B] = (606 plusmn 20) ngμL[C] = (504 plusmn 10) ngμL

Rsup2 = 1000 se(Ct) = 00218

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantDegradation(PQ_d)

ABCStd

[A] = (420 plusmn 18) ngμL[B] = (529 plusmn 16) ngμL[C] = (530 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 00516

18

20

22

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantY(PQ_y)

ABCStd

[A] = (410 plusmn 20) ngμL[B] = 0 ngμL[C] = (139 plusmn 05) ngμL

Rsup2 = 0999 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

72

Table D-1 Summary [DNA] Results for the Commercial qPCR Assays Estimated [DNA] ngμL Ratios

A B C BA CA Assaya xb u(x)c xb u(x)c xb u(x)c yd u(y)e yd u(y)e

IQd 435 28 543 12 510 28 125 008 117 010 IQh 384 26 441 24 516 24 115 010 134 011

IQHYd 353 23 267 15 390 18 076 006 111 009 IQHYh 387 24 448 21 559 31 116 009 144 012

Ph 759 64 775 24 1056 16 102 009 139 012 PQd 420 18 529 16 530 06 126 007 126 006 PQh 436 13 606 20 504 10 139 006 116 004

QFDh 701 53 561 11 472 11 080 006 067 005 QFHh 464 13 529 19 472 22 114 005 102 006

QFHPd 434 15 514 10 466 07 118 005 107 004 QFHPh 340 18 432 05 342 03 127 007 101 005

QFTd 468 08 505 08 523 14 108 003 112 004 QFTh 362 23 430 06 378 07 118 008 104 007 QPHh 554 21 479 05 380 09 086 003 069 003 QPPd 567 18 558 08 484 11 098 003 085 003 QPPh 525 49 480 42 467 45 092 012 089 012

IQHYy 572 13 NDf 137 06

NRg

024 001 Py 596 51 NDf 117 01 020 002

PQy 410 20 NDf 139 05 034 002 QFDy 706 49 NDf 122 02 017 001 QFTy 372 11 NDf 110 06 030 002 QPHy 358 10 NDf 118 03 033 001 QPPy 538 27 0002 0001 162 11 030 003

a Assay codes are listed in Table 8 b Mean of dilution-adjusted results from five dilutions of the SRM 2372a component c Standard uncertainty of the mean estimated from the stdanrd deviation of the five dilutions d Ratio of the estimated [DNA] of component B or C relative to that of component A e Standard uncertainty of the ratio estimated from the standard uncertainties of the individual components f Not detected Component B was prepared using tissue from only female donors it is not expected to

respond to male-specific assays g Not relevant

  • Abstract
  • Keywords
  • Technical Information Contact for this SRM
  • Table of Contents
  • Table of Tables
  • Table of Figures
  • Purpose and Description
  • Warning SRM 2372a is a Human Source Material
  • Storage and Use
  • History and Background
    • Figure 1 Sales History of SRM 2372
      • Experimental Methods
        • Polymerase Chain Reaction Assays
          • Table 1 NIST-Developed Human Nuclear DNA Assays
          • Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays
          • Table 2 NIST-Optimized Human Mitochondrial DNA Assays
          • Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays
            • Droplet Digital Polymerase Chain Reaction Measurements
              • Nuclear DNA (nDNA)
                • Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays
                  • Mitochondrial DNA (mtDNA)
                    • Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays
                    • Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays
                        • Sample Preparation
                        • Absorbance Measurements
                        • SRM Unit Production
                        • Component Homogeneity
                          • Figure 4 Plate Layout Design for Homogeneity Measurements
                            • Component Stability
                              • Figure 5 Plate Layout Design for Stability Measurements
                                • Component Certification Measurements
                                  • Table 6 Tubes used for Certification Measurements
                                  • Figure 6 Basic Plate Layout Design for Certification Measurements
                                    • Mitochondrial DNA Measurements
                                      • Table 7 Run Parameters for Mitochondrial DNA
                                        • Performance in Commercial qPCR Chemistries
                                          • Table 8 Commercial qPCR Chemistries Tested with SRM 2372a
                                          • Figure 7 Plate Layout Design for Performance Assessments
                                              • Measurement Results
                                                • Absorbance Spectrophotometry
                                                  • Table 9 Absorbance at Selected Wavelengths
                                                  • Figure 8 Absorbance Spectra of SRM 2372a Components A B and C
                                                    • Droplet Volume
                                                    • Certification Measurements
                                                      • Homogeneity Results
                                                        • Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component
                                                        • Table 11 Within- and Between-Tube Relative Standard Deviations
                                                          • Stability Results
                                                            • Figure 9 ddPCR Stability Measurements as Function of Time
                                                            • Table 12 Stability Data as λacute Copy Number per Nanoliter of Component
                                                              • Ten-Assay Certification Results
                                                                • Figure 10 Comparison of Assay Performance
                                                                • Table 13 Certification Data as λacute Copy Number per Nanoliter of Component
                                                                  • Estimation of Copy Number Concentration
                                                                    • Table 14 Measured Copy Number Concentrations
                                                                      • Conversion of Copy Number to Mass Concentration
                                                                        • Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution
                                                                            • Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)
                                                                              • Figure 11 mtDNAnDNA for Components A B and C
                                                                              • Table 16 mtDNAnDNA Ratios
                                                                              • Table 17 mtDNAnDNA Summary Results
                                                                                • Performance in Commercial qPCR Chemistries
                                                                                  • Figure 12 qPCR Autosomal Quantitation Performance Summary
                                                                                  • Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary
                                                                                      • Qualification of ddPCR as Fit-For-Purpose
                                                                                        • Figure 14 Absorbance Spectra of SRM 2372 Components A and B
                                                                                        • Digital Polymerase Chain Reaction Assays
                                                                                        • Heat Treatments (Melting of the DNA into Single Strands)
                                                                                        • Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a
                                                                                          • Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a
                                                                                            • Quantitative Evaluation of ssDNA Content
                                                                                              • ddPCR Measurement of ssDNA Proportion
                                                                                              • Proof of Principle
                                                                                                • Table 19 Comparison of Snap-Cooling to Slow-Cooling
                                                                                                  • Time at Denaturation Temperature
                                                                                                    • Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures
                                                                                                      • Temperature
                                                                                                        • Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a
                                                                                                          • DNA Concentration in the ddPCR Solution
                                                                                                            • Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution
                                                                                                              • Equivalence of Assays
                                                                                                                • Table 23 Comparison of Assays
                                                                                                                    • ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials
                                                                                                                      • Table 24 Evaluation of Limiting Ratio
                                                                                                                          • Conclusions and Recommendations
                                                                                                                            • Recommended Certification Values
                                                                                                                              • Table 25 Recommended Values and 95 Uncertainties for Certification
                                                                                                                                • Use of SRM 2372a for Value-Assigning In-House Reference Materials
                                                                                                                                  • Certificate of Analysis
                                                                                                                                  • References
                                                                                                                                  • Appendix A DNA Extraction Method
                                                                                                                                    • Procedure
                                                                                                                                    • Reagents
                                                                                                                                    • Reference
                                                                                                                                      • Appendix B Spectrophotometric Measurement Protocol
                                                                                                                                        • Spectrophotometer Calibration
                                                                                                                                          • Establish Baseline
                                                                                                                                          • Enable Verification of Wavelength Calibration
                                                                                                                                          • Enable Verification of Absorbance Calibration
                                                                                                                                            • For ldquonativerdquo dsDNA Samples
                                                                                                                                              • Verify Optical Balance of Cuvettes
                                                                                                                                                • For NaOH denatured ssDNA Samples
                                                                                                                                                  • Prepare a 04 molL NaOH Stock Solution
                                                                                                                                                  • Prepare the Sample Solutions
                                                                                                                                                  • Measure Sample Solutions
                                                                                                                                                    • After All Sample Measurements are Complete
                                                                                                                                                      • Enable Verification of Spectrophotometer Stability
                                                                                                                                                      • Clean Up
                                                                                                                                                        • Evaluation
                                                                                                                                                        • Reference
                                                                                                                                                          • Appendix C SRM 2372a Statisticianrsquos report
                                                                                                                                                            • 1 Introduction
                                                                                                                                                            • 11 OpenBUGS Analysis
                                                                                                                                                            • 12 OpenBUGS Exemplar Output
                                                                                                                                                            • 2 Homogeneity assessment
                                                                                                                                                            • 21 Data
                                                                                                                                                            • 22 Model
                                                                                                                                                            • 23 OpenBUGS code and data
                                                                                                                                                              • Table C-1 Example OpenBUGS Homogeneity Summary
                                                                                                                                                                • 24 Results
                                                                                                                                                                  • Figure C-1 Homogeneity Assessment
                                                                                                                                                                    • 3 Certification measurements
                                                                                                                                                                    • 31 Data
                                                                                                                                                                    • 32 Model
                                                                                                                                                                    • 33 OpenBUGS code inits and data
                                                                                                                                                                      • Table C-2 Example OpenBUGS Certification Summary for Component A
                                                                                                                                                                      • Table C-3 Example OpenBUGS Certification Summary for Component B
                                                                                                                                                                      • Table C-4 Example OpenBUGS Certification Summary for Component C
                                                                                                                                                                        • 34 Results
                                                                                                                                                                          • Table C-5 Parameter Values for Components A B and C
                                                                                                                                                                          • Table C-6 Certified Values for Components A B and C
                                                                                                                                                                            • 4 Mito-to-nuclear DNA ratios (mtDNAnDNA)
                                                                                                                                                                            • 41 Data
                                                                                                                                                                            • 42 Model
                                                                                                                                                                            • 43 OpenBUGS code inits and data
                                                                                                                                                                              • Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A
                                                                                                                                                                              • Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B
                                                                                                                                                                              • Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C
                                                                                                                                                                              • Table C-10 Results for mtDNAnDNA
                                                                                                                                                                                • 5 References
                                                                                                                                                                                  • Appendix D Performance in Commercial qPCR Chemistries
                                                                                                                                                                                    • Table D-1 Summary [DNA] Results for the Commercial qPCR Assays
Page 9: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

1

Purpose and Description Standard Reference Material (SRM) 2372a is intended for use in the value assignment of human genomic deoxyribonucleic acid (DNA) forensic quantitation materials A unit of SRM 2372a consists of three well-characterized human genomic DNA materials solubilized in 10 mmolL 2-amino-2-(hydroxymethyl)-13 propanediol hydrochloride (Tris HCl) and 01 mmolL ethylenediaminetetraacetic acid disodium salt (disodium EDTA) pH 80 buffer (TE-4) The three component genomic DNA materials labeled A B and C are respectively derived from a single male donor a single female donor and 13 mixture of a male and a female donor A unit of the SRM consists of one 05 mL tube of each component each tube containing approximately 55 microL of DNA solution Each of these tubes is labeled and is sealed with a color-coded screw cap NIST is guided by and adheres to the ethical principles set forth in the Belmont Report1 SRM 2372a was developed after an appropriate human subject research determination by the NIST Human Subjects Protections Office

Warning SRM 2372a is a Human Source Material The three SRM 2372a components are human genomic DNA extracts prepared at NIST from buffy coat white blood cells from single-source anonymous donors All buffy coats were purchased from Interstate Blood Bank (Memphis TN) The supplier reported that each donor unit used in the preparation of this product was tested by FDA-licensed tests and found to be negative for human immunodeficiency virus (HIV-1 RNA and Anti-HIV 12) hepatitis B (HBV DNA and HBsAg) hepatitis C (HCV RNA and Anti-HCV) West Nile virus (WNV RNA) and syphilis Additionally each donor was tested according to FDA guidelines for Trypanosoma cruzi (Chagas) and found to be negativenon-reactive However no known test method can offer complete assurance that infectious agents are absent from this material Accordingly this human blood-based product should be handled at the Biosafety Level 12 SRM 2372a components and derived solutions should be disposed of in accordance with local state and federal regulations

Storage and Use Until required for use SRM 2372a should be stored in the dark between 2 degC and 8 degC The SRM 2372a component tubes should be mixed briefly and centrifuged (without opening the tube cap) prior to removing sample aliquots for analysis For the certified values to be applicable materials should be withdrawn immediately after opening the tubes and processed without delay Certified values do not apply to any material remaining in recapped tubes The certification only applies to the initial use and the same results are not guaranteed if the remaining material is used later

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

2

History and Background SRM 2372a is the second member of the SRM 2372 Human DNA Quantitation Standard series The supply of the initial material SRM 2372 was exhausted in May 2017 SRM 2372a is derived from commercially obtained buffy coat cells from different donors than those used in SRM 2372 A series of interlaboratory studies coordinated by what was then the Biotechnology Division of the National Institute of Standards and Technology (NIST) during the 1990s demonstrated that the DNA quantitation technologies then in use by the forensic human identification community adequately met the needs of the DNA typing tools then available34 However our 2004 study demonstrated that the increasingly sensitive and informative polymerase chain reaction (PCR)-based multiplex typing tools required more reliable determinations of DNA input quantity for best performance5 Quantitative PCR (qPCR) methods for DNA quantitation require calibration to an external standard of defined DNA concentration ([DNA]) expressed in units of nanograms of DNA per microliter of extract solution The true [DNA] of a commercial secondary standard may vary from its nominal concentration depending on supplier and lot number To enable providers of secondary standards to more reliably evaluate their materials and forensic testing laboratories to validate their use of commercial quantitation methods and supplies in 2007 we delivered the human DNA calibration standard SRM 2372 Human DNA Quantitation Standard6 The quantity of DNA in each of the three components of SRM 2372 was originally assigned based on the absorbance of double-stranded DNA (dsDNA) at 260 nm7 However by 2012 the absorbance properties of the materials had increased beyond the certified uncertainty limits due to slow but progressive changes in DNA tertiary structure Sales of the SRM were restricted until the DNA quantities were reassigned based on measurements of single-stranded DNA (ssDNA) in early 20138-10 Figure 1 displays the sales history of this relatively high-selling SRM While SRM 2372 was fit for the practical purpose of harmonizing DNA quantitation measurements to an accessible and reproducible reference material the conversion of both dsDNA and ssDNA absorbance to mass [DNA] is through conventional proportionality constants that are not metrologically traceable to the modern reference system for measurements the International System of Units (SI) Responding to requests from the forensic and clinical communities for calibration standards that provide traceability to the SI we and others have established droplet digital polymerase chain reaction (ddPCR) as a potentially primary measurement process that can provide traceable results11-18 The copy numbers of human nuclear DNA (nDNA) in the reaction mixture per partition symbolized as λ were determined by ddPCR using ten PCR assays optimized for use with our ddPCR platform These assays target loci on eight chromosomes with three assays on widely separated locations on one chromosome (see Table 1)

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

3

Figure 1 Sales History of SRM 2372

Each small black dot represents the sale of one unit of SRM 2372 The red lines summarize the average rate of sales as originally certified (2007 to 2012) and after re-certification (2013 to 2017) The magenta labels state the average yearly sales rate during these two periods

The nDNA copy number concentration human haploid genome equivalents (HHGE) per nanoliter component solution and symbolized here as λacute is calculated as

copy number [DNA] = λ = λ(VF) HHGEnL

where V is the average droplet volume and F is the total dilution factor the volume fraction of the component solution in the ddPCR reaction mixture The manufacturer of the ddPCR platform used a Bio-Rad QX200 Droplet Digital PCR System (Hercules CA) system does not provide metrologically traceable droplet volume information This lack of information prevented metrologically traceable conversion of the number per partition measurements to the desired number per reaction volume units In consequence NIST developed a measurement method that provides metrologically traceable droplet volumes19 We show elsewhere that the mass [DNA] nanograms of human DNA per microliter component can be traceably estimated as

mass [DNA] = 3031timesλacute ngmicroL

where the constant is the product of the number of nucleotide basepairs (bp) in the human reference genome the average bp molecular weight Avogadrorsquos number of bp per mole and

2372

2372(Recertif ied)

050

010

00

1713 unitsyear

1416 unitsyear

2007 2009 2011 2013 2015 2017

Uni

ts S

old

Calendar Year

Human DNA Quantitation StandardSRM 2372

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

4

unit scaling factors18 The three components of SRM 2372a are certified for both copy number and mass [DNA] Mitochondrial DNA (mtDNA) analysis within the forensic community is useful when the nDNA may be degraded or is in low-to-undetectable quantity The challenges of mtDNA testing include contamination of the standard due to routine laboratory use degradation of plasmids or synthetic oligomers or improper commutability of the standard relative to casework samples Commonly cell line materials are used as standards for qPCR with human genomic DNA which may lead to an incorrect mtDNA-to-nDNA ratio (mtDNAnDNA) for unknown samples Non-certified values of mtDNAnDNA for the three components are provided to bridge the current gap between well characterized mitochondrial DNA quantification assays and the availability of commercial standards With the understanding that the forensic DNA community primarily uses qPCR as a method for DNA quantification the performance of the three SRM 2372a materials was evaluated in eleven commercial qPCR chemistries currently used employed within the forensic DNA community

Experimental Methods Production and value-assignment of any certified reference material (CRM) is a complex process In addition to acquiring processing and packaging the materials included in the CRM the materials must be sufficiently homogenous across the units stable over time within its packaging and well-characterized to deliver results that are fit-for-purpose within the intended measurement community20 The process has been particularly complicated for SRM 2372a because we needed to determine the metrological properties of the required ddPCR measurement systems before we could trust the results they provide The following sections describe the methods and processes used to produce and characterize the SRM 2372a materials including sample preparation SRM unit production PCR assays ddPCR measurement processes spectrophotometric absorbance measurements homogeneity testing stability testing ddPCR nDNA and mtDNAnDNA measurements and performance in commercial qPCR assessments Many of these processes were mutually dependent therefore the sections are presented in a ldquodefinitionalrdquo order that sequentially introduces terminology Polymerase Chain Reaction Assays

Ten PCR assays were developed for characterization of SRM 2372a Each assay was confirmed to target one locus per HHGE using the National Center for Biotechnology Informationrsquos BLASTblastn system21 Primers were purchased from Eurofins Operon (Huntsville AL) TaqMan probes labeled with 6-carboxyfluorescein (FAM) were purchased from Thermo Fisher (Waltham MA) FAM-labeled Blackhole Quencher and Blackhole Quencher+ probes were purchased from LGC Biosearch Technologies (Novato CA) Table 1 details the ten nDNA assays used to certify copy number and mass [DNA] of the SRM 2372a components FAM-labeled TaqMan probes were purchased from Thermo Fisher

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

5

and used in monoplex reactions VIC (a proprietary dye)-labeled probes were purchased from the same source and were used when nDNA and mtDNA were analyzed in the same reaction mixture Figure 2 displays exemplar droplet patterns for the assays All assays were developed at NIST and have been optimized for ddPCR18 Table 2 details three mitochondrial DNA (mtDNA) assays used to estimate mtDNAnDNA in the SRM 2372a components Figure 3 displays exemplar droplet patterns for the assays These assays were optimized for ddPCR from real-time quantitative PCR (qPCR) assays from the literature All primers probes were diluted with TE-4 buffer purchased from Affymetrix-USB Products (Affymetrix Inc Cleveland OH) to a nominal 5 micromol working solution for all single-plex assays (individual nDNA and mtDNA assays) For duplex assays (combined nDNA and mtDNA) primer and probe working solutions were diluted to a nominal 10 micromol

Table 1 NIST-Developed Human Nuclear DNA Assays

Assay Target

Chromosome Band Accession Primers and Probe a

Amplicon Length bp

NEIF Gene EIF5B

Chr 2 p111-q111 NC_00000212

F gccaaacttcagccttctcttc R ctctggcaacatttcacactaca PB+ tcatgcagttgtcagaagctg

67

2PR4 Gene RPS27A

Chr 2 p16 NC_00000212

F cgggtttgggttcaggtctt R tgctacaatgaaaacattcagaagtct PB tttgtctaccacttgcaaagctggccttt

97

POTP STR TPOX

Chr 2 p253 NC_00000212

F ccaccttcctctgcttcacttt R acatgggtttttgcctttgg PT caccaactgaaatatg

60

NR4Q Gene DCK

Chr 4 q133-q211 NC_00000412

F tggtgggaatgttcttcagatga R tcgactgagacaggcatatgtt PB+ tgtatgagaaacctgaacgatggt

83

D5 STR D5S2500

Chr 5 q112 NC_00000510

F ttcatacaggcaagcaatgcat R cttaaagggtaaatgtttgcagtaatagat PT ataatatcagggtaaacaggg

75

ND6 STR D6S474

Chr 6 q21-22 NC_00000612

F gcatggctgagtctaaagttcaaag R gcagcctcagggttctcaa PB+ cccagaaccaaggaagatggt

82

D9 STR D9S2157

Chr 9 q342 NC_00000912

F ggctttgctgggtactgctt R ggaccacagcacatcagtcact PT cagggcacatgaat

60

HBB1 Gene HBB

Chr 11 p155 NC_00001110

F gctgagggtttgaagtccaactc R ggtctaagtgatgacagccgtacct PT agccagtgccagaagagccaagga

76

ND14 STR D14S1434

Chr 14 q3213C NC_0000149

F tccaccactgggttctatagttc R ggctgggaagtcccacaatc PB+ tcagactgaatcacaccatcag

109

22C3 Gene PMM1

Chr 22 q132 NC_00002210

F cccctaagaggtctgttgtgttg R aggtctggtggcttctccaat PB caaatcacctgaggtcaaggccagaaca

78

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

6

a F Forward primer R Reverse primer PB Blackhole quencher probe (FAM labeled) PB+ Blackhole Plus quencher probe (FAM labeled) PT TaqMan MGB probe (FAM labeled)

Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays

Table 2 NIST-Optimized Human Mitochondrial DNA Assays

Assay Location Primers and Probe a Amplicon Length bp Ref

mtND1 34853504 F ccctaaaacccgccacatct

69 22 35333553 R gagcgatggtgagagctaaggt 35063522 PT ccatcaccctctacatc

mtBatz 84468473 F aatattaaacacaaactaccacctacct

79 23 85038524 R tggttctcagggtttgttataa 84758491 PT cctcaccaaagcccata

mtKav 1318813308 F ggcatcaaccaaccacaccta

105 24 1336913392 R attgttaaggttgtggatgatgga 1331013325 PT cattcctgcacatctg

a F Forward primer R Reverse primer PT TaqMan MGB probe (FAM labeled)

2PR4

POTP

NEI

F

R4Q

5

D5

ND6

D9

HBB1

ND1

4

22C3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

7

Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays

Droplet Digital Polymerase Chain Reaction Measurements

Nuclear DNA (nDNA) All nDNA ddPCR measurements used in this study followed the same measurement protocol

bull Droplets were generated on the Auto Droplet Generator using the Bio-Rad ldquoddPCR ldquoSupermix for Probes (No dUTP)rdquo (Bio-Rad cat 186-3024 Control 64079080) and ldquoDroplet Generation Oil for Probesrdquo (Bio-Rad cat 186-4110 Control 64046051) Note desoxyuridine 5-triphosphatase (dUTP) is an enzyme sometimes included in PCR reaction mixtures to minimize PCR carryover contamination

bull Table 3 details the composition of the ddPCR mastermix setup for a given assay and sample where ldquomastermix refers to reaction mixture of manufacturerrsquos proprietary PCR reagent ldquoSupermixrdquo forward and reverse PCR primers for a given assay probe for that assay the sample DNA and sometimes additional water to produce the desired value dilution factor F This setup was used for all assays and samples

bull Non-Template Controls (NTCs) for each assay were included in every analysis bull Droplets were thermal cycled on a ProFlex thermal cycler (Applied Biosystems) for

95 degC for 10 min followed by 60 cycles of 94 degC for 05 min and 61 degC for 1 min then 98 degC for 10 min before a 4 degC hold until the plate was removed

bull Droplets were read on the QX200 Droplet Reader and analysis was performed using the QuantaSoft Analysis Software version 1740917

mtND1 mtBatz mtKav

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

8

bull The numbers of positive and negative droplets at the end of 60 cycles were determined and were exported into a spreadsheet for further analysis Assay-specific intensity thresholds were determined by visual inspection for each assay for each measurement session

Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays

Concentration nDNA Mastermix Microliter

per Reaction 2X ldquoSupermix for Probes (no dUTPs)rdquo 125

5 μmolL Forward Primer 1875 5 μmolL Reverse Primer 1875 5 μmolL Probe (FAM) 125

PCR Grade Water 50 14 Diluted Sample 25 Total Volume 250

Mitochondrial DNA (mtDNA) All the mtDNA measurements used in this study followed the nDNA protocol described above but with different mastermix setups The duplex mtDNA measurements used the setup described in Table 4 The monoplex mtDNA measurements used the setup described in Table 5

Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 125

10 μmolL Forward Primer 09375 10 μmolL Reverse Primer 09375 10 μmolL Probe (VIC) 0625

PCR Grade Water 95 nDNA Mastermix from Table 3 05 Total Volume 250

Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 120

10 μmolL Forward Primer 09 10 μmolL Reverse Primer 09 10 μmolL Probe (FAM) 06

PCR Grade Water 72 Diluted Sample (See Table 7) 24

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

9

Total Volume 240 Sample Preparation

The buffy coat white blood cells used to prepare SRM 2372a materials were purchased from Interstate Blood Bank These materials were from four single-source anonymous individuals supplied to us labeled as Female1 Female2 Male1 and Male2 In our laboratories each donor buffy coat bag was aliquoted (5 mL per aliquot) into sterile 50 mL conical tubes and stored at 4 degC prior to extraction The modified salt-out manual extraction protocol described in Appendix A was performed for each aliquot with rehydration of the DNA in TE-4 buffer From 150 mL to 250 mL of solubilized DNA for each component was accumulated through the extraction of the 5 mL aliquots The DNA was stored in perfluoroalkoxy polymer (PFA) containers that had been bleach sterilized thoroughly rinsed with deionized water rinsed with ethanol and air-dried The minimum volume for production of 2000 units was 120 mL Preliminary estimates of the mass [DNA] of the re-solubilized components were obtained using a Nanodrop microvolume spectrophotometer Dilutions of the component materials were made until the absorbance at 260 nm was approximately 1 corresponding to a nominal mass [DNA] of 50 ngμL of dsDNA7 After addition of the TE-4 buffer the containers were placed on an orbital shaker and gently rotated for several hours prior to refrigerated storage at 4 ordmC The four single-source components were initially screened with the nDNA ddPCR measurement process using four nDNA assays 2PR4 NR4Q ND6 and 22C3 Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate The gravimetric 1rarr4 dilution was prepared by weighing an empty PFA vial and recording the weight then pipetting 30 microL TE-4 into the vial and recording the weight then pipetting 10 microL DNA and recording the final weight The dilution factor was calculated in a spreadsheet These screening measurements verified that the individual component materials were fit-for-purpose The four single-source components were evaluated with the nDNA ddPCR measurement process using the ten nDNA assays Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate Quantitative information was derived from the average of the ten nDNA assays to prepare the component C mixture Thirty milliliters of sample Male2 (estimated at 452 ngμL mass [DNA]) were combined with 80 mL of sample Female2 (estimated at 507 ngμL mass [DNA]) and 10 mL of TE-4 buffer to create a 1 male nDNA to 3 female nDNA producing a 14 malehuman mixture After addition of the TE-4 buffer the component C container was gently rotated on the orbital shaker for several hours prior to refrigerated storage at 4 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

10

Absorbance Measurements

In February 2017 absorbance spectra of the recently prepared fully diluted nominally dsDNA and ssDNA solutions of the three components were acquired with a BioCary 100 spectrophotometer using the measurement protocol described in Appendix B In December 2017 additional spectra were acquired for the remaining bulk solution of components A and B and for SRM 2372 components A and B These solutions had been stored at 4 degC in well-sealed PFA containers There was no remaining bulk component C solution SRM Unit Production

In the morning of the day a component was transferred into the sterile component tubes (Sarstedt Biosphere SC micro tube-05 mL Newton NC) used to package the SRM the container holding the diluted component was removed from the refrigerator and placed on a slow orbital shaker for 2 h The tubes were opened and placed under the laminar flow hood in rows of 5 within 80-hole tube racks for a total of 30 tubes per rack Tubes were placed in every second row of the racks to facilitate filling with an Eppendorf Repeater Xstream pipette (Eppendorf North America Inc Hauppauge NY) fitted with a 1-mL positive displacement tip set to dispense 55 microL per tube The component container was removed from the shaker a magnetic stir-bar was added the container was placed on a magnetic stir plate in the laminar flow hood and the material was stirred gently Filling proceeded until 2200 vials were filled with pipet tip replacement after filling every 200 tubes The filling process consisted of two individuals manually filling each vial within a biosafety cabinet (one physically pipetting each sample the other verifying that each tube was filled) and four individuals tightly closing the lids to each tube on a sterilized and covered bench When all tubes in a rack were filled the rack was moved out of the hood the tubes checked for proper filling their lids closed and the tubes transferred to 100-unit pre-labeled storage boxes As the 100-unit boxes were filled they were transferred to the labeling room where SRM labels were applied by hand and colored inserts were added to the lids red for component A white for component B and blue for component C After all tubes for a given component were labeled the 22 boxes were placed in a locked refrigerator Filling of each of the components took 60 min to 90 min All units were transferred to the refrigerator within 4 hours of beginning the filling process All units of the three materials were allowed to equilibrate for a minimum of two weeks prior to beginning homogeneity analysis Component Homogeneity

The ddPCR measurement protocol described above was used to assess the nDNA homogeneity of the three components One tube from the approximate center of each of the 22 storage boxes was assayed using the NEIF and ND6 assays (see Table 1) A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Three technical replicates were obtained for every tube A total

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

11

of six plates were used two plates per component one with the NEIF assay and one with the ND6 assay Figure 4 details the layout of the samples

Figure 4 Plate Layout Design for Homogeneity Measurements

Component Stability

Stability of the materials was evaluated at 4 degC room temperature (RT asymp 22 degC) and 37 degC Tubes from the homogeneity study were divided among the three temperatures for the stability study Additionally two never-opened tubes (one from box 8 and one from box 12 of each component) were held at each temperature Two tubes at each temperature for each component were opened sampled and resealed 4 times over the course of 8 weeks Measurements were made at week 1 week 3 week 5 and week 8 The measurement on week 8 was from an unopened tube (from box 8) for each of the components at all temperatures The data from the homogeneity study was used as timepoint 0 in the stability study anchored to 4 degC by the storage recommendations A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used for both the NEIF and ND6 assays All temperatures components and both assays fit onto one plate for timepoints 1 to 5 Twelve NTCs placed in rows G and H were included for each assay Figure 5 details the sample layout

Figure 5 Plate Layout Design for Stability Measurements

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

ND6 ND6 ND6

NEIFNEIFNEIFComponent A

Component A

Component B

Component B

Component C

Component C

1 2 3 4 5 6 7 8 9 10 11 12A A-7 A-7 B-7 B-7 C-7 C-7 A-7 A-7 B-7 B-7 C-7 C-7B A-10 A-10 B-10 B-10 C-10 C-10 A-10 A-10 B-10 B-10 C-10 C-10C A-9 A-9 B-9 B-9 C-9 C-9 A-9 A-9 B-9 B-9 C-9 C-9D A-12 A-12 B-12 B-12 C-12 C-12 A-12 A-12 B-12 B-12 C-12 C-12E A-8 A-8 B-8 B-8 C-8 C-8 A-8 A-8 B-8 B-8 C-8 C-8F A-11 A-11 B-11 B-11 C-11 C-11 A-11 A-11 B-11 B-11 C-11 C-11G NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTCH NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

NEIF

4C

RT

37C

ND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

12

Component Certification Measurements

After establishing homogeneity and stability measurements intended to be used to certify nDNA values and inform mtDNA values were started using randomly selected SRM tubes from the boxes listed in Table 7 The ten nDNA assays described in Table 1 and two of the three mtDNA assays described in Table 2 were run simultaneously on the same sample plate A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used in each of the six measurement campaigns

Table 6 Tubes used for Certification Measurements

Component Plate A B C

dd20170711 Box 1 Box 1 Box 1 dd20170712a Box 10 tube a Box 10 tube a Box 10 tube a dd20170712b Box 10 tube b Box 10 tube b Box 10 tube b dd20170713 Box 20 Box 20 Box 20 dd20170714 Box 5 Box 5 Box 5 dd20170801 Box 5 Box 5 Box 5

Every sample was analyzed in duplicate with every assay in all six plates Component A was always located in rows B and C component B in rows D and E component C in rows F and G and NTCs in rows A and H Figure 6 displays the basic layout of the ten nDNA and two mtDNA assays In addition to the plate column assignment scheme shown the three assay groupings were also ordered as Group 2 Group 3 Group 1 and Group 3 Group 1 Group 2 to address potential positional bias

Group 1 Group 2 Group 3 2PR4 POTP NEIF NR4Q D5 ND6 D9 HBB1 ND14 22C3 mtND1 mtBatz 1 2 3 4 5 6 7 8 9 10 11 12 A NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Comp A B Comp A C Comp B D Comp B E Comp C F Comp C G

H NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Figure 6 Basic Plate Layout Design for Certification Measurements

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

13

Mitochondrial DNA Measurements

A ldquoduplexed reactionrdquo was used to determine mtDNAnDNA where 1 μL of one of the nDNA mastermix was added to 49 μL of the mtDNA mastermix resulting in a 1rarr50 volumetric dilution of the nDNA The nDNA was targeted with a FAM-labeled probe while the mtDNA was targeted with a VIC labeled probe The primer and probe concentrations were increased from 5 micromol to 10 micromol in the duplexed reactions Additional testing of the mtDNA assays were performed in monoplex with FAM-labeled probes to determine the additional sources of bias between sampling assay setup and dilution factors Table 7 outlines the different methods examined to determine mtDNAnDNA for the three components

Table 7 Run Parameters for Mitochondrial DNA Droplet

Generation nDNA mtDNA

Plate Component Assay Dilution Reps Assay Dilution Reps dd20170824 A manualDG HBB1 1rarr4 3 mtND1 1rarr200 9 dd20170825 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170829 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170831 AB autoDG HBB1 1rarr3 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr6 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr12 2 mtND1 1rarr10rarr100 3 dd20170907 ABC manualDG HBB1 1rarr3 2 mtBatz 1rarr10rarr100 3 dd20170913 AB manualDG NEIF 1rarr4 3 mtND1 mtBatz mtKav 1rarr10rarr100 3 dd20170913 C manualDG NEIF 1rarr4 3 mtBatz mtKav 1rarr10rarr100 3 Performance in Commercial qPCR Chemistries

Table 8 lists the 11 qPCR chemistries that were commercially available in late 2017 All three of the SRM 2372a components were evaluated with these chemistries using component A of SRM 2372 as the reference standard For each component of SRM 2372a four tubes were pulled from box 22 and combined into one tube A 1rarr5 dilution of all samples was prepared followed by three serial 1rarr2 dilutions A five-point serial dilution of each SRM 2372a component was amplified in triplicate for each chemistry This consisted of the neat material the 1rarr5 dilution then the three serial 1rarr2 dilutions A six-point serial dilution of SRM 2372 Component A was also created for the standard curve for the commercial qPCR chemistries This consisted of the neat material followed by a 1rarr5 dilution then followed by four 1rarr2 dilutions serially The six-point serial dilution series of SRM 2372 Component A was also amplified in triplicate These commercial chemistries were amplified in individual plates on an Applied Biosystems 7500 HID Real-Time PCR Instrument (Foster City CA) following the manufacturersrsquo recommended protocols for reaction mix setup and cycling conditions Figure 7 details the plate layout for the qPCR plates Data was analyzed with the HID Real-Time PCR Analysis Software v12 and further analyzed in a spreadsheet

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

14

Table 8 Commercial qPCR Chemistries Tested with SRM 2372a

Manufacturer Commercial Kit Target Codea

Innogenomics

InnoQuant Long IQh Short IQd

InnoQuantHY Long IQHYh Short IQHYd Y IQHYy

Thermo Fisher Quantifiler Duo Human QFDh Male QFDy

Quantifiler HP Large QFHPh Small QFHPd

Quantifiler Human Human QFHh

Quantifiler Trio Large QFTh Small QFTd Y QFTy

Qiagen

Quantiplex Autosomal QPh

Quantiplex Hyres Autosomal QPHh Y QPHy

Quantiplex Pro Human QPPh Degradation QPPd Y QPPy

Promega

Plexor Autosomal Ph Y Py

PowerQuant Autosomal PQh Degradation PQd Y PQy

a Assays are coded with a short acronym derived from the kit name followed by the following codes for the type of target ldquohrdquo autosomal non-degradation described as ldquoAutosomalrdquo ldquoHumanrdquo or ldquoLongrdquo ldquodrdquo autosomal degradation described as ldquoDegradationrdquo or ldquoShortrdquo ldquoyrdquo Y-chromosome described as ldquoMalerdquo or ldquoYrdquo

1 2 3 4 5 6 7 8 9 10 11 12

A

B 2372-A_57 2372-A_57 2372-A_57 2372a-neat_A

2372a-neat_A

2372a-neat_A

2372a-neat_B

2372a-neat_B

2372a-neat_B

2372a-neat_C

2372a-neat_C

2372a-neat_C

C 2372-A_11

4 2372-A_11

4 2372-A_11

4 2372a-15_A

2372a-15_A

2372a-15_A

2372a-15_B

2372a-15_B

2372a-15_B

2372a-15_C

2372a-15_C

2372a-15_C

D 2372-A_5

7 2372-A_5

7 2372-A_5

7 2372a-110_A

2372a-110_A

2372a-110_A

2372a-110_B

2372a-110_B

2372a-110_B

2372a-110_C

2372a-110_C

2372a-110_C

E 2372-A_2

85 2372-A_2

85 2372-A_2

85 2372a-120_A

2372a-120_A

2372a-120_A

2372a-120_B

2372a-120_B

2372a-120_B

2372a-120_C

2372a-120_C

2372a-120_C

F 2372-A_1

425 2372-A_1

425 2372-A_1

425 2372a-140_A

2372a-140_A

2372a-140_A

2372a-140_B

2372a-140_B

2372a-140_B

2372a-140_C

2372a-140_C

2372a-140_C

G 2372-A_0

7125 2372-A_0

7125 2372-A_0

7125

H NTC NTC NTC

Figure 7 Plate Layout Design for Performance Assessments

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

15

Measurement Results After preparing the SRM 2372a bulk solutions their approximate 50 ngmicroL mass [DNA] were confirmed spectrophotometrically The bulk solutions were then packaged into individual tubes The materials contained in the tubes were characterized for homogeneity stability nDNA copy number [DNA] and mtDNAnDNA ratio The following sections detail the results of these measurements Absorbance Spectrophotometry

By widely accepted convention an aqueous DNA solution having an absorbance of 10 at 260 nm corresponds to a mass [DNA] of 50 ngmicroL for dsDNA and (37 to 40) ngmicroL for ssDNA7 Absorbance at (230 270 280 and 330) nm are traditionally used in the assessment of DNA quality725 Due to minor baseline shifts during acquisition of some spectra it was necessary to bias-adjusted all spectra to have absorbance of 00 at 330 nm Table 9 lists the resulting bias-adjusted values for selected wavelengths and the conventional mass [DNA] conversion of the absorbance at 260 nm The close agreement between the mass concentration values for the nominally dsDNA and the converted-to-ssDNA spectra confirm that the bulk solutions were correctly prepared to be asymp 50 ngμL dsDNA The slight increase in absorbance over the 10 months between acquiring the spectra suggests that like SRM 2372 the tertiary structure of the dsDNA in the SRM 2372a materials may be changing slowly with time Figure 8 displays the spectra

Table 9 Absorbance at Selected Wavelengths

Wavelength nm Form Date Component 230 260 270 280 ngμL dsDNA 2142017 A 0394 0931 0758 0488 466

B 0480 1112 0905 0583 556 C 0451 1006 0815 0524 503

ssDNA 2142017 A 0390 0614 0540 0316 454 B 0457 0723 0634 0372 535 C 0411 0641 0563 0330 474

dsDNA 12142017 A 0402 0956 0775 0503 478 B 0484 1128 0914 0589 564

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

16

Figure 8 Absorbance Spectra of SRM 2372a Components A B and C

dsDNA is denatured to ssDNA by diluting with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to DNA mass concentration

Droplet Volume

Droplet volumes for the two lots of ldquoSupermixrdquo used in the studies described in this report were determined using NISTrsquos Special Test Method 11050S-D as described in Dagata et al19 The lot used for most of the ddPCR analyses produced (07349 plusmn 00085) nL droplets where the ldquoplusmnrdquo value is the combined standard uncertainty The lot used for analyses performed after October 11 2017 produced (07376 plusmn 00085) nL droplets NIST measurements have demonstrated that droplet volumes are not influenced by the presence or absence of DNA in the sample solution Measurements repeated over time on an earlier lot of the Supermix product used in these studies suggest that droplet volumes remain constant over the shelf-life of the lot However droplet volumes for different lots of this product as well as for different Supermix products have been observed to differ by several percent We intend to document these measurements in a future report Other researchers have demonstrated that non-constant droplet volume distributions may contribute to ddPCR measurement imprecision and bias2627 However these effects become significant only with distributions wider than observed using Special Test Method 11050S-D19 and at copydroplet values larger than the λ le 05 copydroplet typically used in our measurements Certification Measurements

Homogeneity Results Table 10 lists the λacute nDNA copy number values adjusted for both droplet volume and sample dilution from the homogeneity measurements The data was analyzed using the three-level Gaussian hierarchical ANOVA model described in Appendix C Table 11 lists the values of the within- and between-tube variance components expressed as percent relative standard deviations At the 95 confidence level all components are homogeneous

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372a-A2141721417-NaOH121417

220 260 300 340

Wavelength nm

2372a-B2141721417-NaOH121517

220 260 300 340

Wavelength nm

2372a-C21417

21417-NaOH

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

17

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 NEIF A-1-01 1632 1529 1511 B-1-01 1790 1909 1885 C-1-01 1493 1385 1510 ND6 1616 1640 1582 NAe 1870 1860 1565 1562 1466

NEIF A-1-02 1531 1525 1549 B-1-02 1712 1719 1704 C-1-02 1516 1513 1577 ND6 1615 1589 1601 1775 1718 1756 1490 1547 1570

NEIF A-1-03 1524 1540 1533 B-1-03 1693 1755 1749 C-1-03 1451 1528 1506 ND6 1582 1152 1555 1824 1483 1896 1485 1505 1471

NEIF A-1-04 1454 1458 1427 B-1-04 1754 1745 1739 C-1-04 1440 1486 1520 ND6 1627 1605 1607 1670 1632 1814 1529 1508 1506

NEIF A-1-05 1465 1540 1504 B-1-05 1757 1687 1735 C-1-05 1523 1472 1483 ND6 1607 1633 1586 1683 1759 1703 1513 1520 1491

NEIF A-1-06 1512 1491 1491 B-1-06 1697 1709 1714 C-1-06 1437 1493 1522 ND6 1511 1483 1524 1693 1723 1722 1492 1481 1461

NEIF A-1-07 1418 1480 1455 B-1-07 1722 1625 1703 C-1-07 1485 1477 1463 ND6 1479 1557 1656 1690 1684 1688 1475 1547 1485

NEIF A-1-08 1496 1560 1501 B-1-08 1884 1802 1772 C-1-08 1433 1405 1478 ND6 1514 1615 1540 1864 1887 1821 1497 1501 1495

NEIF A-1-09 1588 1581 1637 B-1-09 1713 1734 1727 C-1-09 1463 1499 1505 ND6 1474 1438 1454 1709 1756 1758 1503 1491 1516

NEIF A-1-10 1574 1513 1570 B-1-10 1759 1809 1738 C-1-10 1534 1458 1454 ND6 1594 1713 1669 1764 1816 1787 1448 1465 1474

NEIF A-1-11 1557 1508 1633 B-1-11 1749 1782 1740 C-1-11 1533 1514 1555 ND6 1571 1611 1631 1741 1792 1784 1488 1470 1483

NEIF A-1-12 1500 1476 1515 B-1-12 1748 1712 1718 C-1-12 1450 1525 1497 ND6 1626 1586 1608 1797 1614 1790 1496 1483 1519

NEIF A-1-13 1672 1634 1804 B-1-13 1632 1668 1701 C-1-13 1537 1527 1515 ND6 1490 1516 1497 1626 1566 1681 1499 1479 1475

NEIF A-1-14 1551 1568 1538 B-1-14 1752 1535 1673 C-1-14 1515 1490 1494 ND6 1617 1614 1639 1724 1741 1685 1469 1480 1466

NEIF A-1-15b 1793 1722 1733 B-1-15 1871 1835 1887 C-1-15 1511 1568 1594 ND6 1818 1753 1762 1770 1752 1798 1515 1543 1565

NEIF A-1-16 1573 1520 1617 B-1-16 1688 1725 1743 C-1-16 1449 1452 1471 ND6 1634 1627 1575 1744 1697 1710 1543 1448 1491

NEIF A-1-17 1516 1516 1628 B-1-17 1789 1769 1825 C-1-17 1388 1421 1394 ND6 1613 1567 1567 1716 1762 1638 1452 NAe 1435

NEIF A-1-18 1612 1535 1672 B-1-18 1826 1780 1849 C-1-18 1475 1455 1445 ND6 1529 1495 1528 1805 1801 1931 1458 1531 1459

NEIF A-1-19 1598 1514 1596 B-1-19 1751 1759 1880 C-1-19 1468 1525 1446 ND6 1571 1535 1511 1827 1766 1786 1435 1478 1421

NEIF A-1-20 1575 1568 1586 B-1-20 1738 1678 1768 C-1-20 1549 1531 1499 ND6 1652 1545 1627 1752 1747 1768 1499 1465 1468

NEIF A-1-21 1658 1611 1578 B-1-21 1742 1836 1713 C-1-21 1530 1440 1484 ND6 1647 1607 1612 1746 1744 1764 1499 1454 NAe

NEIF A-1-22 1547 1528 1532 B-1-22 1761 1743 1660 C-1-22 1442 1487 1436 ND6 1537 1513 1536 1666 1692 1819 1589 1591 1556

NEIF A-1-15c 1456 1448 1523 ND6 1525 1402 1436

NEIF A-2-15d 1679 1599 1674 ND6 1581 1550 1584

a Tube coded as Component-Tube-Box b These results were identified as a probable error in the preparation or volumetric delivery of the sample

solution into the microfluidic device and were not used in the homogeneity evaluations c Repeat assessment of the solution in sample A-1-15 d Results for a second tube from box 15 analyzed at the same time as the repeat assessment of A-1-15 e Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

18

Table 11 Within- and Between-Tube Relative Standard Deviations

Within-Tube Component NEIF ND6 Between-Tube

A 28 40 33 B 27 38 25 C 23 18 19

Stability Results Table 12 lists the λacute volume- and dilution-adjusted nDNA copy number values from the stability measurements Figure 9 summarizes these results The results for the tubes held at 4 degC room temperature (RT asymp 22 degC) and 37 degC are very similar for all three components The absence of between-temperature differences identifies the between-week differences as attributable to variation in the measurement processes rather than sample instability These results indicate that the SRM 2372a genomic copy number is thermally stable from 4 degC to 37 degC over an extended period

Figure 9 ddPCR Stability Measurements as Function of Time

Each dot-and-bar symbol represents the mean λacute and the approximate 95 expanded uncertainty for all results for a given component at a given storage temperature The thick black horizontal lines represent an approximate 95 confidence interval on the λacute values estimated from the homogeneity data

13

14

15

16

17

18

19

0 1 2 3 4 5 6 7 8

λacute H

aplo

id n

DNA

Copi

es μL

Weeks

A

0 1 2 3 4 5 6 7 8

Weeks

B

4 degC 22 degC 37 degC

0 1 2 3 4 5 6 7 8

Weeks

C

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

19

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Week Temp Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3

1 4 degC NEIF A-10 1568 1441 B-10 1666 1653 C-10 1424 1394 ND6 A-10 1474 1494 B-10 1655 1705 C-10 1461 1394

NEIF A-07 1500 1463 B-07 1718 1640 C-07 1446 1452 ND6 A-07 1479 1479 B-07 1708 1686 C-07 1522 1482

RT (22 degC) NEIF A-12 1674 1608 B-12 1805 1777 C-12 1506 1495 ND6 A-12 1522 1507 B-12 1759 1788 C-12 1478 1461 NEIF A-09 1635 1539 B-09 1791 1793 C-09 1411 NAb ND6 A-09 1543 1519 B-09 1661 1682 C-09 1482 1496

37 degC NEIF A-11 1614 1637 B-11 1748 1744 C-11 1515 1464 ND6 A-11 1494 1502 B-11 1740 1697 C-11 1495 1535 NEIF A-08 1592 1614 B-08 1691 1725 C-08 1528 1500 ND6 A-08 1581 1595 B-08 1727 1699 C-08 1458 1483 3 4 degC NEIF A-10 1520 1706 B-10 NAb 1733 C-10 1427 1473 ND6 A-10 1543 1624 B-10 1670 1755 C-10 1443 1364 NEIF A-07 1537 1570 B-07 1549 1565 C-07 1424 1512 ND6 A-07 1517 1479 B-07 1705 1723 C-07 1483 1439

RT (22 degC) NEIF A-12 1521 1607 B-12 1817 1842 C-12 1566 1471 ND6 A-12 1644 1597 B-12 1845 1786 C-12 1504 1511 NEIF A-09 1432 1626 B-09 1825 1856 C-09 1392 1387 ND6 A-09 1539 1528 B-09 1734 1737 C-09 1535 1465

37 degC NEIF A-11 1651 1685 B-11 1759 1779 C-11 1437 1440 ND6 A-11 1680 1594 B-11 1806 1806 C-11 1471 1509 NEIF A-08 1580 1613 B-08 1740 1791 C-08 1582 1590 ND6 A-08 1592 1524 B-08 1769 1694 C-08 1468 1520 5 4 degC NEIF A-10 1486 1440 B-10 1626 1551 C-10 1381 1330 ND6 A-10 1555 1577 B-10 1673 1625 C-10 1547 1395 NEIF A-07 1387 1429 B-07 1723 1787 C-07 1330 1314 ND6 A-07 1538 1522 B-07 1862 1909 C-07 1452 1458

RT (22 degC) NEIF A-12 1601 1631 B-12 1718 1752 C-12 1546 1541 ND6 A-12 1699 1677 B-12 1772 1721 C-12 1490 1471 NEIF A-09 1381 1484 B-09 1675 1694 C-09 1371 1376 ND6 A-09 1625 1615 B-09 1904 1815 C-09 1468 1472

37 degC NEIF A-11 1456 1517 B-11 1720 1704 C-11 1505 1493 ND6 A-11 1650 1658 B-11 1148 1152 C-11 1527 1600 NEIF A-08 1590 1666 B-08 1745 1723 C-08 1469 1540 ND6 A-08 1235 1244 B-08 1809 1749 C-08 1508 1529 8 4 degC NEIF A-08a 1667 1632 1589 B-08a 1880 1884 1940 C-08a 1525 1475 1542 ND6 A-08a 1599 1593 1590 B-08a 1859 1897 1972 C-08a 1550 1502 1520

RT (22 degC) NEIF A-08b 1693 1666 1634 B-08b 1769 1884 1771 C-08b 1595 1557 1545 ND6 A-08b 1644 1620 1647 B-08b 1862 1738 1811 C-08b 1611 1585 1535

37 degC NEIF A-08c 1620 1576 1608 B-08c 1934 1952 1956 C-08c 1595 1600 1569 ND6 A-08c 1576 1553 1532 B-08c 1889 1844 1889 C-08c 1569 1514 1581

a Tube coded as Component-Tube-Box b Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

20

Ten-Assay Certification Results The original intention was to certify the copy number [DNA] of the three components with results from five independently prepared sets of samples analyzed over four days Each sample of the three components was from a new tube from different boxes and was evaluated with 10 nDNA assays The assay positions were alternated across these five ldquo10-assay certificationrdquo plates The λ results for these five plates are systematically somewhat larger than those for the homogeneity and stability data On review we realized that a different workspace and its suite of pipettes had been used to prepare the 1rarr10 dilution of the 1rarr4 diluted components into the reaction mastermix The ldquohighrdquo workstation was used for stability and homogeneity measurements while the ldquolowrdquo workstation was used for certification measurements Gravimetric testing of the pipettes established that the volumes delivered by these alternate pipettes was sufficiently biased to account for the observed λ differences A sixth ldquocertificationrdquo plate was prepared using the original (and now re-verified) pipettes The λ results or this plate agreed well with the homogeneity and stability data Table 13 lists the λacute volume- and dilution-adjusted nDNA copy number values for all six of the ten-assay datasets Figure 10 displays the extent of agreement among the ten assays While there are small-but-consistent differences between the assays there is only a 4 difference between the assays providing smallest and largest λ

Figure 10 Comparison of Assay Performance

The blue dot-and-bars represent the mean and its combined standard uncertainty for the A B and C components for each assay relative to the over-all grand mean for each component The black dot-and-bars represent combination of three component results for each assay The solid horizontal line represents the median difference of the assay means from the grand mean the dotted horizontal lines bound an approximate 95 confidence region about the median

2PR4POTPNEIFR4Q5 D5ND6 D9 HBB1ND14 22C3-10

-8

-6

-4

-2

0

2

4

6

8

10

Rela

tive

Diffe

renc

e fr

om M

ean

Assays

Component MeansAssay MeansMedianU95

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

21

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

7112017 2PR4 1667 1698 1869 1960 1630 1706 POTP 1655 1685 1850 1781 1596 1576 NEIF 1647 1665 1709 1707 1574 1538 R4Q5 1594 1658 1856 1787 1590 1555 D5 1655 1683 1848 1829 1610 1607 ND6 1629 1647 1828 1826 1589 1548 D9 1655 1652 1861 1858 1569 1572 HBB1 1644 1611 1825 1869 1565 1681 ND14 1645 1707 1871 1773 1564 1492 22C3 1545 1603 1776 1902 1548 1655

7122017 2PR4 1668 1739 2174 2231 1499 1497 POTP 1587 1702 2202 2292 1576 1611 NEIF 1678 1614 2172 2216 1545 1560 R4Q5 1628 1762 NA 2078 1584 1546 D5 1738 1649 2230 2192 1577 1529 ND6 1667 1678 2094 2034 1599 1478 D9 1695 1653 1949 2110 1587 1504 HBB1 1757 1750 1971 2123 1520 1551 ND14 1771 1764 2123 2066 1479 1438 22C3 1762 1759 2107 2046 1512 1494

7122017 2PR4 1638 1667 1810 1806 1452 1424 POTP 1676 1638 1865 1801 1347 1397 NEIF 1616 1624 1806 1786 1395 1350 R4Q5 1559 1587 1878 1768 1442 1459 D5 1608 1652 1824 1790 1448 1602 ND6 1611 1589 1768 1670 1463 1435 D9 1517 1607 1811 1749 1402 1433 HBB1 1557 1582 1763 1741 1474 1481 ND14 1604 1625 1865 1781 1466 1482 22C3 1675 1616 1815 1866 1439 1568

7132017 2PR4 1888 1939 1784 1821 1470 1444 POTP 1868 1815 1735 1813 1411 1427 NEIF 1953 1902 1778 1813 1512 1439 R4Q5 1779 1871 1835 1749 1452 1468 D5 1924 2124 1794 1848 1567 1524 ND6 1861 1787 1700 1711 1480 1424 D9 1824 1844 1799 1779 1442 1459 HBB1 1809 1799 1765 1769 1430 1467 ND14 1864 1857 1754 1725 1413 1411 22C3 1930 1942 1792 1814 1486 1480

7142017 2PR4 1597 1590 1718 1768 1485 1458 POTP 1533 1404 1753 1698 1422 1459 NEIF 1462 1533 1773 1792 1438 1465 R4Q5 1493 1519 1777 1740 1439 1468 D5 1538 1586 1803 1795 1515 1457 ND6 1492 1516 1718 1693 1409 1412 D9 1528 1538 1802 1743 1469 1474 HBB1 1548 1581 1834 1745 1453 1428 ND14 1513 1514 1731 1714 1465 1425 22C3 1562 NA 1842 NA 1363 1393

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

22

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

101217 2PR4 1431 1411 1791 1770 1357 1364 POTP 1407 1395 1746 1820 1447 1361 NEIF 1487 1398 1803 1803 1448 1404 R4Q5 1450 1439 1767 1739 1409 1411 D5 1394 1422 1745 1773 1389 1429 ND6 1396 1393 1824 1729 1415 1417 D9 1414 1390 1752 1744 1431 1380 HBB1 1401 1382 1813 1754 1381 1336 ND14 1412 1460 1648 1671 1403 1387 22C3 1435 1410 1818 1741 1435 1372

Estimation of Copy Number Concentration It was decided that all three sets of results (homogeneity stability and certification) would be used to derive the certified values with the λ of the first five ldquoten-assayrdquo plates adjusted by a data-driven estimate of the volumetric bias This bias was incorporated into the statistical model providing better estimates of the λ means while capturing all of the variance components (homogeneity stability between-assay and between data sets) as random effects using Bayesian MCMC programmed in OpenBUGS28 Table 14 lists the estimated values and their uncertainties See Appendix C for details

Table 14 Measured Copy Number Concentrations

Genomic Copies per Nanoliter of Solution Component Mean a u(Mean) b U95(Mean) c CV95

d A 151 07 14 93 B 175 03 06 34 C 145 05 10 69

a Mean the experimentally determined most probable estimate of the true value b u(Mean) the standard uncertainty of the Mean c Mean plusmnU95(Mean) the interval which is believed with about 95 confidence to contain the true value d CV95 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the counting unit one29 2) the validity of the Poisson endpoint transformation for digital PCR endpoint assays when applied to samples providing le 05 copies per droplet 3) the determination that the copies are dispersed as dsDNA and 4) calibrated mean droplet volume measurements made at NIST during sample dilution and mastermix preparation Conversion of Copy Number to Mass Concentration As described in Reference 18 metrologically valid conversion from copies per nanoliter to nanograms per microliter requires quantitative estimates for the number of nucleotide bases per reference human genome and the average nucleotide mass Using the values estimated in that report and asserting that each target copy corresponds to one HHGE the required transformation is

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

23

[DNA] ngmicroL

= [DNA] HHGE

nLtimes

3301 ngHHGE

The combined standard uncertainty is

119906119906 [DNA] ng

microL = 1199061199062 [DNA] HHGE

nL + 0531100

times[DNA] HHGE

nL 2

Both the measured and the conversion factor uncertainties are associated with ldquolargerdquo degrees of freedom Therefor the approximate 95 confidence uncertainty is estimated

11988011988095 [DNA] ng

microL = 2 times 119906119906

[DNA] ngmicroL

Table 15 lists the transformation of the Table 14 copy number values These values agree well with the conventional 260 nm absorbance values averaging (4 plusmn 5) larger

Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution

ngmicroL Component Value a u(Measurement)b u(Conversion)c U95(Value)d CV95

e A 498 23 03 47 93 B 578 10 03 21 36 C 479 17 03 33 70

a Estimated true value b Standard uncertainty component from the ddPCR measurements c Standard uncertainty estimated from the relative uncertainty of the conversion factor d Value plusmnU95(Value) is believed with about 95 confidence to contain the true value e 100U95(Value)Value the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the measured estimates of copies per nanoliter and 2) determinations made by us from internationally recognized literature resources of the average number of nucleotide bases per HHGE and the mean molecular weights of the bases18 Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)

While not intended to provide certified values mtDNAnDNA was measured as part of 13 experiments over 5 months Measurements were made using dilution-adjusted λ for the mtND1 mtBatz andor mtKav mitochondrial assays referenced against the dilution-adjusted λ of either the NEIF or HBB1 nDNA assays By chance both DNAs in the component C mixture have a binding site mutation that silences the mtND1 probe Figure 11 graphically summarizes the results by assay and component Table 16 lists the data Table 17 summarizes results from a random effects model that accounts for between-assay variability See Appendix C for details

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

24

Due to the limited number of mtDNA assays used in these measurements these results are metrologically traceable only to the assays used

Figure 11 mtDNAnDNA for Components A B and C

Each dot-and-bar symbol represents a mean mtDNAnDNA and its approximate 95 expanded uncertainty The dotted horizontal lines represent approximate 95 confidence intervals on the ratios

Table 16 mtDNAnDNA Ratios

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

20170711a 1 mtND1 1931 1870 2461 2401 1 mtBatz 1845 1764 2555 2608 3201 3299

20170711b 1 mtND1 1810 1906 2229 2200 1 mtBatz 1785 1837 2108 2192 3347 3287

20170712 1 mtND1 1743 1761 2127 2198 1 mtBatz 1829 1847 2455 2506 3217 3384

20170713 1 mtND1 1716 1800 2211 2199 1 mtBatz 1740 1813 2380 2249 3200 3363

20170714 1 mtND1 1923 1835 2105 1943 1 mtBatz 1935 1851 2136 2119 3294 3013

20170814 1 mtND1 1780 1776 1706 2046 2057 2024 1 mtBatz 1804 1751 1738 1998 2043 1952 2927 2776 2762 1 mtKav 1769 1786 1761 2077 2117 2041 3044 2830 2825

20170824 1 mtND1 1947 1928 1922 2 mtND1 1829 1895 1851 3 mtND1 1674 1617

20170825 1 mtBatz 1667 1813 1708 1598 1928 1916 1915 2832 2840 2788 2725 2 mtBatz 1566 1687 1663 1589 1947 1946 1961 1909 2835 2790 2736 2679 3 mtBatz 1643 1668 1604 1573 1953 1952 1941 1881 2803 2773 2757 2752

20170829 1 mtBatz 1719 1716 1670 1666 1865 1863 1828 1828 2599 2508 2551 2609 2 mtBatz 1581 1573 1601 1612 1944 1964 1947 1895 2617 2483 2626 2602 3 mtBatz 1546 1542 1538 1587 1908 1909 1940 1839 2665 2535 2570 2650

20170831 1 mtND1 1617 1599 1688 1783

ND1

ND1

ND1Ka

v

Kav

Kav

Batz

Batz

Batz

A

B

C

160

210

260

mtD

NAn

DNA

Ratio

Assay

Bind

ing s

ite m

utat

ion

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

25

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

2 mtND1 1622 1579 1823 1631 3 mtND1 1575 1528 1920 1887

20170907 1 mtBatz 1584 1544 1890 1886 2501 2522 20170913 1 mtND1 1757 1724 1768 1990 1921 1982

1 mtBatz 1743 1671 1766 1990 1924 1958 2805 2825 2829 1 mtKav 1692 1662 1767 1903 1831 1919 2709 2738 2727

20171016 1 mtND1 1924 1944 1893 2304 2341 2332 1 mtBatz 1786 1752 1743 2048 2078 2067 2797 2642 2792 1 mtKav 1530 1542 1523 2399 2479 2493 2804 2789 2842 2 mtBatz 1580 1557 1592 2715 2606 2687 2 mtKav 2103 2100 2049 2799 2628 2559 3 mtND1 1749 1769 1777 2065 2166 2164 3 mtBatz 1965 1889 1947 2314 2280 2258 2639 2535 2657 3 mtKav 1813 1864 1864 1867 1987 1971 2953 2757 2802

Table 17 mtDNAnDNA Summary Results

mtDNAnDNA Component Value a u(Value)b U95(Value)c CV95

d A 1737 16 38 17 B 2058 22 44 21 C 2791 23 47 17

a Estimated true value b Standard uncertainty c Value plusmnU95(Value) is believed with about 95 confidence to contain the true value d 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) Performance in Commercial qPCR Chemistries

Traditional qPCR assays only quantify relative to a reference standard Thus qPCR assays do not provide absolute [DNA] for the individual materials but they can provide [DNA] of one component relative to another Figure 12 summarizes the BA and CA ratios for the commercial autosomal (Human) targets from the eleven assays evaluated (see Table 8) The Y-chromosome (Male) assays cannot be summarized in this manner as the (all female) component B does not include a Y chromosome Figure 13 summaries the mass [DNA] results for the (all male) component A and (male-female mixture) component C Since this material was prepared as a nominal 13 malefemale mixture the expected malehuman (total) ratio is 14 sjc As displayed in these figures there is intrinsic variability between the SRM 2372a components and between the commercial qPCR chemistries examined This reflects the difference in the targets and amplification chemistries for each of the commercial kits Many of the commercial chemistries on the market to date contain multi-copy targets for both autosomal and Y-chromosomal quantification to increase the sensitivity of the assays

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

26

Figure 12 qPCR Autosomal Quantitation Performance Summary

Open circles denote the BA CA ratios the 17 commercial qPCR assays evaluated each labeled with the code listed in Table 8 The solid black circle represents ratios as estimated by ddPCR The red lines bound the approximate 95 confidence intervals on the ddPCR-estimated ratios

Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary

Open squares denote the ngμL values for SRM 2372a Components A (male) and C (mixture of male and female) for the seven commercial Y-chromosome qPCR assays evaluated each labeled with the code listed in Table 8 The dotted red lines represent malehuman ratios of 13 14 and 15

Appendix D provides a graphical analysis of all the data generated in the performance assessments for each of the targets amplified in the eleven commercial qPCR chemistries and the [DNA] values for all of the assays

dd

IQ_h

IQ_d

IQHY_h

IQHY_d

QFD_h

QFH_h QFHP_h

QFHP_dQFT_h

QFT_d

QPH_h

QPP_hQPP_d

P_h

PQ_h

PQ_d

06

08

10

12

14

07 09 11 13

CA

BA

IQHY_y

QFD_y

QFT_y

QPH_y

QPP_y

P_y

PQ_y

1413

15

10

12

14

16

18

30 40 50 60 70

[DNA

] C n

gμL

[DNA]A nguL

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

27

Qualification of ddPCR as Fit-For-Purpose We have documented the requirements for establishing metrological traceability of ddPCR human genomic assay results expressed as copies per nanoliter of sample in a series of three peer-reviewed papers151617 Through the experimental evidence they are based upon these papers demonstrate that ddPCR can accurately enumerate the number of independently migrating genomic copies in an aqueous extract Our most recent paper establishes the metrological traceability of these ddPCR results when transformed into the nanogram per microliter units used by the forensic and clinical communities18 However our previous studies have not addressed a fundamental requirement for correctly interpreting digital PCR measurements In broad simplification end-point assay systems count the number of independently partitioning entities regardless of the number of copies the entities contain With assays that target just one site per HHGE entities that migrate as ssDNA will yield counts that are twice the value had they migrated as dsDNA30 Should some originally dsDNA be converted over time to ssDNA ddPCR measurements of the number of independently migrating entities would increase although the number of HHGE would be unchanged Measurements of mass [DNA] provided by absorbance spectrophotometry and ddPCR reflect different properties absorbance at 260 nm is sensitive to the spatial configuration of the nucleotides as well as their number while ddPCR counts independently migrating entities The SRM 2372a materials have absorbance spectra shown in Figure 8 consistent with being predominantly dsDNA The SRM 2372 materials now have the absorbance spectra shown in Figure 14 which are consistent with them being entirely ssDNA

Figure 14 Absorbance Spectra of SRM 2372 Components A and B dsDNA is denatured to ssDNA by diluting the source material with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to mass [DNA]

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372-A200720122012-NaOH121417

0

4

6

0

4

6

220 260 300 340

Wavelength nm

2372-B200720122012-NaOH121517

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

28

To ensure the validity of ddPCR-based value assignments it is thus necessary to evaluate the nature of the partitioning entities Should a solution of what is thought to be dsDNA contain an appreciable fraction of ssDNA then the ddPCR results would require adjustment31 To avoid further delay in making SRM 2372a available to the forensic community the following sections detail a series of experiments that have enabled us to conclude with high confidence that the SRM 2372a components have a dsDNA conformation and that for this material ldquoentitiesrdquo can be considered dsDNA ldquocopiesrdquo The remaining stock of the SRM 2372 components also have a dsDNA conformation however the strands in the old materials are much less robustly connected than they are in the new materials Digital Polymerase Chain Reaction Assays

Three nDNA assays were used HBB1 NEIF and ND6 The HBB1 and NEIF assays reproducibly yield very similar results for DNA extracts from many different donors The HBB1 target is at the tip of the short (p) arm of chromosome 11 The NEIF target is in the centromere region of chromosome 2 The ND6 assay was selected for use because it typically yields the lowest results of the ten nDNA assays used to certify the SRM 2372a components (see Figure 10 below) The ND6 target is about halfway down the long (q) arm of chromosome 6 Heat Treatments (Melting of the DNA into Single Strands)

In all experiments the non-heated control and the heat-treated diluted solutions were stored in the same type of PCR reaction tube The starting point for the heat treatment was to determine if there was a difference between snap-cooling on ice and slower cooling at 25 degC Two tubes of the diluted material were placed in a GeneAmp 9700 thermal cycler (Applied Biosystems) and cycled up to 98 degC and held for 15 min After 15 min one tube was removed and placed on ice while the remaining tube was cooled at 1 degCs in the thermal cycler to 25 degC then placed on ice until prepared for ddPCR Additional heat treatments included cycling samples on a GeneAmp 9700 thermal cycler

a) from 98 degC for 5 min then cooling to 25 degC for 2 min During the 25 degC hold a randomly selected tube was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 45 min at 98 degC

b) from 94 degC for 1 min then cooling to 25 degC for 1 min During the 1 min hold at 25 degC a random sample was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 9 min at 94 degC

c) from 90 degC for 1 min then cooling to 25 degC for 1 min for a total of 20 cycles Samples were pulled at 25 degC after 1 min 3 min 5 min 10 min and 20 min and placed on ice

After the above experiments heat treatments were performed using a ProFlex thermal cycler This thermal cycler has six temperature zones that can be le 5 degC apart This enables heating

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

29

tubes at the same time but at different temperatures After treatment tubes were cooled at 5 degCs to 4 degC in the thermal cycler While designed for use with PCR plates a MicroAmp 96-well tray for the VeriFlex block enabled its use with single PCR reaction tubes Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a

As a first step in evaluating whether the DNA in the SRM materials disperse as dsDNA or ssDNA a direct comparison of the five materials was made using the HBB1 and NEIF assays Table 18 reports summary statistics for the ddPCR measurements reported as λ copiesdroplet and the estimated mass [DNA] values from the ddPCR results18 from the conventional dsDNA relationship7 and from the conventional ssDNA relationship after denaturing with 04 molL NaOH89 While the mass [DNA] values differ somewhat between the three methods there is no evidence for the SRM materials segregating exclusively or even mostly as ssDNA The extent of the hypochromic interactions among the nucleotides does not accurately reflect the extent to which the two strands of dsDNA have separated

Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a

λ copies per droplet λ copies per droplet ngmicroL Sample Assay Nrep

a Mean CVb ncmbc Mean CVb ddPCRd dsDNAe ssDNAf

2372-A HBB1 4 03645 19 8 03644 20 566g 524 57 2372-A NEIF 4 03644 23 2372-B HBB1 4 03513 23 8 03487 35 580h 536 61 2372-B NEIF 4 03461 47 2372a-A HBB1 4 02603 27 8 02649 30 474 465 454 2372a-A NEIF 4 02694 24 2372a-B HBB1 4 02908 34 8 02891 37 517 556 534 2372a-B NEIF 4 02873 45 2372a-C HBB1 4 02757 28 8 02711 28 485 502 474 2372a-C NEIF 4 02665 16

a Number of technical replicates for the given assay b CV = 100(standard deviation)mean c Combined number of technical replicates ignoring potential differences between the assays d Estimated from the ddPCR λ measurements listed in this Table18 e Estimated from the conventional relationship between the mass [DNA] of dsDNA and absorbance at

260 nm7 using absorbance spectra acquired shortly after solutions were produced f Estimated from the relationship between the mass [DNA] of ssDNA and absorbance at 260 nm after 11

dilution with 04 molL NaOH89 g Values adjusted to measured (132 plusmn19) volume loss of archived units due to evaporation h Values adjusted to measured (71 plusmn25) volume loss of archived due to evaporation

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

30

Quantitative Evaluation of ssDNA Content

Absorbance spectrophotometry is routinely used to quantitatively assess DNA tertiary structure3233 On storage ldquoat physiological pH the intrastrand repulsion between the negatively charged phosphate groups forces the double helix into more rigid rodlike conformationrdquo34 suggesting a mechanism for the observed slow changes in the SRM 2372 materials As demonstrated in Table 18 absorbance at 260 nm does not necessarily correlate with ddPCR behavior Yet ldquofurthermore the repulsion between phosphate groups on opposite strands tends to separate the complementary strandsrdquo34 Therefore a method that is independent of hypochromism is needed to adequately evaluate the proportion (if any) of ldquotruerdquo (completely disjoint) ssDNA in the SRM materials Circular dichroism spectroscopy is sensitive to the conformational structures of DNA in solution35 and different DNA configurations migrate differentially in agarose gel electrophoresis7 Both techniques have been used to qualitatively discriminate ssDNA from dsDNA in ldquopurerdquo single-sequence materials but neither technique appears to be able to quantify small proportions of ssDNA in a complex multi-chromosome genomic material Wilson and Ellison have proposed a real-time chamber digital PCR (cdPCR) technique that can estimate excess proportions of ssDNA copies from the crossing-threshold distribution31 However their method is insufficiently sensitive for low ssDNA proportions It is well known that heating DNA beyond the melting temperature separates mammalian dsDNA into ssDNA36 albeit we have shown using cdPCR that boiling dsDNA for some minutes also renders some target sites non-amplifiable15 In the following section we document a ddPCR-based approach for estimating the proportion if any of ssDNA in a sample that involves less destructive heating While the method is not yet optimized the results from these preliminary studies demonstrate that the SRM 2372 and 2372a components disperse into the ddPCR droplets as dsDNA ddPCR Measurement of ssDNA Proportion Given two otherwise identical aliquots of a given DNA material one subjected to some treatment and the other untreated the ratio between the average number of copies per droplet of treated material and that of its untreated sibling λtλu quantitatively estimates the change in the number of independently dispersing entities Should the treatment reduce the number of accessible amplifiable targets (AAT) λtλu will be less than 1 Reduction in AAT may result from damage to the target itself andor conformational changes to the fragment containing the target that render the target inaccessible or non-amplifiable16 Assuming no denaturation of dsDNA to ssDNA either because of an ineffective treatment or the material having started as ldquopurerdquo ssDNA if the treatment does not itself damage targets then λtλu will equal 1 Should treatment convert at least some dsDNA to ssDNA without damaging targets the number of AAT will increase and λtλu will be greater than 1 Assuming complete conversion of ldquopurerdquo dsDNA to ldquopurerdquo ssDNA without other damage then λtλu will equal 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

31

Values greater than 1 but less than 2 indicate some conversion of dsDNA to ssDNA but also some combination of (a) incomplete denaturation of dsDNA to ssDNA (b) partial renaturation of ssDNA to dsDNA after treatment but before ddPCR analysis (c) damage to targets during the process or ndash of greatest interest ndash (d) the original sample material was an impure mixture of dsDNA and ssDNA The following discussions graphically summarize sets of λtλu followed by tabular listings of the relevant experimental conditions ddPCR measurements and the numerical λtλu values The information in these tables includes

Plate Index Experiment identifier indicates date performed in YYYYMMDD format Material Sample material SRM 2372-A 2372-B 2372a-A 2372a-B or 2372a-C Assay ddPCR assay HBB1 NEIF or ND6 degC Denaturation temperature in degC Min Length of time the treated material was held at denaturation temperature in

minutes Cool Method used to cool treated material from denaturation temperature Snap

Slow R1 or R5 where ldquoSnaprdquo indicates cooling on ice ldquoSlowrdquo cooling at room temperature to 25 degC ldquoR1rdquo cooling in a thermal cycler to 25 degC at 1 degCs and ldquoR5rdquo cooling in a thermal cycler to 4 degC at 5 degCs

F Dilution factor expressed as (volume Material) per (final volume of diluted solution) 12 14 or 18

n Number of technical replicates Mean Mean of technical replicates CV Coefficient of Variation expressed as a percentage 100 Standard Deviation

Mean

λtλu the copies per droplet result of the treated solution λt divided by the copies per droplet result of its untreated sibling λu

u(λtλu) Standard uncertainty of λtλu 119906119906(120582120582t 120582120582ufrasl ) = 120582120582t 120582120582ufrasl CV(120582120582t)100

2

+ CV(120582120582u)100

2

Proof of Principle Figure 15 summarizes two replicate sets of measurements performed to assess whether the SRM 2372 and 2372a materials respond similarly whether the denatured materials renature if allowed to cool slowly and whether results are (reasonably) repeatable Table 19 details the experimental conditions The λtλu values for the four materials are similar all much greater than 1 but not approaching 2 Most of the ratios are on or close to the ldquoslow cooling is the same as snap coolingrdquo line suggesting that renaturation is not a significant concern While the replicate results for the SRM 2372-A and 2372a-A materials do not overlap as closely as those for SRM 2372-B and 2372a-B similar treatment conditions produce similar results

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

32

Figure 15 Comparison of Snap-Cooling to Slow-Cooling

Symbols mark the means of replicate measurements made using the same dilutions on successive days with bars representing plusmn one standard combined uncertainty Results are labeled by material ldquoArdquo and ldquoBrdquo for SRM 2372 components and ldquoaArdquo ldquoaBrdquo and ldquoaCrdquo for SRM 2372a components The diagonal black line represents equality between the snap- and slow-cooling treatments

Table 19 Comparison of Snap-Cooling to Slow-Cooling

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171218 2372-A HBB1 14 4 03830 29

dd20171218 2372-A HBB1 98 15 Slow 14 2 05644 18 1474 0051 dd20171218 2372-A HBB1 98 15 Snap 14 4 05852 15 1528 0051

dd20171218 2372a-A HBB1 14 4 03029 18

dd20171218 2372a-A HBB1 98 15 Slow 14 4 04430 21 1463 0040 dd20171218 2372a-A HBB1 98 15 Snap 14 4 04454 33 1471 0055

dd20171218 2372-B HBB1 14 4 03594 31

dd20171218 2372-B HBB1 98 15 Slow 14 4 05806 18 1615 0058 dd20171218 2372-B HBB1 98 15 Snap 14 4 05825 13 1620 0055

dd20171218 2372a-B HBB1 14 4 03383 15

dd20171218 2372a-B HBB1 98 15 Slow 14 4 05043 38 1491 0061 dd20171218 2372a-B HBB1 98 15 Snap 14 4 05254 20 1553 0039

dd20171219 2372-A HBB1 14 4 03357 23

dd20171219 2372-A HBB1 98 15 Slow 14 4 05340 16 1591 0046 dd20171219 2372-A HBB1 98 15 Snap 14 4 05340 12 1591 0042

A

aA

B

aB

A

aA

B

aB

14

15

16

17

14 15 16 17

λ tλ

u Sn

ap C

oole

d

λtλu Slow Cooled

Rep 1Rep 2Snap = Slow

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

33

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171219 2372a-A HBB1 14 4 02735 24

dd20171219 2372a-A HBB1 98 15 Slow 14 4 04151 03 1518 0037 dd20171219 2372a-A HBB1 98 15 Snap 14 4 03918 21 1433 0045 dd20171219 2372-B HBB1 14 4 03346 22

dd20171219 2372-B HBB1 98 15 Slow 14 4 05476 17 1637 0045 dd20171219 2372-B HBB1 98 15 Snap 14 4 05480 29 1638 0059 dd20171219 2372a-B HBB1 14 4 03158 19

dd20171219 2372a-B HBB1 98 15 Slow 14 4 04798 16 1519 0038 dd20171219 2372a-B HBB1 98 15 Snap 14 4 04856 17 1538 0039

Time at Denaturation Temperature Figure 16 summarizes λtλu for the SRM 2372-B material as a function of denaturation temperature and the cumulative time the solution is held at that temperature Table 20 details the experimental conditions

Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures

The symbols represent the ratio of the mean result of three technical replicates of the untreated and each of the treated solutions the bars represent combined standard uncertainties The lines denote regression fits to the function λtλu = atimese-bt where t is the cumulative time Values and the standard uncertainties of the coefficients are listed in the legend

12

14

16

18

20

0 5 10 15 20

λ t λ

u

Cummulative Treatment Time min

degC a u (a ) b u (b )90 1804 0039 00198 0000494 1788 0029 00099 0002998 1855 0019 00067 00019

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

34

For all three of the denaturation temperatures studied the relationship between λtλu and cumulative time t can be modeled as

λtλu = ae-bt

where a is the extrapolated value of λtλu at t = 0 and b is the rate of change in AAT per minute As expected the rate of loss due to the treatment increases with increasing treatment temperature and cumulative time at that temperature37 Given the much longer treatment time for the 98 degC series the agreement among the three t = 0 parameter (a) values is remarkable This suggests that the first 1 min of treatment efficiently separates dsDNA into ssDNA The pooled relative standard uncertainty of these regression parameter is 17 somewhat smaller than the 25 pooled relative standard uncertainties of the individual λtλu However the plusmnstandard uncertainty interval of the 1-min λtλu results for the 94 degC and 90 degC both overlap the extrapolated t = 0 value suggesting that reliable results can be obtained using a single treatment of short duration

Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171228 2372-B HBB1 14 3 03650 17 dd20171228 2372-B HBB1 98 5 R1 14 3 06157 12 1687 0035 dd20171228 2372-B HBB1 98 10 R1 14 3 05493 08 1505 0028 dd20171228 2372-B HBB1 98 15 R1 14 3 05019 13 1375 0029 dd20171228 2372-B HBB1 98 20 R1 14 3 04583 21 1256 0034 dd20171228 2372-B HBB1 98 25 R1 14 3 04154 32 1138 0041 dd20171228 2372-B HBB1 98 30 R1 14 3 03784 14 1037 0023 dd20171228 2372-B HBB1 98 35 R1 14 3 03408 10 0934 0019 dd20171228 2372-B HBB1 98 40 R1 14 3 02980 01 0816 0014 dd20171228 2372-B HBB1 98 40 R1 14 3 02815 08 0771 0015

dd20171229 2372-B HBB1 14 3 03708 27

dd20171229 2372-B HBB1 94 1 R1 14 3 06598 26 1779 0056 dd20171229 2372-B HBB1 94 2 R1 14 3 06553 32 1767 0063 dd20171229 2372-B HBB1 94 3 R1 14 3 06301 20 1699 0045 dd20171229 2372-B HBB1 94 4 R1 14 3 06252 14 1686 0037 dd20171229 2372-B HBB1 94 5 R1 14 3 06577 14 1773 0039 dd20171229 2372-B HBB1 94 6 R1 14 3 06156 07 1660 0030 dd20171229 2372-B HBB1 94 7 R1 14 3 06118 10 1650 0033 dd20171229 2372-B HBB1 94 8 R1 14 3 06256 12 1687 0035 dd20171229 2372-B HBB1 94 9 R1 14 3 05999 10 1618 0032

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

35

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180104 2372-B HBB1 14 4 03713 25 dd20180104 2372-B HBB1 90 1 R1 14 3 06916 22 1863 0052 dd20180104 2372-B HBB1 90 3 R1 14 3 06465 06 1741 0032 dd20180104 2372-B HBB1 90 5 R1 14 3 06431 20 1732 0045 dd20180104 2372-B HBB1 90 10 R1 14 3 06014 25 1620 0049 dd20180104 2372-B HBB1 90 15 R1 14 3 06024 10 1623 0032 dd20180104 2372-B HBB1 90 20 R1 14 3 06003 14 1617 0035

Temperature Figure 17 summarizes the effect of denaturation temperature using a single treatment of 05 min on the 2372-B and 2372a-B materials Table 21 details the experimental conditions

Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

The symbols represent the ratio of the mean result of four technical replicates of the untreated and three replicates of the treated solutions the bars represent combined standard uncertainties In all cases the duration of the treatment was 05 min The lines denote regression fits to the linear function λtλu = a +bT where Tis the denaturation temperature

The SRM 2372 and 2372a materials respond very differently to changes in denaturation temperature The older spectrophotometrically ssDNA material suffers somewhat greater loss in λtλu as temperature increases The new spectrophotometrically dsDNA material suffers much greater loss in λtλu as temperature decreases

14

15

16

17

18

19

20

82 86 90 94 98

λ tλ

u

Temperature degC

2372-B

2372a-B

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

36

Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180109 2372-B HBB1 14 4 03577 25 dd20180109 2372-B HBB1 100 05 R5 14 3 06314 16 1765 0051

dd20180104a 2372-B HBB1 14 4 03577 05

dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032 dd20180104a 2372-B HBB1 94 05 R5 14 3 06459 25 1806 0046 dd20180104a 2372-B HBB1 90 05 R5 14 2 06591 08 1842 0017 dd20180104a 2372-B HBB1 88 05 R5 14 3 06589 05 1842 0013 dd20180104a 2372-B HBB1 84 05 R5 14 3 06729 10 1881 0022 dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

We hypothesize that in the older SRM 2372 material the between-strand bond is relatively fragile and can be completely severed at relatively low temperature while avoiding much temperature-dependent AAT loss In contrast the between-strand bond in the SRM 2372a material requires higher temperature to completely disassociate the strands This suggests that there is not a sample-independent ldquooptimumrdquo combination of denaturation temperature and treatment duration However 98 degC for 05 min appears to provide comparable results for these two materials DNA Concentration in the ddPCR Solution Figure 18 summarizes the effect of [DNA] in the ddPCR solution with SRM 2372-B at a denaturation temperature of 98 degC and treatment duration of 05 min Table 22 details the experimental conditions

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

37

Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution

The solid black symbols represent the ratio of the mean result of four technical replicates of the untreated and treated solutions with the λtλu values shown on the left-hand vertical axis the bars represent combined standard uncertainties The solid blue circles represent λu and the open red circles λt with values shown on the right-hand vertical axis at this graphical scale the standard deviations of the technical replicates are covered by the symbol The diagonal blue and red lines represent the log-linear relationships between just the 18 and 14 λ values

The primary effect of [DNA] appears to be the loss of linearity in λt with copiesdroplet values greater than about 1 copiesdroplet There may be some advantage to use somewhat greater dilutions than the routine 14 however the small increase in λtλu may be offset by the increased variability with small λu This suggests the use of dilutions that give λu values of (01 to 04) copiesdroplet

Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180117 2372-B HBB1 12 4 07127 11 dd20180117 2372-B HBB1 98 05 R5 12 4 12331 14 1730 0031

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 3 06151 15 1843 0040

dd20180117 2372-B HBB1 18 4 01549 18 dd20180117 2372-B HBB1 98 05 R5 18 4 02897 20 1870 0051

12 14 18

170

175

180

185

190

02

04

06

08

101214

λ T

arge

t D

ropl

et

λ tλ

u

Dilution Factor

λ₊λᵤλᵤλ₊

λtλu

λu

λt

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

38

Equivalence of Assays Figure 19 summarizes the agreement between the HBB1 NEIF and ND6 assays at a denaturation temperature of 98 degC and treatment duration of 05 min Table 23 details the experimental conditions

Figure 19 Comparison of Assays

The symbols compared the λtλu results from the HBB1 and either the NEIF or ND6 assays Each λtλu value is the ratio of the mean result of four technical replicates of an untreated and three or four replicates of the treated solutions The bars represent combined standard uncertainties The diagonal line represents equality between the assays

There is little or no difference between the λtλu results from the HBB1 NEIF and ND6 assays suggesting that the chromosomal location of the assay target has little influence Since the NEIFHBB1 comparison includes results for both the SRM 2372 and 2372a materials this equivalence is not material-specific

14

16

18

20

14 16 18 20

λ tλ

u NE

IF o

r ND6

λtλu HBB1

NEIFHBB1

ND6HBB1

2nd Assay = HBB1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

39

Table 23 Comparison of Assays

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu)

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045 dd20180105 2372a-B ND6 94 05 R5 14 3 05526 36 1711 0074 dd20180105 2372a-B ND6 90 05 R5 14 3 05326 18 1649 0051 dd20180105 2372a-B ND6 88 05 R5 14 3 05077 22 1572 0052 dd20180105 2372a-B ND6 86 05 R5 14 3 05038 23 1560 0053 dd20180105 2372a-B ND6 84 05 R5 14 3 04906 10 1519 0041

dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

40

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials

Figure 20 summarizes the currently available λtλu results obtained by denaturing at 98 degC for 05 min rapidly cooling in the thermal cycler with dilutions that provide asymp 03 copiesdroplet The figure also displays the consensus result and its 95 confidence region estimated using the Hierarchical Bayes method provided in the NIST Consensus Builder (NICOB) system3839 Table 24 details the experimental conditions

Figure 20 Evaluation of Limiting Ratio

The symbols represent the ratio of the mean result of three or four technical replicates of the untreated and treated solutions the bars represent combined standard uncertainties The thick horizontal line represents a consensus value covering all displayed results the dotted horizontal lines bracket the 95 confidence interval about the consensus value

The Hierarchical Bayes consensus result for these values is 183 with a 95 confidence interval of plusmn003 All but three of the 16 λtλu values are within one standard uncertainty of this consensus value with most of the λtλu means lying within the 95 confidence interval This suggests that there are no statistically significant differences between the SRM 2372 and 2372a materials with regard to ssDNA proportion

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

17

18

19

20

λ tλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

41

Table 24 Evaluation of Limiting Ratio

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180104a 2372-B HBB1 14 4 03577 05 dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032

dd20180108 2372-B HBB1 14 3 03463 09 dd20180108 2372-B HBB1 98 05 R5 14 3 06562 14 1895 0031

dd20180109 2372-B HBB1 14 3 03577 25 dd20180109 2372-B HBB1 98 05 R5 14 3 06495 21 1816 0059

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 4 06151 15 1843 0040

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

42

The agreement among the measured λtλu is more directly visualized using estimated unilateral ldquoDegrees of Equivalencerdquo d Each d is the absolute difference between the measured and consensus values

119889119889 = |120582120582119905119905 120582120582119906119906frasl minus consensus|

and is associated with a 95 confidence interval U95(d) that combines the uncertainty estimates of the measured and consensus values The U95(d) calculation details are specific to the consensus estimation method39 Values of d plusmn U95(d) that include 0 indicate that the λtλu result does not differ from the majority at about a 95 level of confidence Figure 21 displays the d for the λtλu results for the Hierarchical Bayes consensus value39 By this figure of merit all the λtλu can be considered as members of a single population

Figure 21 Degrees of Equivalence

The symbols represent the Degree of Equivalence d for the λtλu results relative to the Hierarchical Bayes consensus value The bars represent 95 level of confidence uncertainty intervals d plusmn U95(d) The horizontal line marks the d = 0 ldquono differencerdquo value Results with d plusmn U95(d) that include this line are not significantly different from the majority

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

-024

-018

-012

-006

000

006

012

018

024

Degr

ees o

f Equ

ival

ence

λtλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

43

Conclusions and Recommendations With high confidence all tubes of the SRM 2372a solution have the same nDNA copy number content within measurement repeatability Additionally within measurement repeatability all tubes of the SRM 2372a have the same mtDNAnDNA With high confidence the nDNA copy number content of the SRM 2372a solution is thermally stable from 4 degC to 37 degC over an extended period The solution should not be shipped or stored below 4 degC With high confidence the ddPCR measurements used to establish the entity and mass concentrations and the mtDNAnDNA ratios are adequately precise and unbiased With high confidence the DNA in the SRM 2372a solutions is now dsDNA and can be expected to remain dsDNA for at least ten years when stored at 4 degC Recommended Certification Values

Since the component materials are believed to be homogeneous are of very similar composition and were prepared and processed similarly the observed differences in the measured entity concentration uncertainties are likely artifacts of the measurement process To limit over-interpretation of these differences we recommend that the copy per nanoliter values of the three components be assigned same 95 relative uncertainty interval (CV95) based on the largest of the measured CV95 values (93 ) rounded up to 10 Since the ngμL transformation does not appreciably inflate this relative uncertainty we recommend that the mass concentration values also be assigned CV95 of 10 Similarly we recommend that the mtDNAnDNA ratios be rounded to integer values all three components be assigned the same CV95 and that this uncertainty be based on the largest (21 ) of the measured CV95 rounded up to 25 Table 25 lists the values that we recommend be used in the Certificate of Analysis

Table 25 Recommended Values and 95 Uncertainties for Certification

Component Units Value U95(Value) a A Copies per nanoliter 151 15 B Copies per nanoliter 175 18 C Copies per nanoliter 145 15 A ngμL 498 50

B ngμL 578 58 C ngμL 479 48 A mtDNAnDNA 174 4

B mtDNAnDNA 206 5 C mtDNAnDNA 279 7

a The interval Value plusmn U95(Value) is believed with about 95 confidence to contain the true value of the measurand

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

44

Use of SRM 2372a for Value-Assigning In-House Reference Materials

The within-component and between-assay variability observed for the commercial qPCR assays is not unexpected since different qPCR methods exploit qualitatively different molecular targets Given the intrinsic variability of the human genome the prospect of developing a human DNA source material that responds equally to all current and future qPCR assays is improbable All three components should be used to elucidate potential bias when using these materials to value assign in-house DNA reference and control materials for qPCR assays

Certificate of Analysis In accordance with ISO Guide 31 2000 a NIST SRM certificate is a document containing the name description and intended purpose of the material the logo of the US Department of Commerce the name of NIST as a certifying body instructions for proper use and storage of the material certified property value(s) with associated uncertainty(ies) method(s) used to obtain property values the period of validity if appropriate and any other technical information deemed necessary for its proper use A Certificate is issued for an SRM certified for one or more specific physical or engineering performance properties and may contain NIST reference information or both values in addition to certified values A Certificate of Analysis is issued for an SRM certified for one or more specific chemical properties Note ISO Guide 31 is updated periodically check with ISO for the latest version [httpswwwnistgovsrmsrm-definitions] For the most current version of the Certificate of Analysis for NIST SRM 2372a Human DNA Quantification Standard please visit httpswww-snistgovsrmorsview_detailcfmsrm=2372a

References

1 National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research Department of Health Education and Welfare (1978) Belmont Report Ethical Principles and Guidelines for the Protection of Human Subjects of Research DHEW Publication No (OS) 78-0012 httpvideocastnihgovpdfohrp_belmont_reportpdf

2 CDCNIH Biosafety in Microbiological and Biomedical Laboratories 5th ed HHS publication No (CDC) 21-1112 Chosewood LC Wilson DE Eds US Government Printing Office Washington DC (2009) httpwwwcdcgovbiosafetypublicationsbmbl5indexhtm

3 Duewer DL Kline MC Redman JW Newall PJ Reeder DJ NIST Mixed Stain Studies 1 and 2 Interlaboratory Comparison of DNA Quantification Practice and Short Tandem Repeat Multiplex Performance with Multiple-Source Samples J Forensic Sci 200146(5)1199-1210 PMID 11569565

4 Kline MC Duewer DL Redman JW Butler JM NIST Mixed Stain Study 3 DNA quantitation accuracy and its influence on short tandem repeat multiplex signal intensity Anal Chem 200375(10)2463-2469 httpsdoiorg101021ac026410i

5 Kline MC Duewer DL Redman JW Butler JM Results from the NIST 2004 DNA Quantitation Study J Forensic Sci 200550(3)571-578 PMID 15932088

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

45

6 Kline MC Duewer DL Travis JC Smith MV Redman JW Vallone PM Decker AE Butler JM Production and Certification of Standard Reference Material 2372 Human DNA Quantitation Standard Anal Bioanal Chem 2009394(4)1183-1192 httpsdoiorg101007s00216-009-2782-0

7 Sambrook J and Russell DW (2001) Molecular Cloning a Laboratory Manual Cold Spring Harbor Laboratory Press Cold Spring Harbor New York

8 EU-JRC Protocol NK603 ndash Community Reference Laboratory Event-specific method for the quantification of maize line NK603 using real-time PCR Section 33 Spectrophotometric Measurement of DNA Concentration httpgmo-crljrceceuropaeusummariesNK603-WEB-ProtocolValidationpdf

9 ISO 215712005(E) Foodstuffs ndash Methods of analysis for the detection of genetically modified organisms and derived products ndash Nucleic acid extraction Annex B Methods for quantitation of extracted DNA pp 34-36 International Organization for Standardization Geneva (2005)

10 Certificate of Analysis Standard Reference Material 2372 Human DNA Quantitation Standard (2013) Gaithersburg MD USA Standard Reference Materials Program National Institute of Standards and Technology httpswww-snistgovsrmorsview_certcfmsrm=2372

11 Vogelstein B Kinzler KW Digital PCR Proc Natl Acad Sci USA 1999969236-9241 12 Hindson BJ Ness KD Masquelier DA Belgrader P Heredia NJ Makarewicz AJ Bright IJ Lucero

MY Hiddessen AL Legler TC Kitano TK Hodel MR Petersen JF Wyatt PW Steenblock ER Shah PH Bousse LJ Troup CB Mellen JC Wittmann DK Erndt NG Cauley TH Koehler RT So AP Dube S Rose KA Montesclaros L Wang S Stumbo DP Hodges SP Romine S Milanovich FP White HE Regan JF Karlin-Neumann GA Hindson CM Saxonov S Colston BW High-Throughput Droplet Digital PCR System for Absolute Quantitation of DNA Copy Number Anal Chem 2011838604-8610 httpsdoiorg101021ac202028g

13 Pinheiro LB Coleman VA Hindson CM Herrmann J Hindson BJ Bhat S Emslie KR Evaluation of a Droplet Digital Polymerase Chain Reaction Format for DNA Copy Number Quantification AnalChem 2012841003-1011 httpsdoiorg101021ac202578x

14 Corbisier P Pinheiro L Mazoua S Kortekaas AM Chung PY Gerganova T Roebben G Emons H Emslie K DNA copy number concentration measured by digital and droplet digital quantitative PCR using certified reference materials Anal Bioanal Chem 2015407(7)1831-1840 httpsdoiorg101007s00216-015-8458-z

15 Duewer DL Kline MC Romsos EL Real-time cdPCR opens a window into events occurring in the first few PCR amplification cycles Anal Bioanal Chem 2015407(30)9061-9069 httpsdoiorg101007s00216-015-9073-8

16 Kline MC Romsos EL Duewer DL Evaluating Digital PCR for the Quantification of Human Genomic DNA Accessible Amplifiable Targets Anal Chem 201688(4)2132-2139 httpsdoiorg101021acsanalchem5b03692

17 Kline MC Duewer DL Evaluating Droplet Digital Polymerase Chain Reaction for the Quantification of Human Genomic DNA Lifting the Traceability Fog Anal Chem 201789(8)4648-4654 httpsdoiorg101021acsanalchem7b00240

18 Duewer DL Kline MC Romsos EL Toman B Evaluating Droplet Digital PCR for the quantification of human genomic DNA Conversion from Copies per Nanoliter to Nanogram Genomic DNA per Microliter Anal Bioanal Chem 2018in press httpsdoiorg101007s00216-018-0982-1

19 Dagata JA Farkas N Kramer JA Method for measuring the volume of nominally 100 μm diameter spherical water in oil emulsion droplets NIST Special Publication 260-184 2016 httpsdoiorg106028NISTSP260-184

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

20 ISO 170342016 General requirements for the competence of reference material producers International Organization for Standardization Geneva (2016)

21 NCBI Standard Nucleotide BLAST httpsblastncbinlmnihgovBlastcgiPAGE_TYPE=BlastSearch

22 Timken MD Swango KL Orrego C Buoncristiani MR A Duplex Real-Time qPCR Assay for the Quantification of Human Nuclear and Mitochondrial DNA in Forensic Samples Implications for Quantifying DNA in Degraded Samples J Forensic Sci 200550(5)1044-60 PMID 16225209

23 Walker JA Hedges DJ Perodeau BP Landry KE Stoilova N Laborde ME Shewale J Sinha SK Batzer MA Multiplex polymerase chain reaction for simultaneous quantitation of human nuclear mitochondrial and male Y-chromosome DNA application in human identification Anal Biochem 200533789ndash97 httpsdoiorg101016jab200409036

24 Kavlick MF Lawrence HS Merritt RT Fisher C Isenberg A Robertson JM Budowle B Quantification of Human Mitochondrial DNA Using Synthesized DNA Standards J Forensic Sci 201156(6) 1457ndash1463 httpsdoiorg101111j1556-4029201101871x

25 Stulnig TM Amberger A Exposing contaminating phenol in nucleic acid preparations BioTechniques 199416(3) 403-404

26 Majumdar N Wessel T Marks J Digital PCR Modeling for Maximal Sensitivity Dynamic Range and Measurement Precision PLoS ONE 201510(3)e0118833 httpsdoi101371journalpone0118833

27 Tellinghuisen J Partition Volume Variability in Digital Polymerase Chain Reaction Methods Polydispersity Causes Bias but Can Improve Precision Anal Chem 2016881283-12187 httpsdoiorg101021acsanalchem6b03139

28 Lunn DJ Spiegelhalter D Thomas A Best N (2009) Statistics in Medicine 283049ndash3082 29 De Biegravevre P Dybkaer R Fajgelj A Hibbert DB Metrological traceability of measurement results in

chemistry Concepts and implementation Pure Appl Chem 201183(10)1873-1935 httpsdoiorg101007s00769-011-0805-y

30 Huggett JF Foy CA Benes V Emslie K Garson JA Haynes R Hellemans J Kubista M Mueller RD Nolan T Pfaffl MW Shipley GL Vandesompele J Wittwer CT Bustin SA The Digital MIQE Guidelines Minimum Information for Publication of Quantitative Digital PCR Experiments Clin Chem 201359(6)892-902 httpsdoiorg101373clinchem2013206375

31 Wilson PJ Ellison SLR Extending digital PCR analysis by modelling quantification cycle data BMC Bioinfo 201617421 httpsdoiorg101186s12859-016-1275-3

32 Wang X Son A Effects of pretreatment on the denaturation and fragmentation of genomic DNA for DNA hybridization Environ Sci Processes Impacts 2013152204 httpsdoiorg101039c3em00457k

33 Nwokeoji AO Kilby PM Portwood DE Dickman MJ Accurate Quantification of Nucleic Acids Using Hypochromicity Measurements in Conjunction with UV Spectrophotometry Anal Chem 201789(24)13567-13574 https101021acsanalchem7b04000

34 Farkas DH Kaul KL Wiedbrauk DL Kiechle FL Specimen collection and storage for diagnostic molecular pathology investigation Arch Pathol Lab Med 1996120(6)591-6

35 Kypr J Kejnovskaacute I Renčiuk D Vorličkovaacute M Circular dichroism and conformational polymorphism of DNA Nucleic Acids Res 200937(6)1713-1725 httpsdoiorg101093nargkp026

36 Strachan T Read AP (1999) Human Molecular Genetics 2nd Ed John Wiley amp Sons NY

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

37 Bhat S McLaughlin JL Emslie KR Effect of sustained elevated temperature prior to amplification on template copy number estimation using digital polymerase chain reaction Analyst 2011136724ndash32 httpsdoiorg101039c0an00484g

38 NIST Consensus Builder (NICOB) httpsconsensusnistgov 39 Koepke A Lafarge T Toman B Possolo A NIST Consensus Builder Userrsquos Manual

httpsconsensusnistgovNISTConsensusBuilder-UserManualpdf

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

Appendix A DNA Extraction Method The following method has been adapted from Miller et al (1988)A1 Procedure

1 Transfer approximately 5 mL of buffy coat into a 50 mL sterile polypropylene centrifuge tube and add 30 mL Red Blood Cell Lysis Buffer

2 Mix by gentle inversion and let stand at room temperature for 15 min

3 Centrifuge at approximately 209 rads (2000 rpm) for 20 min

4 Discard supernatant

5 Add 9 mL of Nucleic Lysis Buffer and re-suspend the pelleted cells by agitating with a plastic transfer pipette

6 Add 300 microL of 20 sodium dodecyl sulfate (SDS) and 18 mL of Protease K Solution (freshly prepared 2 mg to 5 mg Protease K per 1 mL Protease K Buffer)

7 Mix by inversion followed by incubation overnight or longer at 37 ordmC to 40 ordmC in a shaking water bath The length of incubation is determined by the time required to solubilize the cells present in the tube

8 Add 4 mL saturated ammonium acetate solution (in deionized water stored at 2 ordmC to 8 ordmC) Shake vigorously for 15 s

9 Centrifuge at approximately 471 rads (4500 rpm) for 10 min

10 Pipette supernatant into a 50 mL sterile polypropylene centrifuge tube and add two volumes of room temperature absolute ethanol Mix by gentle inversion until the DNA aggregates and floats to the top of the tube

11 Spool the aggregated DNA from the current tube into a fresh tube of 80 ethanol using a sterile inoculation loop At this step multiple tubes of precipitated DNA may be pooled to make a large lot of the purified material Mix by gentle inversion

12 Transfer the aggregated DNA to a fresh dry 50 mL polypropylene centrifuge tube Remove as much ethanol as possible by squeezing the DNA against the wall of the tube with a sterile inoculation loop Air-dry the material in a laminar flow hood

13 Resolubilize DNA in TE-4 buffer and store material at 4 ordmC Reagents

Red Blood Cell Lysis Buffer 144 mmolL ammonium chloride 1 mmolL sodium bicarbonate in deionized (DI) water Store at 2 ordmC to 8 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

Nucleic Lysis Buffer 10 mmolL Tris-HCl 400 mmolL sodium chloride 20 mmolL disodium ethylenediaminetetraacetic acid in DI Water adjust pH to 82 Store at 2 ordmC to 8 ordmC

Protease K Buffer

20 mmolL disodium ethylenediaminetetraacetic acid 10 (mass per volume) Sodium Dodecyl Sulfate (SDS) in DI water Store at 2 ordmC to 8 ordmC

Protease K

Sigma-Aldrich cat P6556 Tritirachium album lyophilized powder Store at ndash20 ordmC

Ammonium Acetate

192 molL ammonium acetate (saturated) in DI water Store at 2 ordmC to 8 ordmC

Absolute Ethanol (200 proof)

Store at room temperature Reference

A1 Miller SA Dykes DD Polesky HF A simple salting out procedure for extracting DNA from human nucleated cells Nucleic Acids Res 198816(3)1215 httpsdoiorg101093nar1631215

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

48

Appendix B Spectrophotometric Measurement Protocol The following is for use with a scanning recording spectrophotometer The procedures include the spectrophotometer calibration measurement of the ldquonativerdquo dsDNA and measurement of the NaOH denatured ssDNA The procedure for evaluating solutions after treatment with NaOH is adapted from references89 Spectrophotometer Calibration

Establish Baseline Set the spectrophotometer to acquire data from 220 nm to 345 nm in steps of 01 nm Zero the instrument with nothing (air) in the sample and reference cells Acquire an air-air baseline of the instrument with nothing (air) in the sample and reference cells If the trace has significant structure or excessive noise take corrective action If the trace is acceptable proceed with data collection Enable Verification of Wavelength Calibration SRM 2034 is a solution of holmium oxide in 10 perchloric acid contained in a sealed quartz cuvette delivering an intrinsic UVVis wavelength standard for absorption spectrophotometryB1 Evaluation of the location of peak maxima in the spectra of this solution enables verification of the spectrophotometers wavelength calibration Place SRM 2034 in the sample cell Create and name a file to hold the verification spectra Perform three replicate SRM 2034-to-air scans saving each scan to the file Remove SRM 2034 from the spectrophotometer and store appropriately Enable Verification of Absorbance Calibration SRM 2031 consists of three metal-on-fused silica filters held in cuvette-sized filter holders plus an empty filter holder The three filters are certified for transmittance and transmittance density (-log10(transmittance) effectively absorbance) at several wavelengths with nominal wavelength-independent transmittances of 10 30 and 90 Evaluation of the absorbance of each of these three filters at the certified 250 nm 280 nm and 340 nm wavelengths enables verification of the spectrophotometers absorbance calibration Place the empty SRM 2031 filter holder in the reference cell Remove the protective covers from the 10 transmittance filter and place in the sample cell Perform three replicate scans saving each scan to file Remove the filter from the spectrophotometer replace the protective covers and store appropriately Repeat for the 30 and 90 filters Close the file containing the verification spectra For ldquonativerdquo dsDNA Samples

Verify Optical Balance of Cuvettes The measurements of the DNA-containing samples are made using a 100 microL quartz cuvette in the sample cell compared to an optically matched cuvette in the reference cell The optical equivalence of these two cuvettes can be verified by filling both cuvettes with appropriate volumes of a sample solution matrix (TE-4 buffer) that does not contain DNA Fill two dry

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

49

clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes followed by a methanol rinse in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquonativerdquo dsDNA remove and clean both the reference cuvette and the sample cuvette For NaOH denatured ssDNA Samples

Prepare a 04 molL NaOH Stock Solution To a 15 mL disposable centrifuge tube add 016 g NaOH pellets add deionized (DI) water to the 10 mL line and mix until all the NaOH pellets have dissolved Centrifuge at 524 rads (5000 rpm) for 5 min For every 9 samples to be analyzed aliquot 1 mL of this 04 molL NaOH stock solution into a 15 mL screw top tube Use these tubes to prepare the NaOH blanks and DNA samples In a 15 mL screw cap vial mix 200 microL TE-4 buffer with 200 microL 04 molL sodium hydroxide solution to achieve a final NaOH concentration of 02 molL Vortex 5 s then centrifuge at 1257 rads (12 000 rpm) for 1 min Fill two dry clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Prepare the Sample Solutions For each unit of material to be measured (Tube or vial)

a) Vortex the unit for 5 s b) Centrifuge at 1257 rads (12 000 rpm) for 1 min c) Remove a 100 microL aliquot and place in a new labeled vial d) Add 100 microL of the freshly prepared 04 molL NaOH solution (with the same pipet) e) Vortex 5 s f) Centrifuge at 1257 rads (12 000 rpm) for 1 min

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

50

Measure Sample Solutions Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes rinse with methanol in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the NaOH denatured DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquodenaturedrdquo ssDNA close the file holding the spectra for this series Remove and clean both the reference cuvette and the sample cuvette After All Sample Measurements are Complete

Enable Verification of Spectrophotometer Stability The stability of the spectrophotometer over the course of the data collection can be evaluated by comparing spectra for SRM 2034 and 2031 taken at the start of data collection with spectra taken at the end of data collection Repeat the wavelength and absorbance verification measurements described above collecting the spectra in a suitably named file Clean Up Transfer all relevant data files from the spectrophotometers storage to a transfer device Clean the spectrophotometer work area as necessary and restore the SRMs and cuvettes to their appropriate storage locations Evaluation

Export the files to a spreadsheet Average the replicate spectra Subtract the average absorbance at A320 from the absorbance at A260 resulting in the corrected absorbance at A260 The conventional spectrophotometric estimate of the mass concentration of the dsDNA in the sample is achieved by multiplying the corrected absorbance by 50 The conventional estimate for the NaOH-treated ssDNA is achieved by multiplying the corrected absorbance by 37times2 where the ldquotimes2rdquo adjusts for the 1rarr2 volumetric dilution of the sample Note reliable spectrophotometric measurements require corrected A260 to be greater than 005 Reference

B1 Travis JC Acosta JC Andor G Bastie J Blattner P Chunnilall CJ Crosson SC Duewer DL Early EA Hengstberger F Kim C-S Liedquist L Manoocheri F Mercader F Metzdorf J Mito A Monardh LAG Nevask S Nilsson M Noeumll M Rodriguez AC Ruiacutez A Schirmacher A Smith MV Valencia G van Tonder N Zwinkels J Intrinsic Wavelength Standard Absorption Bands in Holmium Oxide Solution for UVvisible Molecular Absorption Spectrophotometry J Phys Chem Ref Data 200534(1)41-56

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

51

Appendix C SRM 2372a Statisticianrsquos report Extracted from Blaza Toman

November 30 2017 1 Introduction Candidate material for SRM 2372a is composed of three components labeled A B and C All components are human genomic DNA and prepared at NIST from buffy coat white blood cells from single-source anonymous donors Component A and B are single source (malefemale respectively) and Component C is a 1 male to 3 female mixture of two single-source donors Droplet digital PCR (ddPCR) assays were used to obtain measurements of the concentration of copies per microliter of sample Ten NIST-developed human nuclear DNA and three literature-derived human mitochondrial DNA PCR assays were used in this study

11 OpenBUGS Analysis All parameters in the following sections were estimated using Markov Chain Monte Carlo (MCMC) implemented in the OpenBUGS freewareC1 OpenBUGS is a software application for the Bayesian analysis of complex statistical models

12 OpenBUGS Exemplar Output All following sections list the OpenBUGs model (ldquocoderdquo) initialization values (ldquoinitsrdquo) if they are needed and the data used to evaluate the various data sets discussed in the body of this report homogeneity stability nDNA certification and mtDNAnDNA value assignment The provided information in each of the following sections is sufficient to produce estimates but the estimates may not exactly reproduce the exemplar results provided The MCMC paradigm does not produce ldquoanalyticalrdquo values but approaches them asymptotically as the number o model updates increases All exemplar results presented here are based upon the following bull 10-fold ldquothinningrdquo to minimize autocorrelation there were 10 MCMC estimates for every

model update used in parameter estimation bull a ldquoburn inrdquo of 5 000 updates before beginning parameter estimation and bull 10 000 model updates

The following statistics are provided in the tables listing the exemplar results

mean arithmetic mean of the parameterrsquos model update values sd standard deviation of the parameterrsquos model update values MC_error Monte Carlo standard deviation of the mean meanradic(number of model updates) val25pc 25th percentile of the model update values median 50th percentile of the model update values val975pc 975th percentile of the model update values startnumber of ldquoburn inrdquo updates sample number of model update values

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

52

2 Homogeneity assessment 21 Data

Samples of the material of all three components were drawn from 22 different boxes Two assays (NEIF ND6) were used to obtain 3 replicates of the measurement in copies per nanoliter for each sample and component For the complete homogeneity data see

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

22 Model The data were analyzed using a three-level Gaussian hierarchical ANOVA model

119910119910119894119894119894119894119894119894~119873119873120572120572119894119894119894119894119894119894 120590120590119908119908119894119894119905119905ℎ119894119894119894119894 119887119887119887119887119887119887 1198941198941198941198942

where ~N() denotes ldquodistributed as a normal distribution of some mean μ and variancerdquo 119894119894 denotes component (1 for A 2 for B 3 for C) 119895119895 denotes assay (1 for NEIF 2 for ND6) 119896119896 denotes box (1hellip22) and 120572120572119894119894119894119894119894119894~119873119873micro119894119894119894119894120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119887119887119887119887119887119887119894119894

2

All variance components were given a Half Cauchy prior as described in Reference C2 Note The Half Cauchy prior is specified by ldquo~dnorm(0 00016)I(0001)rdquo The ldquoIrdquo notation is for censoring It cuts off any random draws from the Normal distribution that are lower than 0001 An informative prior is needed when there are less than four data per distribution and the half Cauchy is a good general-purpose prior

23 OpenBUGS code and data Homogeneity code ModelBegin for(i in 13)xiNsb[i] ~ dnorm(0 00016)I(0001) chSqNsb[i] ~ dgamma(0505) sigb[i] lt- xiNsb[i]sqrt(chSqNsb[i]) for(i in 13)for(j in 12)mu[ij]~dnorm(010E-5) for(i in 13) muA[i]lt-mean(mu[i])sigbP[i]lt-100(muA[i]sqrt(sigb[i])) for(i in 13)for(j in 12)xiNs[ij] ~ dnorm(0 00016)I(0001) chSqNs[ij]~dgamma(0505) sigt[ij]lt-xiNs[ij]sqrt(chSqNs[ij]) sigtP[ij]lt-100(mu[ij]sqrt(sigt[ij])) for(i in 13)for(j in 12)for(k in 122)alpha[ijk]~dnorm(mu[ij]sigb[i]) for(idx in 1393)lamb[idx]~dnorm(alpha[cmp[idx]asy[idx]box[idx]]sigt[cmp[idx]asy[idx]]) ModelEnd Inits None required

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

53

Homogeneity Data lamb[] asy[] box[] cmp[] rep[] 1632 1 1 1 1 1529 1 1 1 2 1511 1 1 1 3 1616 2 1 1 1 1640 2 1 1 2 1582 2 1 1 3 1531 1 2 1 1 1525 1 2 1 2 1549 1 2 1 3 1615 2 2 1 1 1589 2 2 1 2 1601 2 2 1 3 1524 1 3 1 1 1540 1 3 1 2 1533 1 3 1 3 1582 2 3 1 1 1152 2 3 1 2 1555 2 3 1 3 1454 1 4 1 1 1458 1 4 1 2 1427 1 4 1 3 1627 2 4 1 1 1605 2 4 1 2 1607 2 4 1 3 1465 1 5 1 1 1540 1 5 1 2 1504 1 5 1 3 1607 2 5 1 1 1633 2 5 1 2 1586 2 5 1 3 1512 1 6 1 1 1491 1 6 1 2 1491 1 6 1 3 1511 2 6 1 1 1483 2 6 1 2 1524 2 6 1 3 1418 1 7 1 1 1480 1 7 1 2 1455 1 7 1 3 1479 2 7 1 1 1557 2 7 1 2 1656 2 7 1 3 1496 1 8 1 1 1560 1 8 1 2 1501 1 8 1 3 1514 2 8 1 1 1615 2 8 1 2 1540 2 8 1 3 1588 1 9 1 1 1581 1 9 1 2 1637 1 9 1 3 1474 2 9 1 1 1438 2 9 1 2 1454 2 9 1 3 1574 1 10 1 1 1513 1 10 1 2 1570 1 10 1 3 1594 2 10 1 1 1713 2 10 1 2 1669 2 10 1 3

1557 1 11 1 1 1508 1 11 1 2 1633 1 11 1 3 1571 2 11 1 1 1611 2 11 1 2 1631 2 11 1 3 1500 1 12 1 1 1476 1 12 1 2 1515 1 12 1 3 1626 2 12 1 1 1586 2 12 1 2 1608 2 12 1 3 1672 1 13 1 1 1634 1 13 1 2 1804 1 13 1 3 1490 2 13 1 1 1516 2 13 1 2 1497 2 13 1 3 1551 1 14 1 1 1568 1 14 1 2 1538 1 14 1 3 1617 2 14 1 1 1614 2 14 1 2 1639 2 14 1 3 1456 1 15 1 1 1448 1 15 1 2 1523 1 15 1 3 1525 2 15 1 1 1402 2 15 1 2 1436 2 15 1 3 1573 1 16 1 1 1520 1 16 1 2 1617 1 16 1 3 1634 2 16 1 1 1627 2 16 1 2 1575 2 16 1 3 1516 1 17 1 1 1516 1 17 1 2 1628 1 17 1 3 1613 2 17 1 1 1567 2 17 1 2 1567 2 17 1 3 1612 1 18 1 1 1535 1 18 1 2 1672 1 18 1 3 1529 2 18 1 1 1495 2 18 1 2 1528 2 18 1 3 1598 1 19 1 1 1514 1 19 1 2 1596 1 19 1 3 1571 2 19 1 1 1535 2 19 1 2 1511 2 19 1 3 1575 1 20 1 1 1568 1 20 1 2 1586 1 20 1 3 1652 2 20 1 1 1545 2 20 1 2 1627 2 20 1 3 1658 1 21 1 1

1611 1 21 1 2 1578 1 21 1 3 1647 2 21 1 1 1607 2 21 1 2 1612 2 21 1 3 1547 1 22 1 1 1528 1 22 1 2 1532 1 22 1 3 1537 2 22 1 1 1513 2 22 1 2 1536 2 22 1 3 1790 1 1 2 1 1909 1 1 2 2 1885 1 1 2 3 1870 2 1 2 2 1860 2 1 2 3 1712 1 2 2 1 1719 1 2 2 2 1704 1 2 2 3 1775 2 2 2 1 1718 2 2 2 2 1756 2 2 2 3 1693 1 3 2 1 1755 1 3 2 2 1749 1 3 2 3 1824 2 3 2 1 1483 2 3 2 2 1896 2 3 2 3 1754 1 4 2 1 1745 1 4 2 2 1739 1 4 2 3 1670 2 4 2 1 1632 2 4 2 2 1814 2 4 2 3 1757 1 5 2 1 1687 1 5 2 2 1735 1 5 2 3 1683 2 5 2 1 1759 2 5 2 2 1703 2 5 2 3 1697 1 6 2 1 1709 1 6 2 2 1714 1 6 2 3 1693 2 6 2 1 1723 2 6 2 2 1722 2 6 2 3 1722 1 7 2 1 1625 1 7 2 2 1703 1 7 2 3 1690 2 7 2 1 1684 2 7 2 2 1688 2 7 2 3 1884 1 8 2 1 1802 1 8 2 2 1772 1 8 2 3 1864 2 8 2 1 1887 2 8 2 2 1821 2 8 2 3 1713 1 9 2 1 1734 1 9 2 2 1727 1 9 2 3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

54

1709 2 9 2 1 1756 2 9 2 2 1758 2 9 2 3 1759 1 10 2 1 1809 1 10 2 2 1738 1 10 2 3 1764 2 10 2 1 1816 2 10 2 2 1787 2 10 2 3 1749 1 11 2 1 1782 1 11 2 2 1740 1 11 2 3 1741 2 11 2 1 1792 2 11 2 2 1784 2 11 2 3 1748 1 12 2 1 1712 1 12 2 2 1718 1 12 2 3 1797 2 12 2 1 1614 2 12 2 2 1790 2 12 2 3 1632 1 13 2 1 1668 1 13 2 2 1701 1 13 2 3 1626 2 13 2 1 1566 2 13 2 2 1681 2 13 2 3 1752 1 14 2 1 1535 1 14 2 2 1673 1 14 2 3 1724 2 14 2 1 1741 2 14 2 2 1685 2 14 2 3 1871 1 15 2 1 1835 1 15 2 2 1887 1 15 2 3 1770 2 15 2 1 1752 2 15 2 2 1798 2 15 2 3 1688 1 16 2 1 1725 1 16 2 2 1743 1 16 2 3 1744 2 16 2 1 1697 2 16 2 2 1710 2 16 2 3 1789 1 17 2 1 1769 1 17 2 2 1825 1 17 2 3 1716 2 17 2 1 1762 2 17 2 2 1638 2 17 2 3 1826 1 18 2 1 1780 1 18 2 2 1849 1 18 2 3 1805 2 18 2 1 1801 2 18 2 2 1931 2 18 2 3 1751 1 19 2 1 1759 1 19 2 2 1880 1 19 2 3 1827 2 19 2 1 1766 2 19 2 2

1786 2 19 2 3 1738 1 20 2 1 1678 1 20 2 2 1768 1 20 2 3 1752 2 20 2 1 1747 2 20 2 2 1768 2 20 2 3 1742 1 21 2 1 1836 1 21 2 2 1713 1 21 2 3 1746 2 21 2 1 1744 2 21 2 2 1764 2 21 2 3 1761 1 22 2 1 1743 1 22 2 2 1660 1 22 2 3 1666 2 22 2 1 1692 2 22 2 2 1819 2 22 2 3 1493 1 1 3 1 1385 1 1 3 2 1510 1 1 3 3 1565 2 1 3 1 1562 2 1 3 2 1466 2 1 3 3 1516 1 2 3 1 1513 1 2 3 2 1577 1 2 3 3 1490 2 2 3 1 1547 2 2 3 2 1570 2 2 3 3 1451 1 3 3 1 1528 1 3 3 2 1506 1 3 3 3 1485 2 3 3 1 1505 2 3 3 2 1471 2 3 3 3 1440 1 4 3 1 1486 1 4 3 2 1520 1 4 3 3 1529 2 4 3 1 1508 2 4 3 2 1506 2 4 3 3 1523 1 5 3 1 1472 1 5 3 2 1483 1 5 3 3 1513 2 5 3 1 1520 2 5 3 2 1491 2 5 3 3 1437 1 6 3 1 1493 1 6 3 2 1522 1 6 3 3 1492 2 6 3 1 1481 2 6 3 2 1461 2 6 3 3 1485 1 7 3 1 1477 1 7 3 2 1463 1 7 3 3 1475 2 7 3 1 1547 2 7 3 2 1485 2 7 3 3 1433 1 8 3 1

1405 1 8 3 2 1478 1 8 3 3 1497 2 8 3 1 1501 2 8 3 2 1495 2 8 3 3 1463 1 9 3 1 1499 1 9 3 2 1505 1 9 3 3 1503 2 9 3 1 1491 2 9 3 2 1516 2 9 3 3 1534 1 10 3 1 1458 1 10 3 2 1454 1 10 3 3 1448 2 10 3 1 1465 2 10 3 2 1474 2 10 3 3 1533 1 11 3 1 1514 1 11 3 2 1555 1 11 3 3 1488 2 11 3 1 1470 2 11 3 2 1483 2 11 3 3 1450 1 12 3 1 1525 1 12 3 2 1497 1 12 3 3 1496 2 12 3 1 1483 2 12 3 2 1519 2 12 3 3 1537 1 13 3 1 1527 1 13 3 2 1515 1 13 3 3 1499 2 13 3 1 1479 2 13 3 2 1475 2 13 3 3 1515 1 14 3 1 1490 1 14 3 2 1494 1 14 3 3 1469 2 14 3 1 1480 2 14 3 2 1466 2 14 3 3 1511 1 15 3 1 1568 1 15 3 2 1594 1 15 3 3 1515 2 15 3 1 1543 2 15 3 2 1565 2 15 3 3 1449 1 16 3 1 1452 1 16 3 2 1471 1 16 3 3 1543 2 16 3 1 1448 2 16 3 2 1491 2 16 3 3 1388 1 17 3 1 1421 1 17 3 2 1394 1 17 3 3 1452 2 17 3 1 1435 2 17 3 3 1475 1 18 3 1 1455 1 18 3 2 1445 1 18 3 3 1458 2 18 3 1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

55

1531 2 18 3 2 1459 2 18 3 3 1468 1 19 3 1 1525 1 19 3 2 1446 1 19 3 3 1435 2 19 3 1 1478 2 19 3 2 1421 2 19 3 3 1549 1 20 3 1

1531 1 20 3 2 1499 1 20 3 3 1499 2 20 3 1 1465 2 20 3 2 1468 2 20 3 3 1530 1 21 3 1 1440 1 21 3 2 1484 1 21 3 3 1499 2 21 3 1

1454 2 21 3 2 1442 1 22 3 1 1487 1 22 3 2 1436 1 22 3 3 1589 2 22 3 1 1591 2 22 3 2 1556 2 22 3 3 END

Table C-1 Example OpenBUGS Homogeneity Summary

mean sd MC_error val25pc median val975pc start sample mu[11] 15460 0122 000123 1522 1546 1570 5001 10000 mu[12] 15640 0137 000141 1537 1564 1591 5001 10000 mu[21] 17490 0113 000118 1727 1749 1772 5001 10000 mu[22] 17480 0129 000125 1722 1748 1774 5001 10000 mu[31] 14870 0073 000073 1473 1487 1501 5001 10000 mu[32] 14960 0069 000063 1482 1496 1509 5001 10000 muA[1] 15550 0092 000092 1536 1555 1573 5001 10000 muA[2] 17490 0085 000086 1732 1749 1765 5001 10000 muA[3] 14910 0050 000046 1481 1491 1501 5001 10000 sigbP[1] 3269 04985 000514 2376 3244 4324 5001 10000 sigbP[2] 2548 04385 000461 1733 2533 3472 5001 10000 sigbP[3] 1850 0292 000260 1329 1832 2477 5001 10000

sigtP[11] 2774 0298 000260 2268 2750 3442 5001 10000 sigtP[12] 4036 0429 000438 3290 4005 4986 5001 10000 sigtP[21] 2664 0301 000321 2155 2640 3313 5001 10000 sigtP[22] 3807 0395 000377 3116 3772 4662 5001 10000 sigtP[31] 2313 0250 000209 1880 2291 2873 5001 10000 sigtP[32] 1812 0195 000170 1473 1797 2243 5001 10000

24 Results The material appears to be homogeneous as can be seen in Figure C-1 Based on these results the between-tube variability is of the same order as within-tube variability

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

56

Component A Assay NEIF Component A Assay ND6 box plot gamma[11]

[111][112][113]

[114]

[115][116]

[117]

[118]

[119]

[1110][1111]

[1112]

[1113]

[1114]

[1115]

[1116][1117]

[1118]

[1119][1120]

[1121]

[1122]

114

box plot gamma[12]

[121][122]

[123]

[124][125]

[126]

[127][128]

[129]

[1210]

[1211][1212]

[1213]

[1214]

[1215]

[1216]

[1217]

[1218][1219]

[1220][1221]

[1222]

115

Component B Assay NEIF Component B Assay ND6

box plot gamma[21]

[211]

[212]

[213][214]

[215][216]

[217]

[218]

[219]

[2110][2111]

[2112]

[2113][2114]

[2115]

[2116]

[2117]

[2118]

[2119]

[2120]

[2121]

[2122]

129

box plot gamma[22]

[221]

[222][223]

[224][225][226][227]

[228]

[229]

[2210][2211]

[2212]

[2213]

[2214]

[2215]

[2216][2217]

[2218]

[2219]

[2220][2221][2222]

128

Component C Assay NEIF Component C Assay ND6

box plot gamma[31]

[311]

[312]

[313][314]

[315][316]

[317]

[318]

[319][3110]

[3111]

[3112]

[3113]

[3114]

[3115]

[3116]

[3117]

[3118]

[3119]

[3120]

[3121]

[3122]109

box plot gamma[32]

[321][322]

[323]

[324][325]

[326]

[327][328][329]

[3210]

[3211]

[3212]

[3213][3214]

[3215]

[3216]

[3217]

[3218]

[3219]

[3220][3221]

[3222]

11

Figure C-1 Homogeneity Assessment

Note These results were obtained using the 1-4 gravimetric dilution-adjusted λ but uncorrected for droplet volume

or the 1-10 volumetric dilution That is the vertical axis in each panel should be multiplied by 07349times10 = 7349 This constant difference in scale does not impact the homogeneity analysis Error bars span approximate 95 confidence intervals

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

57

3 Certification measurements 31 Data

Originally the experiment consisted of 5 plates containing the 3 components (A B C) and 12 assays Ten of the assays were for genomic DNA identification (for the certified value) and two of the assays were for mitochondrial detection which is used for informational values (and are not included in this data set) Each of the plates contained samples from a new vial from different boxes (noted in the ldquovial usedrdquo portion the number references the box) The assay positions were alternated across the five plates as well For the plate design and which tubes were used see

Figure 6 Basic Plate Layout Design for Certification Measurements It was later determined that the homogeneity and stability data sets were obtained using a different set of pipettes than the certification data set and since all data was to be combined for certification another plate was obtained using the original pipette For the complete certification data see

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component Stability data was measured at different temperatures over several weeks using assays NEIF and ND6 Only the 4degC data was also used for certification For the complete stability data see

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

32 Model It was decided that all three sets of data should be used to derive the certified values In this way the uncertainty of the certified values would better reflect any potential heterogeneity in the samples as well as any additional unknown factors that may have effected the results Because two different pipettes were used to obtain different sets of data there was a significant bias in some of the measurements Specifically the ldquocertificationrdquo data from the first five plates was obtained using pipette set ldquo1rdquo and the remaining measurements were obtained using pipette set ldquo2rdquo This bias was introduced into the statistical model as part of the mean Possible effects of different assays and type of data (heterogeneity certification and stability) were accounted for in the statistical model in terms of a random effect Specifically the response for each component (A B C) was represented as

119910119910119894119894119894119894119894119894~1198731198731205741205741198941198941198941198941198941198941205901205901198941198942 where ~N() denotes ldquodistributed as a normal distribution of some mean and variancerdquo i = 123 denotes type of measurement heterogeneity certification and stability j = 1hellip10 denotes assay 2PR4 POTP NEIF R4Q5 D5 ND6 D9 HBB1 ND14 22C3 k = 1 2 denotes pipette set 120574120574119894119894119894119894119894119894 = 120572120572119894119894119894119894 + 120573120573119894119894 120572120572119894119894119894119894~119873119873(micro1198941198941205901205901205721205722) 120573120573119894119894 = 120573120573 for data type 1 and 120573120573119894119894 = 0 for data types 2 and 3 For a given 2372a component the variance 1205901205901205721205722 measures the between assay variability and 1205901205901198941198942 measures the repeatability variability for each data type The data was analyzed using Bayesian MCMC programmed in OpenBUGSC2 with non-informative Gaussian priors for the means and Gamma priors for the variance components

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

58

33 OpenBUGS code inits and data Certification code ModelBegin for(i in 13)mu[i]~dnorm(010E-5) sig[i]~dgamma(10E-510E-5) sigalph~dgamma(10E-510E-5) for(i in 110)alpha[1i]~dnorm(mu[1] sigalph) for(i in 12)alpha[2i]~dnorm(mu[2] sigalph) alpha[3i]~dnorm(mu[3]sigalph) b~dnorm(010E-5) beta[1]lt-b beta[2]lt-00 for(i in 1Ntot)mean[i]lt-alpha[typ[i]asy[i]]+beta[pip[i]] for(i in 1Ntot)lamb[i]~dnorm(mean[i]sig[typ[i]]) ModelEnd Certification Inits list(sig=c(111) sigalph=1) Certification Data for Component A list(Ntot=275) lamb[] asy[] cmp[] typ[] pip[] 1667 1 1 1 1 1655 2 1 1 1 1647 3 1 1 1 1594 4 1 1 1 1655 5 1 1 1 1629 6 1 1 1 1655 7 1 1 1 1644 8 1 1 1 1645 9 1 1 1 1545 10 1 1 1 1698 1 1 1 1 1685 2 1 1 1 1665 3 1 1 1 1658 4 1 1 1 1683 5 1 1 1 1647 6 1 1 1 1652 7 1 1 1 1611 8 1 1 1 1707 9 1 1 1 1603 10 1 1 1 1668 1 1 1 1 1587 2 1 1 1 1678 3 1 1 1 1628 4 1 1 1 1738 5 1 1 1 1667 6 1 1 1 1695 7 1 1 1 1757 8 1 1 1 1771 9 1 1 1 1762 10 1 1 1 1739 1 1 1 1 1702 2 1 1 1 1614 3 1 1 1 1762 4 1 1 1 1649 5 1 1 1 1678 6 1 1 1 1653 7 1 1 1

175 8 1 1 1 1764 9 1 1 1 1759 10 1 1 1 1638 1 1 1 1 1676 2 1 1 1 1616 3 1 1 1 1559 4 1 1 1 1608 5 1 1 1 1611 6 1 1 1 1517 7 1 1 1 1557 8 1 1 1 1604 9 1 1 1 1675 10 1 1 1 1667 1 1 1 1 1638 2 1 1 1 1624 3 1 1 1 1587 4 1 1 1 1652 5 1 1 1 1589 6 1 1 1 1607 7 1 1 1 1582 8 1 1 1 1625 9 1 1 1 1616 10 1 1 1 1888 1 1 1 1 1868 2 1 1 1 1953 3 1 1 1 1779 4 1 1 1 1924 5 1 1 1 1861 6 1 1 1 1824 7 1 1 1 1809 8 1 1 1 1864 9 1 1 1 193 10 1 1 1 1939 1 1 1 1 1815 2 1 1 1 1902 3 1 1 1 1871 4 1 1 1 2124 5 1 1 1

1787 6 1 1 1 1844 7 1 1 1 1799 8 1 1 1 1857 9 1 1 1 1942 10 1 1 1 1597 1 1 1 1 1533 2 1 1 1 1462 3 1 1 1 1493 4 1 1 1 1538 5 1 1 1 1492 6 1 1 1 1528 7 1 1 1 1548 8 1 1 1 1513 9 1 1 1 1562 10 1 1 1 159 1 1 1 1 1404 2 1 1 1 1533 3 1 1 1 1519 4 1 1 1 1586 5 1 1 1 1516 6 1 1 1 1538 7 1 1 1 1581 8 1 1 1 1514 9 1 1 1 1431 1 1 1 2 1407 2 1 1 2 1487 3 1 1 2 145 4 1 1 2 1394 5 1 1 2 1396 6 1 1 2 1414 7 1 1 2 1401 8 1 1 2 1412 9 1 1 2 1435 10 1 1 2 1411 1 1 1 2 1395 2 1 1 2 1398 3 1 1 2 1439 4 1 1 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

59

1422 5 1 1 2 1393 6 1 1 2 139 7 1 1 2 1382 8 1 1 2 146 9 1 1 2 141 10 1 1 2 1632 1 1 2 2 1531 1 1 2 2 1524 1 1 2 2 1454 1 1 2 2 1465 1 1 2 2 1512 1 1 2 2 1418 1 1 2 2 1496 1 1 2 2 1588 1 1 2 2 1574 1 1 2 2 1557 1 1 2 2 15 1 1 2 2 1672 1 1 2 2 1551 1 1 2 2 1573 1 1 2 2 1516 1 1 2 2 1612 1 1 2 2 1598 1 1 2 2 1575 1 1 2 2 1658 1 1 2 2 1547 1 1 2 2 1529 1 1 2 2 1525 1 1 2 2 154 1 1 2 2 1458 1 1 2 2 154 1 1 2 2 1491 1 1 2 2 148 1 1 2 2 156 1 1 2 2 1581 1 1 2 2 1513 1 1 2 2 1508 1 1 2 2 1476 1 1 2 2 1634 1 1 2 2 1568 1 1 2 2 152 1 1 2 2 1516 1 1 2 2 1535 1 1 2 2 1514 1 1 2 2 1568 1 1 2 2 1611 1 1 2 2 1528 1 1 2 2 1511 1 1 2 2 1549 1 1 2 2 1533 1 1 2 2 1427 1 1 2 2 1504 1 1 2 2 1491 1 1 2 2 1455 1 1 2 2

1501 1 1 2 2 1637 1 1 2 2 157 1 1 2 2 1633 1 1 2 2 1515 1 1 2 2 1804 1 1 2 2 1538 1 1 2 2 1617 1 1 2 2 1628 1 1 2 2 1672 1 1 2 2 1596 1 1 2 2 1586 1 1 2 2 1578 1 1 2 2 1532 1 1 2 2 1616 2 1 2 2 1615 2 1 2 2 1582 2 1 2 2 1627 2 1 2 2 1607 2 1 2 2 1511 2 1 2 2 1479 2 1 2 2 1514 2 1 2 2 1474 2 1 2 2 1594 2 1 2 2 1571 2 1 2 2 1626 2 1 2 2 149 2 1 2 2 1617 2 1 2 2 1634 2 1 2 2 1613 2 1 2 2 1529 2 1 2 2 1571 2 1 2 2 1652 2 1 2 2 1647 2 1 2 2 1537 2 1 2 2 164 2 1 2 2 1589 2 1 2 2 1152 2 1 2 2 1605 2 1 2 2 1633 2 1 2 2 1483 2 1 2 2 1557 2 1 2 2 1615 2 1 2 2 1438 2 1 2 2 1713 2 1 2 2 1611 2 1 2 2 1586 2 1 2 2 1516 2 1 2 2 1614 2 1 2 2 1627 2 1 2 2 1567 2 1 2 2 1495 2 1 2 2 1535 2 1 2 2 1545 2 1 2 2 1607 2 1 2 2

1513 2 1 2 2 1582 2 1 2 2 1601 2 1 2 2 1555 2 1 2 2 1607 2 1 2 2 1586 2 1 2 2 1524 2 1 2 2 1656 2 1 2 2 154 2 1 2 2 1454 2 1 2 2 1669 2 1 2 2 1631 2 1 2 2 1608 2 1 2 2 1497 2 1 2 2 1639 2 1 2 2 1575 2 1 2 2 1567 2 1 2 2 1528 2 1 2 2 1511 2 1 2 2 1627 2 1 2 2 1612 2 1 2 2 1536 2 1 2 2 1568 1 1 3 2 152 1 1 3 2 1486 1 1 3 2 1441 1 1 3 2 1706 1 1 3 2 144 1 1 3 2 15 1 1 3 2 1537 1 1 3 2 1387 1 1 3 2 1463 1 1 3 2 157 1 1 3 2 1429 1 1 3 2 1667 1 1 3 2 1632 1 1 3 2 1589 1 1 3 2 1474 2 1 3 2 1543 2 1 3 2 1555 2 1 3 2 1599 2 1 3 2 1494 2 1 3 2 1624 2 1 3 2 1577 2 1 3 2 1593 2 1 3 2 1479 2 1 3 2 1517 2 1 3 2 1538 2 1 3 2 159 2 1 3 2 1479 2 1 3 2 1479 2 1 3 2 1522 2 1 3 2 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

60

Table C-2 Example OpenBUGS Certification Summary for Component A

mean sd MC_error val25pc median val975pc start sample b 2584 02948 0005376 2001 2588 3158 100001 100000 mu[1] 1417 02706 0005366 1364 1416 147 100001 100000 mu[2] 1559 008368 4776E-4 1543 1559 1576 100001 100000 mu[3] 1533 01504 0001419 1504 1533 1563 100001 100000

Certification Data for Component B list(Ntot=278) lamb[] asy[] cmp[] typ[] pip[] 1869 1 2 1 1 185 2 2 1 1 1709 3 2 1 1 1856 4 2 1 1 1848 5 2 1 1 1828 6 2 1 1 1861 7 2 1 1 1825 8 2 1 1 1871 9 2 1 1 1776 10 2 1 1 196 1 2 1 1 1781 2 2 1 1 1707 3 2 1 1 1787 4 2 1 1 1829 5 2 1 1 1826 6 2 1 1 1858 7 2 1 1 1869 8 2 1 1 1773 9 2 1 1 1902 10 2 1 1 2174 1 2 1 1 2202 2 2 1 1 2172 3 2 1 1 223 5 2 1 1 2094 6 2 1 1 1949 7 2 1 1 1971 8 2 1 1 2123 9 2 1 1 2107 10 2 1 1 2231 1 2 1 1 2292 2 2 1 1 2216 3 2 1 1 2078 4 2 1 1 2192 5 2 1 1 2034 6 2 1 1 211 7 2 1 1 2123 8 2 1 1 2066 9 2 1 1 2046 10 2 1 1 181 1 2 1 1 1865 2 2 1 1 1806 3 2 1 1 1878 4 2 1 1 1824 5 2 1 1 1768 6 2 1 1 1811 7 2 1 1 1763 8 2 1 1 1865 9 2 1 1 1815 10 2 1 1

1806 1 2 1 1 1801 2 2 1 1 1786 3 2 1 1 1768 4 2 1 1 179 5 2 1 1 167 6 2 1 1 1749 7 2 1 1 1741 8 2 1 1 1781 9 2 1 1 1866 10 2 1 1 1784 1 2 1 1 1735 2 2 1 1 1778 3 2 1 1 1835 4 2 1 1 1794 5 2 1 1 17 6 2 1 1 1799 7 2 1 1 1765 8 2 1 1 1754 9 2 1 1 1792 10 2 1 1 1821 1 2 1 1 1813 2 2 1 1 1813 3 2 1 1 1749 4 2 1 1 1848 5 2 1 1 1711 6 2 1 1 1779 7 2 1 1 1769 8 2 1 1 1725 9 2 1 1 1814 10 2 1 1 1718 1 2 1 1 1753 2 2 1 1 1773 3 2 1 1 1777 4 2 1 1 1803 5 2 1 1 1718 6 2 1 1 1802 7 2 1 1 1834 8 2 1 1 1731 9 2 1 1 1842 10 2 1 1 1768 1 2 1 1 1698 2 2 1 1 1792 3 2 1 1 174 4 2 1 1 1795 5 2 1 1 1693 6 2 1 1 1743 7 2 1 1 1745 8 2 1 1 1714 9 2 1 1 1791 1 2 1 2

1746 2 2 1 2 1803 3 2 1 2 1767 4 2 1 2 1745 5 2 1 2 1824 6 2 1 2 1752 7 2 1 2 1813 8 2 1 2 1648 9 2 1 2 1818 10 2 1 2 177 1 2 1 2 182 2 2 1 2 1803 3 2 1 2 1739 4 2 1 2 1773 5 2 1 2 1729 6 2 1 2 1744 7 2 1 2 1754 8 2 1 2 1671 9 2 1 2 1741 10 2 1 2 179 1 2 2 2 1712 1 2 2 2 1693 1 2 2 2 1754 1 2 2 2 1757 1 2 2 2 1697 1 2 2 2 1722 1 2 2 2 1884 1 2 2 2 1713 1 2 2 2 1759 1 2 2 2 1749 1 2 2 2 1748 1 2 2 2 1632 1 2 2 2 1752 1 2 2 2 1871 1 2 2 2 1688 1 2 2 2 1789 1 2 2 2 1826 1 2 2 2 1751 1 2 2 2 1738 1 2 2 2 1742 1 2 2 2 1761 1 2 2 2 1909 1 2 2 2 1719 1 2 2 2 1755 1 2 2 2 1745 1 2 2 2 1687 1 2 2 2 1709 1 2 2 2 1625 1 2 2 2 1802 1 2 2 2 1734 1 2 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

61

1809 1 2 2 2 1782 1 2 2 2 1712 1 2 2 2 1668 1 2 2 2 1535 1 2 2 2 1835 1 2 2 2 1725 1 2 2 2 1769 1 2 2 2 178 1 2 2 2 1759 1 2 2 2 1678 1 2 2 2 1836 1 2 2 2 1743 1 2 2 2 1885 1 2 2 2 1704 1 2 2 2 1749 1 2 2 2 1739 1 2 2 2 1735 1 2 2 2 1714 1 2 2 2 1703 1 2 2 2 1772 1 2 2 2 1727 1 2 2 2 1738 1 2 2 2 174 1 2 2 2 1718 1 2 2 2 1701 1 2 2 2 1673 1 2 2 2 1887 1 2 2 2 1743 1 2 2 2 1825 1 2 2 2 1849 1 2 2 2 188 1 2 2 2 1768 1 2 2 2 1713 1 2 2 2 166 1 2 2 2 1775 2 2 2 2 1824 2 2 2 2 167 2 2 2 2 1683 2 2 2 2 1693 2 2 2 2 169 2 2 2 2 1864 2 2 2 2 1709 2 2 2 2 1764 2 2 2 2

1741 2 2 2 2 1797 2 2 2 2 1626 2 2 2 2 1724 2 2 2 2 177 2 2 2 2 1744 2 2 2 2 1716 2 2 2 2 1805 2 2 2 2 1827 2 2 2 2 1752 2 2 2 2 1746 2 2 2 2 1666 2 2 2 2 187 2 2 2 2 1718 2 2 2 2 1483 2 2 2 2 1632 2 2 2 2 1759 2 2 2 2 1723 2 2 2 2 1684 2 2 2 2 1887 2 2 2 2 1756 2 2 2 2 1816 2 2 2 2 1792 2 2 2 2 1614 2 2 2 2 1566 2 2 2 2 1741 2 2 2 2 1752 2 2 2 2 1697 2 2 2 2 1762 2 2 2 2 1801 2 2 2 2 1766 2 2 2 2 1747 2 2 2 2 1744 2 2 2 2 1692 2 2 2 2 186 2 2 2 2 1756 2 2 2 2 1896 2 2 2 2 1814 2 2 2 2 1703 2 2 2 2 1722 2 2 2 2 1688 2 2 2 2 1821 2 2 2 2 1758 2 2 2 2 1787 2 2 2 2

1784 2 2 2 2 179 2 2 2 2 1681 2 2 2 2 1685 2 2 2 2 1798 2 2 2 2 171 2 2 2 2 1638 2 2 2 2 1931 2 2 2 2 1786 2 2 2 2 1768 2 2 2 2 1764 2 2 2 2 1819 2 2 2 2 1666 1 2 3 2 1626 1 2 3 2 188 1 2 3 2 1653 1 2 3 2 1733 1 2 3 2 1551 1 2 3 2 1884 1 2 3 2 1718 1 2 3 2 1549 1 2 3 2 1723 1 2 3 2 194 1 2 3 2 164 1 2 3 2 1565 1 2 3 2 1787 1 2 3 2 1655 2 2 3 2 167 2 2 3 2 1673 2 2 3 2 1705 2 2 3 2 1755 2 2 3 2 1625 2 2 3 2 1708 2 2 3 2 1705 2 2 3 2 1862 2 2 3 2 1686 2 2 3 2 1723 2 2 3 2 1909 2 2 3 2 1859 2 2 3 2 1897 2 2 3 2 1972 2 2 3 2 END

Table C-3 Example OpenBUGS Certification Summary for Component B

mean sd MC_error val25pc median val975pc start sample b 09337 03387 0007504 02695 09371 1592 100001 100000 mu[1] 1764 03101 0007483 1704 1763 1825 100001 100000 mu[2] 1748 008014 4719E-4 1733 1748 1764 100001 100000 mu[3] 1735 02338 0003093 1688 1736 1781 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

62

Certification data for Component C list(Ntot=280) lamb[] asy[] cmp[] typ[] pip[] 163 1 3 1 1 1596 2 3 1 1 1574 3 3 1 1 159 4 3 1 1 161 5 3 1 1 1589 6 3 1 1 1569 7 3 1 1 1565 8 3 1 1 1564 9 3 1 1 1548 10 3 1 1 1706 1 3 1 1 1576 2 3 1 1 1538 3 3 1 1 1555 4 3 1 1 1607 5 3 1 1 1548 6 3 1 1 1572 7 3 1 1 1681 8 3 1 1 1492 9 3 1 1 1655 10 3 1 1 1499 1 3 1 1 1576 2 3 1 1 1545 3 3 1 1 1584 4 3 1 1 1577 5 3 1 1 1599 6 3 1 1 1587 7 3 1 1 152 8 3 1 1 1479 9 3 1 1 1512 10 3 1 1 1497 1 3 1 1 1611 2 3 1 1 156 3 3 1 1 1546 4 3 1 1 1529 5 3 1 1 1478 6 3 1 1 1504 7 3 1 1 1551 8 3 1 1 1438 9 3 1 1 1494 10 3 1 1 1452 1 3 1 1 1347 2 3 1 1 1395 3 3 1 1 1442 4 3 1 1 1448 5 3 1 1 1463 6 3 1 1 1402 7 3 1 1 1474 8 3 1 1 1466 9 3 1 1 1439 10 3 1 1 1424 1 3 1 1 1397 2 3 1 1 135 3 3 1 1 1459 4 3 1 1 1602 5 3 1 1 1435 6 3 1 1 1433 7 3 1 1 1481 8 3 1 1 1482 9 3 1 1

1568 10 3 1 1 147 1 3 1 1 1411 2 3 1 1 1512 3 3 1 1 1452 4 3 1 1 1567 5 3 1 1 148 6 3 1 1 1442 7 3 1 1 143 8 3 1 1 1413 9 3 1 1 1486 10 3 1 1 1444 1 3 1 1 1427 2 3 1 1 1439 3 3 1 1 1468 4 3 1 1 1524 5 3 1 1 1424 6 3 1 1 1459 7 3 1 1 1467 8 3 1 1 1411 9 3 1 1 148 10 3 1 1 1485 1 3 1 1 1422 2 3 1 1 1438 3 3 1 1 1439 4 3 1 1 1515 5 3 1 1 1409 6 3 1 1 1469 7 3 1 1 1453 8 3 1 1 1465 9 3 1 1 1363 10 3 1 1 1458 1 3 1 1 1459 2 3 1 1 1465 3 3 1 1 1468 4 3 1 1 1457 5 3 1 1 1412 6 3 1 1 1474 7 3 1 1 1428 8 3 1 1 1425 9 3 1 1 1393 10 3 1 1 1357 1 3 1 2 1447 2 3 1 2 1448 3 3 1 2 1409 4 3 1 2 1389 5 3 1 2 1415 6 3 1 2 1431 7 3 1 2 1381 8 3 1 2 1403 9 3 1 2 1435 10 3 1 2 1364 1 3 1 2 1361 2 3 1 2 1404 3 3 1 2 1411 4 3 1 2 1429 5 3 1 2 1417 6 3 1 2 138 7 3 1 2 1336 8 3 1 2 1387 9 3 1 2

1372 10 3 1 2 1493 1 3 2 2 1516 1 3 2 2 1451 1 3 2 2 144 1 3 2 2 1523 1 3 2 2 1437 1 3 2 2 1485 1 3 2 2 1433 1 3 2 2 1463 1 3 2 2 1534 1 3 2 2 1533 1 3 2 2 145 1 3 2 2 1537 1 3 2 2 1515 1 3 2 2 1511 1 3 2 2 1449 1 3 2 2 1388 1 3 2 2 1475 1 3 2 2 1468 1 3 2 2 1549 1 3 2 2 153 1 3 2 2 1442 1 3 2 2 1385 1 3 2 2 1513 1 3 2 2 1528 1 3 2 2 1486 1 3 2 2 1472 1 3 2 2 1493 1 3 2 2 1477 1 3 2 2 1405 1 3 2 2 1499 1 3 2 2 1458 1 3 2 2 1514 1 3 2 2 1525 1 3 2 2 1527 1 3 2 2 149 1 3 2 2 1568 1 3 2 2 1452 1 3 2 2 1421 1 3 2 2 1455 1 3 2 2 1525 1 3 2 2 1531 1 3 2 2 144 1 3 2 2 1487 1 3 2 2 151 1 3 2 2 1577 1 3 2 2 1506 1 3 2 2 152 1 3 2 2 1483 1 3 2 2 1522 1 3 2 2 1463 1 3 2 2 1478 1 3 2 2 1505 1 3 2 2 1454 1 3 2 2 1555 1 3 2 2 1497 1 3 2 2 1515 1 3 2 2 1494 1 3 2 2 1594 1 3 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

63

1471 1 3 2 2 1394 1 3 2 2 1445 1 3 2 2 1446 1 3 2 2 1499 1 3 2 2 1484 1 3 2 2 1436 1 3 2 2 1565 2 3 2 2 149 2 3 2 2 1485 2 3 2 2 1529 2 3 2 2 1513 2 3 2 2 1492 2 3 2 2 1475 2 3 2 2 1497 2 3 2 2 1503 2 3 2 2 1448 2 3 2 2 1488 2 3 2 2 1496 2 3 2 2 1499 2 3 2 2 1469 2 3 2 2 1515 2 3 2 2 1543 2 3 2 2 1452 2 3 2 2 1458 2 3 2 2 1435 2 3 2 2 1499 2 3 2 2 1499 2 3 2 2 1589 2 3 2 2 1562 2 3 2 2 1547 2 3 2 2 1505 2 3 2 2 1508 2 3 2 2 152 2 3 2 2

1481 2 3 2 2 1547 2 3 2 2 1501 2 3 2 2 1491 2 3 2 2 1465 2 3 2 2 147 2 3 2 2 1483 2 3 2 2 1479 2 3 2 2 148 2 3 2 2 1543 2 3 2 2 1448 2 3 2 2 1531 2 3 2 2 1478 2 3 2 2 1465 2 3 2 2 1454 2 3 2 2 1591 2 3 2 2 1466 2 3 2 2 157 2 3 2 2 1471 2 3 2 2 1506 2 3 2 2 1491 2 3 2 2 1461 2 3 2 2 1485 2 3 2 2 1495 2 3 2 2 1516 2 3 2 2 1474 2 3 2 2 1483 2 3 2 2 1519 2 3 2 2 1475 2 3 2 2 1466 2 3 2 2 1565 2 3 2 2 1491 2 3 2 2 1435 2 3 2 2 1459 2 3 2 2

1421 2 3 2 2 1468 2 3 2 2 1556 2 3 2 2 1424 1 3 3 2 1427 1 3 3 2 1381 1 3 3 2 1394 1 3 3 2 1473 1 3 3 2 133 1 3 3 2 1446 1 3 3 2 1424 1 3 3 2 133 1 3 3 2 1452 1 3 3 2 1512 1 3 3 2 1314 1 3 3 2 1525 1 3 3 2 1475 1 3 3 2 1542 1 3 3 2 1461 2 3 3 2 1443 2 3 3 2 1547 2 3 3 2 1394 2 3 3 2 1364 2 3 3 2 1395 2 3 3 2 1522 2 3 3 2 1483 2 3 3 2 1452 2 3 3 2 1482 2 3 3 2 1439 2 3 3 2 1458 2 3 3 2 155 2 3 3 2 1502 2 3 3 2 152 2 3 3 2 END

Table C-4 Example OpenBUGS Certification Summary for Component C

mean sd MC_error val25pc median val975pc start sample b 09626 01701 0002428 06285 09633 1293 100001 100000 mu[1] 1399 01566 0002413 1369 1399 143 100001 100000 mu[2] 1491 005701 2386E-4 148 1491 1503 100001 100000 mu[3] 1449 0132 0001181 1423 1449 1475 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

64

34 Results Table C-5 lists the results in CopiesnL for the three components

Table C-5 Parameter Values for Components A B and C

Posterior Distribution Summary Statistics Component Parameter Mean SD 25 50 975

A

β 260 029 203 260 318 μ1 1415 027 1362 1415 1467 μ2 1559 010 1539 1559 158 μ3 1533 016 1501 1533 1565

B

β 095 034 029 095 162 μ1 1762 031 1701 1762 1822 μ2 1748 010 1728 1748 1769 μ3 1735 024 1687 1735 1783

C

β 097 017 064 097 130 μ1 1399 016 1368 1399 1429 μ2 1491 008 1476 1491 1507 μ3 1449 014 1421 1449 1477

Since in each case the three data set types represent independent measurements the three estimates can be combined to obtain a consensus value In this way a between data set type variability is included in the final standard uncertainty Table C-6 lists the results in Copies per Nanoliter obtained using the NIST Consensus Builder (NICOB)C3

Table C-6 Certified Values for Components A B and C

Component Mean u(Mean) Uk=2(Mean) A 151 07 14 B 175 03 06 C 145 05 10

4 Mito-to-nuclear DNA ratios (mtDNAnDNA)

41 Data Measurements were also obtained on the same samples using three mitochondrial detection assays mtND1 mtBatz and mtKav and a single nDNA assay (genomic) to compute the ratios mtDNAnDNA For the complete mtDNAnDNA data see

Table 16 mtDNAnDNA Ratios

42 Model For each component the data was analyzed using the following random effects model to account for between assay variability

119903119903119894119894~119873119873120572120572119894119894 1205901205901198861198861198861198861198861198861198861198861198861198862

where 119894119894 denotes assay (for A and B 1 = mtND1 2 = mtBatz 3 = mtKav for C 1 = mtBatz 2 = mtKav) and 120572120572119894119894~119873119873micro120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119886119886119886119886119886119886119886119886119886119886

2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

65

43 OpenBUGS code inits and data Two versions of the same model are required one for the results from the three mtDNA assays available for components A and B mtND1 mtBatz and mtKav and the other for the results from the two mtDNA assays available for component C mtBatz and mtKav mtDNAnDNA code for components A and B ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 13)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for components A and B list(sigt=c(111)) Data for Component A list(Ndata=102) set[] rep[] asy[] rat[] 1 1 1 1931 1 2 1 187 1 1 2 1845 1 2 2 1764 2 1 1 181 2 2 1 1906 2 1 2 1785 2 2 2 1837 3 1 1 1743 3 2 1 1761 3 1 2 1829 3 2 2 1847 4 1 1 1716 4 2 1 180 4 1 2 174 4 2 2 1813 5 1 1 1923 5 2 1 1835 5 1 2 1935 5 2 2 1851 6 1 1 1757 6 2 1 1724 6 3 1 1768 6 1 2 1743 6 2 2 1671 6 3 2 1766 6 1 3 1692 6 2 3 1662 6 3 3 1767 7 1 1 1924 7 2 1 1944 7 3 1 1893 7 1 2 1786

7 2 2 1752 7 3 2 1743 7 1 3 153 7 2 3 1542 7 3 3 1523 7 1 2 158 7 2 2 1557 7 3 2 1592 7 1 3 2103 7 2 3 210 7 3 3 2049 7 1 1 1749 7 2 1 1769 7 3 1 1777 7 1 2 1965 7 2 2 1889 7 3 2 1947 7 1 3 1813 7 2 3 1864 7 3 3 1864 8 1 1 178 8 2 1 1776 8 3 1 1706 8 1 2 1804 8 2 2 1751 8 3 2 1738 8 1 3 1769 8 2 3 1786 8 3 3 1761 9 1 2 1584 9 2 2 1544 10 1 1 1947 10 2 1 1928 10 3 1 1922 10 1 1 1829

10 2 1 1895 10 3 1 1851 10 1 1 1674 10 2 1 1617 11 1 2 1667 11 2 2 1813 11 3 2 1708 11 4 2 1598 11 1 2 1566 11 2 2 1687 11 3 2 1663 11 4 2 1589 11 1 2 1643 11 2 2 1668 11 3 2 1604 11 4 2 1573 12 1 2 1719 12 2 2 1716 12 3 2 167 12 4 2 1666 12 1 2 1581 12 2 2 1573 12 3 2 1601 12 4 2 1612 12 1 2 1546 12 2 2 1542 12 3 2 1538 12 4 2 1587 13 1 1 1617 13 2 1 1599 13 1 1 1622 13 2 1 1579 13 1 1 1575 13 2 1 1528 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

66

Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A

mean sd MC_error val25pc median val975pc start sample mu 1730 1443 004419 1703 1730 1759 100001 100000

data for Component B list(Ndata=87) set[] rep[] asy[] rat[] 1 1 1 2461 1 2 1 2401 1 1 2 2555 1 2 2 2608 2 1 1 2229 2 2 1 2200 2 1 2 2108 2 2 2 2192 3 1 1 2127 3 2 1 2198 3 1 2 2455 3 2 2 2506 4 1 1 2211 4 2 1 2199 4 1 2 2380 4 2 2 2249 5 1 1 2105 5 2 1 1943 5 1 2 2136 5 2 2 2119 6 1 1 1990 6 2 1 1921 6 3 1 1982 6 1 2 1990 6 2 2 1924 6 3 2 1958 6 1 3 1903 6 2 3 1831

6 3 3 1919 7 1 1 2304 7 2 1 2341 7 3 1 2332 7 1 2 2048 7 2 2 2078 7 3 2 2067 7 1 3 2399 7 2 3 2479 7 3 3 2493 7 1 1 2065 7 2 1 2166 7 3 1 2164 7 1 2 2314 7 2 2 2280 7 3 2 2258 7 1 3 1867 7 2 3 1987 7 3 3 1971 8 1 1 2046 8 2 1 2057 8 3 1 2024 8 1 2 1998 8 2 2 2043 8 3 2 1952 8 1 3 2077 8 2 3 2117 8 3 3 2041 9 1 2 1890 9 2 2 1886

10 1 2 1928 10 2 2 1916 10 3 2 1915 10 1 2 1947 10 2 2 1946 10 3 2 1961 10 4 2 1909 10 1 2 1953 10 2 2 1952 10 3 2 1941 10 4 2 1881 11 1 2 1865 11 2 2 1863 11 3 2 1828 11 4 2 1828 11 1 2 1944 11 2 2 1964 11 3 2 1947 11 4 2 1895 11 1 2 1908 11 2 2 1909 11 3 2 1940 11 4 2 1839 12 1 1 1688 12 2 1 1783 12 1 1 1823 12 2 1 1631 12 1 1 1920 12 2 1 1887 END

Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B

mean sd MC_error val25pc median val975pc start sample mu 2045 2246 0111 1999 2045 2088 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

67

mtDNAnDNA code for component C ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 12)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for component C list(sigt=c(11)) mtDNAnDNA data for component C list(Ndata=66) set[] rep[] asy[] rat[] 1 1 1 3201 1 2 1 3299 2 1 1 3347 2 2 1 3287 3 1 1 3217 3 2 1 3384 4 1 1 3200 4 2 1 3363 5 1 1 3294 5 2 1 3013 6 1 1 2805 6 2 1 2825 6 3 1 2829 6 1 2 2709 6 2 2 2738 6 3 2 2727 7 1 1 2797 7 2 1 2642 7 3 1 2792 7 1 2 2804 7 2 2 2789

7 3 2 2842 7 1 1 2715 7 2 1 2606 7 3 1 2687 7 1 2 2799 7 2 2 2628 7 3 2 2559 7 1 1 2639 7 2 1 2535 7 3 1 2657 7 1 2 2953 7 2 2 2757 7 3 2 2802 8 1 1 2927 8 2 1 2776 8 3 1 2762 8 1 2 3044 8 2 2 2830 8 3 2 2825 9 1 1 2501 9 2 1 2522 10 1 1 2832 10 2 1 2840

10 3 1 2788 10 4 1 2725 10 1 1 2835 10 2 1 2790 10 3 1 2736 10 4 1 2679 10 1 1 2803 10 2 1 2773 10 3 1 2757 10 4 1 2752 11 1 1 2599 11 2 1 2508 11 3 1 2551 11 4 1 2609 11 1 1 2617 11 2 1 2483 11 3 1 2626 11 4 1 2602 11 1 1 2665 11 2 1 2535 11 3 1 2570 11 4 1 2650 END

Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C

mean sd MC_error val25pc median val975pc start sample mu 2774 2367 01176 2729 2774 2821 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

68

Table C-10 Results for mtDNAnDNA Posterior Distribution Summary Statistics

Component Mean Std Dev 25 50 975 A 1730 14 1703 1730 1758 B 2045 23 1994 2047 2087 C 2774 23 2729 2774 2821

5 ReferencesC1 Lunn DJ Spiegelhalter D Thomas A Best N Statistics in Medicine 2009283049ndash3082C2 Polson NG Scott JG Bayesian Analysis 20127887ndash902C3 NIST Consensus Builder Software and userrsquos guide freely available at

httpsconsensusnistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

69

Appendix D Performance in Commercial qPCR Chemistries The performance of the SRM 2372a materials in commercial qPCR chemistries was assessed using manufacturerrsquos recommended protocols Each panel in Figures D-1 to D-4 displays the average cycle threshold (Ct) as a function of DNA concentration for the 5-point dilutions of SRM 2372a components In all cases the external calibrant was SRM 2372 component A Error bars represent approximate 95 confidence intervals Table D-1 lists the estimated [DNA] for the three SRM 2372a components for all commercial assays evaluated This Table also lists the Component BComponent A (BA) and Component CComponent A (CA) ratios calculated from the [DNA] estimates

Figure D-1 Results for the Innogenomics Commercial qPCR assays

20

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantLong(IQ_h)

ABCStd

[A] = (384 plusmn 26) ngμL[B] = (441 plusmn 24) ngμL[C] = (516 plusmn 24) ngμL

Rsup2 = 0999 se(Ct) = 01016

17

18

19

20

21

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantShort(IQ_d)

ABCStd

[A] = (435 plusmn 28) ngμL[B] = (543 plusmn 12) ngμL[C] = (510 plusmn 28) ngμL

Rsup2 = 0999 se(Ct) = 007

24

25

26

27

28

29

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYLong(IQHY_h)

ABCStd

[A] = (387 plusmn 24) ngμL[B] = (448 plusmn 21) ngμL[C] = (559 plusmn 31) ngμL

Rsup2 = 0998 se(Ct) = 01314

16

18

20

22

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYShort(IQHY_d)

ABCStd

[A] = (353 plusmn 23) ngμL[B] = (267 plusmn 15) ngμL[C] = (390 plusmn 18) ngμL

Rsup2 = 1000 se(Ct) = 00420

22

24

26

28

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYY(IQHY_y)

ABCStd

[A] = (572 plusmn 13) ngμL[B] = 0 ngμL[C] = (137 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

70

Figure D-2 Results for the Thermo Fisher Commercial qPCR assays

21

22

23

24

25

26

27

28

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HumanHuman(QFH_h)

ABCStd

[A] = (464 plusmn 13) ngμL[B] = (529 plusmn 19) ngμL[C] = (472 plusmn 22) ngμL

Rsup2 = 1000 se(Ct) = 004

22

23

24

25

26

27

28

29

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoHuman(QFD_h)

ABCStd

[A] = (701 plusmn 53) ngμL[B] = (561 plusmn 11) ngμL[C] = (472 plusmn 11) ngμL

Rsup2 = 0994 se(Ct) = 01922

24

26

28

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoMale(QFD_y)

ABCStd

[A] = (706 plusmn 49) ngμL[B] = 0 ngμL[C] = (122 plusmn 02) ngμL

Rsup2 = 0999 se(Ct) = 007

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPLarge(QFHP_h)

ABCStd

[A] = (340 plusmn 18) ngμL[B] = (432 plusmn 05) ngμL[C] = (342 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 00720

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPSmall(QFHP_d)

ABCStd

[A] = (434 plusmn 15) ngμL[B] = (514 plusmn 10) ngμL[C] = (466 plusmn 07) ngμL

Rsup2 = 1000 se(Ct) = 005

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioLarge(QFT_h)

ABCStd

[A] = (362 plusmn 23) ngμL[B] = (430 plusmn 06) ngμL[C] = (378 plusmn 07) ngμL

Rsup2 = 0999 se(Ct) = 00920

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioSmall(QFT_d)

ABCStd

[A] = (468 plusmn 08) ngμL[B] = (505 plusmn 08) ngμL[C] = (523 plusmn 14) ngμL

Rsup2 = 1000 se(Ct) = 00318

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioY(QFT_y)

ABCStd

[A] = (372 plusmn 11) ngμL[B] = 0 ngμL[C] = (110 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 003

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

71

Figure D-3 Results for the Qiagen Commercial qPCR assays

Figure D-4 Results for the Promega Commercial qPCR assays

19

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresAutosomal(QPH_h)

ABCStd

[A] = (554 plusmn 21) ngμL[B] = (479 plusmn 05) ngμL[C] = (381 plusmn 09) ngμL

Rsup2 = 1000 se(Ct) = 00418

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresY(QPH_y)

ABCStd

[A] = (358 plusmn 10) ngμL[B] = 0 ngμL[C] = (118 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 007

16

17

18

19

20

21

22

23

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProHuman(QPP_h)

ABCStd

[A] = (525 plusmn 49) ngμL[B] = (480 plusmn 42) ngμL[C] = (467 plusmn 45) ngμL

Rsup2 = 0999 se(Ct) = 01019

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProDegradation(QPP_d)

ABCStd

[A] = (567 plusmn 18) ngμL[B] = (558 plusmn 08) ngμL[C] = (484 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 00410

15

20

25

30

35

-50 -40 -30 -20 -10 00 10

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProY(QPP_y)

ABCStd

[A] = (538 plusmn 27) ngμL[B] = (00 plusmn 00) ngμL[C] = (162 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 003

16

18

20

22

24

-05 00 05 10 15 20

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorAutosomal(P_h)

ABCStd

[A] = (759 plusmn 64) ngμL[B] = (775 plusmn 24) ngμL[C] = (1056 plusmn 17) ngμL

Rsup2 = 0995 se(Ct) = 01818

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorY(P_y)

ABCStd

[A] = (596 plusmn 51) ngμL[B] = 0 ngμL[C] = (117 plusmn 01) ngμL

Rsup2 = 0994 se(Ct) = 020

18

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantAutosomal(PQ_h)

ABCStd

[A] = (436 plusmn 13) ngμL[B] = (606 plusmn 20) ngμL[C] = (504 plusmn 10) ngμL

Rsup2 = 1000 se(Ct) = 00218

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantDegradation(PQ_d)

ABCStd

[A] = (420 plusmn 18) ngμL[B] = (529 plusmn 16) ngμL[C] = (530 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 00516

18

20

22

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantY(PQ_y)

ABCStd

[A] = (410 plusmn 20) ngμL[B] = 0 ngμL[C] = (139 plusmn 05) ngμL

Rsup2 = 0999 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

72

Table D-1 Summary [DNA] Results for the Commercial qPCR Assays Estimated [DNA] ngμL Ratios

A B C BA CA Assaya xb u(x)c xb u(x)c xb u(x)c yd u(y)e yd u(y)e

IQd 435 28 543 12 510 28 125 008 117 010 IQh 384 26 441 24 516 24 115 010 134 011

IQHYd 353 23 267 15 390 18 076 006 111 009 IQHYh 387 24 448 21 559 31 116 009 144 012

Ph 759 64 775 24 1056 16 102 009 139 012 PQd 420 18 529 16 530 06 126 007 126 006 PQh 436 13 606 20 504 10 139 006 116 004

QFDh 701 53 561 11 472 11 080 006 067 005 QFHh 464 13 529 19 472 22 114 005 102 006

QFHPd 434 15 514 10 466 07 118 005 107 004 QFHPh 340 18 432 05 342 03 127 007 101 005

QFTd 468 08 505 08 523 14 108 003 112 004 QFTh 362 23 430 06 378 07 118 008 104 007 QPHh 554 21 479 05 380 09 086 003 069 003 QPPd 567 18 558 08 484 11 098 003 085 003 QPPh 525 49 480 42 467 45 092 012 089 012

IQHYy 572 13 NDf 137 06

NRg

024 001 Py 596 51 NDf 117 01 020 002

PQy 410 20 NDf 139 05 034 002 QFDy 706 49 NDf 122 02 017 001 QFTy 372 11 NDf 110 06 030 002 QPHy 358 10 NDf 118 03 033 001 QPPy 538 27 0002 0001 162 11 030 003

a Assay codes are listed in Table 8 b Mean of dilution-adjusted results from five dilutions of the SRM 2372a component c Standard uncertainty of the mean estimated from the stdanrd deviation of the five dilutions d Ratio of the estimated [DNA] of component B or C relative to that of component A e Standard uncertainty of the ratio estimated from the standard uncertainties of the individual components f Not detected Component B was prepared using tissue from only female donors it is not expected to

respond to male-specific assays g Not relevant

  • Abstract
  • Keywords
  • Technical Information Contact for this SRM
  • Table of Contents
  • Table of Tables
  • Table of Figures
  • Purpose and Description
  • Warning SRM 2372a is a Human Source Material
  • Storage and Use
  • History and Background
    • Figure 1 Sales History of SRM 2372
      • Experimental Methods
        • Polymerase Chain Reaction Assays
          • Table 1 NIST-Developed Human Nuclear DNA Assays
          • Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays
          • Table 2 NIST-Optimized Human Mitochondrial DNA Assays
          • Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays
            • Droplet Digital Polymerase Chain Reaction Measurements
              • Nuclear DNA (nDNA)
                • Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays
                  • Mitochondrial DNA (mtDNA)
                    • Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays
                    • Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays
                        • Sample Preparation
                        • Absorbance Measurements
                        • SRM Unit Production
                        • Component Homogeneity
                          • Figure 4 Plate Layout Design for Homogeneity Measurements
                            • Component Stability
                              • Figure 5 Plate Layout Design for Stability Measurements
                                • Component Certification Measurements
                                  • Table 6 Tubes used for Certification Measurements
                                  • Figure 6 Basic Plate Layout Design for Certification Measurements
                                    • Mitochondrial DNA Measurements
                                      • Table 7 Run Parameters for Mitochondrial DNA
                                        • Performance in Commercial qPCR Chemistries
                                          • Table 8 Commercial qPCR Chemistries Tested with SRM 2372a
                                          • Figure 7 Plate Layout Design for Performance Assessments
                                              • Measurement Results
                                                • Absorbance Spectrophotometry
                                                  • Table 9 Absorbance at Selected Wavelengths
                                                  • Figure 8 Absorbance Spectra of SRM 2372a Components A B and C
                                                    • Droplet Volume
                                                    • Certification Measurements
                                                      • Homogeneity Results
                                                        • Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component
                                                        • Table 11 Within- and Between-Tube Relative Standard Deviations
                                                          • Stability Results
                                                            • Figure 9 ddPCR Stability Measurements as Function of Time
                                                            • Table 12 Stability Data as λacute Copy Number per Nanoliter of Component
                                                              • Ten-Assay Certification Results
                                                                • Figure 10 Comparison of Assay Performance
                                                                • Table 13 Certification Data as λacute Copy Number per Nanoliter of Component
                                                                  • Estimation of Copy Number Concentration
                                                                    • Table 14 Measured Copy Number Concentrations
                                                                      • Conversion of Copy Number to Mass Concentration
                                                                        • Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution
                                                                            • Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)
                                                                              • Figure 11 mtDNAnDNA for Components A B and C
                                                                              • Table 16 mtDNAnDNA Ratios
                                                                              • Table 17 mtDNAnDNA Summary Results
                                                                                • Performance in Commercial qPCR Chemistries
                                                                                  • Figure 12 qPCR Autosomal Quantitation Performance Summary
                                                                                  • Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary
                                                                                      • Qualification of ddPCR as Fit-For-Purpose
                                                                                        • Figure 14 Absorbance Spectra of SRM 2372 Components A and B
                                                                                        • Digital Polymerase Chain Reaction Assays
                                                                                        • Heat Treatments (Melting of the DNA into Single Strands)
                                                                                        • Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a
                                                                                          • Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a
                                                                                            • Quantitative Evaluation of ssDNA Content
                                                                                              • ddPCR Measurement of ssDNA Proportion
                                                                                              • Proof of Principle
                                                                                                • Table 19 Comparison of Snap-Cooling to Slow-Cooling
                                                                                                  • Time at Denaturation Temperature
                                                                                                    • Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures
                                                                                                      • Temperature
                                                                                                        • Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a
                                                                                                          • DNA Concentration in the ddPCR Solution
                                                                                                            • Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution
                                                                                                              • Equivalence of Assays
                                                                                                                • Table 23 Comparison of Assays
                                                                                                                    • ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials
                                                                                                                      • Table 24 Evaluation of Limiting Ratio
                                                                                                                          • Conclusions and Recommendations
                                                                                                                            • Recommended Certification Values
                                                                                                                              • Table 25 Recommended Values and 95 Uncertainties for Certification
                                                                                                                                • Use of SRM 2372a for Value-Assigning In-House Reference Materials
                                                                                                                                  • Certificate of Analysis
                                                                                                                                  • References
                                                                                                                                  • Appendix A DNA Extraction Method
                                                                                                                                    • Procedure
                                                                                                                                    • Reagents
                                                                                                                                    • Reference
                                                                                                                                      • Appendix B Spectrophotometric Measurement Protocol
                                                                                                                                        • Spectrophotometer Calibration
                                                                                                                                          • Establish Baseline
                                                                                                                                          • Enable Verification of Wavelength Calibration
                                                                                                                                          • Enable Verification of Absorbance Calibration
                                                                                                                                            • For ldquonativerdquo dsDNA Samples
                                                                                                                                              • Verify Optical Balance of Cuvettes
                                                                                                                                                • For NaOH denatured ssDNA Samples
                                                                                                                                                  • Prepare a 04 molL NaOH Stock Solution
                                                                                                                                                  • Prepare the Sample Solutions
                                                                                                                                                  • Measure Sample Solutions
                                                                                                                                                    • After All Sample Measurements are Complete
                                                                                                                                                      • Enable Verification of Spectrophotometer Stability
                                                                                                                                                      • Clean Up
                                                                                                                                                        • Evaluation
                                                                                                                                                        • Reference
                                                                                                                                                          • Appendix C SRM 2372a Statisticianrsquos report
                                                                                                                                                            • 1 Introduction
                                                                                                                                                            • 11 OpenBUGS Analysis
                                                                                                                                                            • 12 OpenBUGS Exemplar Output
                                                                                                                                                            • 2 Homogeneity assessment
                                                                                                                                                            • 21 Data
                                                                                                                                                            • 22 Model
                                                                                                                                                            • 23 OpenBUGS code and data
                                                                                                                                                              • Table C-1 Example OpenBUGS Homogeneity Summary
                                                                                                                                                                • 24 Results
                                                                                                                                                                  • Figure C-1 Homogeneity Assessment
                                                                                                                                                                    • 3 Certification measurements
                                                                                                                                                                    • 31 Data
                                                                                                                                                                    • 32 Model
                                                                                                                                                                    • 33 OpenBUGS code inits and data
                                                                                                                                                                      • Table C-2 Example OpenBUGS Certification Summary for Component A
                                                                                                                                                                      • Table C-3 Example OpenBUGS Certification Summary for Component B
                                                                                                                                                                      • Table C-4 Example OpenBUGS Certification Summary for Component C
                                                                                                                                                                        • 34 Results
                                                                                                                                                                          • Table C-5 Parameter Values for Components A B and C
                                                                                                                                                                          • Table C-6 Certified Values for Components A B and C
                                                                                                                                                                            • 4 Mito-to-nuclear DNA ratios (mtDNAnDNA)
                                                                                                                                                                            • 41 Data
                                                                                                                                                                            • 42 Model
                                                                                                                                                                            • 43 OpenBUGS code inits and data
                                                                                                                                                                              • Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A
                                                                                                                                                                              • Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B
                                                                                                                                                                              • Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C
                                                                                                                                                                              • Table C-10 Results for mtDNAnDNA
                                                                                                                                                                                • 5 References
                                                                                                                                                                                  • Appendix D Performance in Commercial qPCR Chemistries
                                                                                                                                                                                    • Table D-1 Summary [DNA] Results for the Commercial qPCR Assays
Page 10: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

2

History and Background SRM 2372a is the second member of the SRM 2372 Human DNA Quantitation Standard series The supply of the initial material SRM 2372 was exhausted in May 2017 SRM 2372a is derived from commercially obtained buffy coat cells from different donors than those used in SRM 2372 A series of interlaboratory studies coordinated by what was then the Biotechnology Division of the National Institute of Standards and Technology (NIST) during the 1990s demonstrated that the DNA quantitation technologies then in use by the forensic human identification community adequately met the needs of the DNA typing tools then available34 However our 2004 study demonstrated that the increasingly sensitive and informative polymerase chain reaction (PCR)-based multiplex typing tools required more reliable determinations of DNA input quantity for best performance5 Quantitative PCR (qPCR) methods for DNA quantitation require calibration to an external standard of defined DNA concentration ([DNA]) expressed in units of nanograms of DNA per microliter of extract solution The true [DNA] of a commercial secondary standard may vary from its nominal concentration depending on supplier and lot number To enable providers of secondary standards to more reliably evaluate their materials and forensic testing laboratories to validate their use of commercial quantitation methods and supplies in 2007 we delivered the human DNA calibration standard SRM 2372 Human DNA Quantitation Standard6 The quantity of DNA in each of the three components of SRM 2372 was originally assigned based on the absorbance of double-stranded DNA (dsDNA) at 260 nm7 However by 2012 the absorbance properties of the materials had increased beyond the certified uncertainty limits due to slow but progressive changes in DNA tertiary structure Sales of the SRM were restricted until the DNA quantities were reassigned based on measurements of single-stranded DNA (ssDNA) in early 20138-10 Figure 1 displays the sales history of this relatively high-selling SRM While SRM 2372 was fit for the practical purpose of harmonizing DNA quantitation measurements to an accessible and reproducible reference material the conversion of both dsDNA and ssDNA absorbance to mass [DNA] is through conventional proportionality constants that are not metrologically traceable to the modern reference system for measurements the International System of Units (SI) Responding to requests from the forensic and clinical communities for calibration standards that provide traceability to the SI we and others have established droplet digital polymerase chain reaction (ddPCR) as a potentially primary measurement process that can provide traceable results11-18 The copy numbers of human nuclear DNA (nDNA) in the reaction mixture per partition symbolized as λ were determined by ddPCR using ten PCR assays optimized for use with our ddPCR platform These assays target loci on eight chromosomes with three assays on widely separated locations on one chromosome (see Table 1)

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

3

Figure 1 Sales History of SRM 2372

Each small black dot represents the sale of one unit of SRM 2372 The red lines summarize the average rate of sales as originally certified (2007 to 2012) and after re-certification (2013 to 2017) The magenta labels state the average yearly sales rate during these two periods

The nDNA copy number concentration human haploid genome equivalents (HHGE) per nanoliter component solution and symbolized here as λacute is calculated as

copy number [DNA] = λ = λ(VF) HHGEnL

where V is the average droplet volume and F is the total dilution factor the volume fraction of the component solution in the ddPCR reaction mixture The manufacturer of the ddPCR platform used a Bio-Rad QX200 Droplet Digital PCR System (Hercules CA) system does not provide metrologically traceable droplet volume information This lack of information prevented metrologically traceable conversion of the number per partition measurements to the desired number per reaction volume units In consequence NIST developed a measurement method that provides metrologically traceable droplet volumes19 We show elsewhere that the mass [DNA] nanograms of human DNA per microliter component can be traceably estimated as

mass [DNA] = 3031timesλacute ngmicroL

where the constant is the product of the number of nucleotide basepairs (bp) in the human reference genome the average bp molecular weight Avogadrorsquos number of bp per mole and

2372

2372(Recertif ied)

050

010

00

1713 unitsyear

1416 unitsyear

2007 2009 2011 2013 2015 2017

Uni

ts S

old

Calendar Year

Human DNA Quantitation StandardSRM 2372

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

4

unit scaling factors18 The three components of SRM 2372a are certified for both copy number and mass [DNA] Mitochondrial DNA (mtDNA) analysis within the forensic community is useful when the nDNA may be degraded or is in low-to-undetectable quantity The challenges of mtDNA testing include contamination of the standard due to routine laboratory use degradation of plasmids or synthetic oligomers or improper commutability of the standard relative to casework samples Commonly cell line materials are used as standards for qPCR with human genomic DNA which may lead to an incorrect mtDNA-to-nDNA ratio (mtDNAnDNA) for unknown samples Non-certified values of mtDNAnDNA for the three components are provided to bridge the current gap between well characterized mitochondrial DNA quantification assays and the availability of commercial standards With the understanding that the forensic DNA community primarily uses qPCR as a method for DNA quantification the performance of the three SRM 2372a materials was evaluated in eleven commercial qPCR chemistries currently used employed within the forensic DNA community

Experimental Methods Production and value-assignment of any certified reference material (CRM) is a complex process In addition to acquiring processing and packaging the materials included in the CRM the materials must be sufficiently homogenous across the units stable over time within its packaging and well-characterized to deliver results that are fit-for-purpose within the intended measurement community20 The process has been particularly complicated for SRM 2372a because we needed to determine the metrological properties of the required ddPCR measurement systems before we could trust the results they provide The following sections describe the methods and processes used to produce and characterize the SRM 2372a materials including sample preparation SRM unit production PCR assays ddPCR measurement processes spectrophotometric absorbance measurements homogeneity testing stability testing ddPCR nDNA and mtDNAnDNA measurements and performance in commercial qPCR assessments Many of these processes were mutually dependent therefore the sections are presented in a ldquodefinitionalrdquo order that sequentially introduces terminology Polymerase Chain Reaction Assays

Ten PCR assays were developed for characterization of SRM 2372a Each assay was confirmed to target one locus per HHGE using the National Center for Biotechnology Informationrsquos BLASTblastn system21 Primers were purchased from Eurofins Operon (Huntsville AL) TaqMan probes labeled with 6-carboxyfluorescein (FAM) were purchased from Thermo Fisher (Waltham MA) FAM-labeled Blackhole Quencher and Blackhole Quencher+ probes were purchased from LGC Biosearch Technologies (Novato CA) Table 1 details the ten nDNA assays used to certify copy number and mass [DNA] of the SRM 2372a components FAM-labeled TaqMan probes were purchased from Thermo Fisher

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

5

and used in monoplex reactions VIC (a proprietary dye)-labeled probes were purchased from the same source and were used when nDNA and mtDNA were analyzed in the same reaction mixture Figure 2 displays exemplar droplet patterns for the assays All assays were developed at NIST and have been optimized for ddPCR18 Table 2 details three mitochondrial DNA (mtDNA) assays used to estimate mtDNAnDNA in the SRM 2372a components Figure 3 displays exemplar droplet patterns for the assays These assays were optimized for ddPCR from real-time quantitative PCR (qPCR) assays from the literature All primers probes were diluted with TE-4 buffer purchased from Affymetrix-USB Products (Affymetrix Inc Cleveland OH) to a nominal 5 micromol working solution for all single-plex assays (individual nDNA and mtDNA assays) For duplex assays (combined nDNA and mtDNA) primer and probe working solutions were diluted to a nominal 10 micromol

Table 1 NIST-Developed Human Nuclear DNA Assays

Assay Target

Chromosome Band Accession Primers and Probe a

Amplicon Length bp

NEIF Gene EIF5B

Chr 2 p111-q111 NC_00000212

F gccaaacttcagccttctcttc R ctctggcaacatttcacactaca PB+ tcatgcagttgtcagaagctg

67

2PR4 Gene RPS27A

Chr 2 p16 NC_00000212

F cgggtttgggttcaggtctt R tgctacaatgaaaacattcagaagtct PB tttgtctaccacttgcaaagctggccttt

97

POTP STR TPOX

Chr 2 p253 NC_00000212

F ccaccttcctctgcttcacttt R acatgggtttttgcctttgg PT caccaactgaaatatg

60

NR4Q Gene DCK

Chr 4 q133-q211 NC_00000412

F tggtgggaatgttcttcagatga R tcgactgagacaggcatatgtt PB+ tgtatgagaaacctgaacgatggt

83

D5 STR D5S2500

Chr 5 q112 NC_00000510

F ttcatacaggcaagcaatgcat R cttaaagggtaaatgtttgcagtaatagat PT ataatatcagggtaaacaggg

75

ND6 STR D6S474

Chr 6 q21-22 NC_00000612

F gcatggctgagtctaaagttcaaag R gcagcctcagggttctcaa PB+ cccagaaccaaggaagatggt

82

D9 STR D9S2157

Chr 9 q342 NC_00000912

F ggctttgctgggtactgctt R ggaccacagcacatcagtcact PT cagggcacatgaat

60

HBB1 Gene HBB

Chr 11 p155 NC_00001110

F gctgagggtttgaagtccaactc R ggtctaagtgatgacagccgtacct PT agccagtgccagaagagccaagga

76

ND14 STR D14S1434

Chr 14 q3213C NC_0000149

F tccaccactgggttctatagttc R ggctgggaagtcccacaatc PB+ tcagactgaatcacaccatcag

109

22C3 Gene PMM1

Chr 22 q132 NC_00002210

F cccctaagaggtctgttgtgttg R aggtctggtggcttctccaat PB caaatcacctgaggtcaaggccagaaca

78

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

6

a F Forward primer R Reverse primer PB Blackhole quencher probe (FAM labeled) PB+ Blackhole Plus quencher probe (FAM labeled) PT TaqMan MGB probe (FAM labeled)

Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays

Table 2 NIST-Optimized Human Mitochondrial DNA Assays

Assay Location Primers and Probe a Amplicon Length bp Ref

mtND1 34853504 F ccctaaaacccgccacatct

69 22 35333553 R gagcgatggtgagagctaaggt 35063522 PT ccatcaccctctacatc

mtBatz 84468473 F aatattaaacacaaactaccacctacct

79 23 85038524 R tggttctcagggtttgttataa 84758491 PT cctcaccaaagcccata

mtKav 1318813308 F ggcatcaaccaaccacaccta

105 24 1336913392 R attgttaaggttgtggatgatgga 1331013325 PT cattcctgcacatctg

a F Forward primer R Reverse primer PT TaqMan MGB probe (FAM labeled)

2PR4

POTP

NEI

F

R4Q

5

D5

ND6

D9

HBB1

ND1

4

22C3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

7

Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays

Droplet Digital Polymerase Chain Reaction Measurements

Nuclear DNA (nDNA) All nDNA ddPCR measurements used in this study followed the same measurement protocol

bull Droplets were generated on the Auto Droplet Generator using the Bio-Rad ldquoddPCR ldquoSupermix for Probes (No dUTP)rdquo (Bio-Rad cat 186-3024 Control 64079080) and ldquoDroplet Generation Oil for Probesrdquo (Bio-Rad cat 186-4110 Control 64046051) Note desoxyuridine 5-triphosphatase (dUTP) is an enzyme sometimes included in PCR reaction mixtures to minimize PCR carryover contamination

bull Table 3 details the composition of the ddPCR mastermix setup for a given assay and sample where ldquomastermix refers to reaction mixture of manufacturerrsquos proprietary PCR reagent ldquoSupermixrdquo forward and reverse PCR primers for a given assay probe for that assay the sample DNA and sometimes additional water to produce the desired value dilution factor F This setup was used for all assays and samples

bull Non-Template Controls (NTCs) for each assay were included in every analysis bull Droplets were thermal cycled on a ProFlex thermal cycler (Applied Biosystems) for

95 degC for 10 min followed by 60 cycles of 94 degC for 05 min and 61 degC for 1 min then 98 degC for 10 min before a 4 degC hold until the plate was removed

bull Droplets were read on the QX200 Droplet Reader and analysis was performed using the QuantaSoft Analysis Software version 1740917

mtND1 mtBatz mtKav

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

8

bull The numbers of positive and negative droplets at the end of 60 cycles were determined and were exported into a spreadsheet for further analysis Assay-specific intensity thresholds were determined by visual inspection for each assay for each measurement session

Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays

Concentration nDNA Mastermix Microliter

per Reaction 2X ldquoSupermix for Probes (no dUTPs)rdquo 125

5 μmolL Forward Primer 1875 5 μmolL Reverse Primer 1875 5 μmolL Probe (FAM) 125

PCR Grade Water 50 14 Diluted Sample 25 Total Volume 250

Mitochondrial DNA (mtDNA) All the mtDNA measurements used in this study followed the nDNA protocol described above but with different mastermix setups The duplex mtDNA measurements used the setup described in Table 4 The monoplex mtDNA measurements used the setup described in Table 5

Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 125

10 μmolL Forward Primer 09375 10 μmolL Reverse Primer 09375 10 μmolL Probe (VIC) 0625

PCR Grade Water 95 nDNA Mastermix from Table 3 05 Total Volume 250

Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 120

10 μmolL Forward Primer 09 10 μmolL Reverse Primer 09 10 μmolL Probe (FAM) 06

PCR Grade Water 72 Diluted Sample (See Table 7) 24

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

9

Total Volume 240 Sample Preparation

The buffy coat white blood cells used to prepare SRM 2372a materials were purchased from Interstate Blood Bank These materials were from four single-source anonymous individuals supplied to us labeled as Female1 Female2 Male1 and Male2 In our laboratories each donor buffy coat bag was aliquoted (5 mL per aliquot) into sterile 50 mL conical tubes and stored at 4 degC prior to extraction The modified salt-out manual extraction protocol described in Appendix A was performed for each aliquot with rehydration of the DNA in TE-4 buffer From 150 mL to 250 mL of solubilized DNA for each component was accumulated through the extraction of the 5 mL aliquots The DNA was stored in perfluoroalkoxy polymer (PFA) containers that had been bleach sterilized thoroughly rinsed with deionized water rinsed with ethanol and air-dried The minimum volume for production of 2000 units was 120 mL Preliminary estimates of the mass [DNA] of the re-solubilized components were obtained using a Nanodrop microvolume spectrophotometer Dilutions of the component materials were made until the absorbance at 260 nm was approximately 1 corresponding to a nominal mass [DNA] of 50 ngμL of dsDNA7 After addition of the TE-4 buffer the containers were placed on an orbital shaker and gently rotated for several hours prior to refrigerated storage at 4 ordmC The four single-source components were initially screened with the nDNA ddPCR measurement process using four nDNA assays 2PR4 NR4Q ND6 and 22C3 Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate The gravimetric 1rarr4 dilution was prepared by weighing an empty PFA vial and recording the weight then pipetting 30 microL TE-4 into the vial and recording the weight then pipetting 10 microL DNA and recording the final weight The dilution factor was calculated in a spreadsheet These screening measurements verified that the individual component materials were fit-for-purpose The four single-source components were evaluated with the nDNA ddPCR measurement process using the ten nDNA assays Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate Quantitative information was derived from the average of the ten nDNA assays to prepare the component C mixture Thirty milliliters of sample Male2 (estimated at 452 ngμL mass [DNA]) were combined with 80 mL of sample Female2 (estimated at 507 ngμL mass [DNA]) and 10 mL of TE-4 buffer to create a 1 male nDNA to 3 female nDNA producing a 14 malehuman mixture After addition of the TE-4 buffer the component C container was gently rotated on the orbital shaker for several hours prior to refrigerated storage at 4 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

10

Absorbance Measurements

In February 2017 absorbance spectra of the recently prepared fully diluted nominally dsDNA and ssDNA solutions of the three components were acquired with a BioCary 100 spectrophotometer using the measurement protocol described in Appendix B In December 2017 additional spectra were acquired for the remaining bulk solution of components A and B and for SRM 2372 components A and B These solutions had been stored at 4 degC in well-sealed PFA containers There was no remaining bulk component C solution SRM Unit Production

In the morning of the day a component was transferred into the sterile component tubes (Sarstedt Biosphere SC micro tube-05 mL Newton NC) used to package the SRM the container holding the diluted component was removed from the refrigerator and placed on a slow orbital shaker for 2 h The tubes were opened and placed under the laminar flow hood in rows of 5 within 80-hole tube racks for a total of 30 tubes per rack Tubes were placed in every second row of the racks to facilitate filling with an Eppendorf Repeater Xstream pipette (Eppendorf North America Inc Hauppauge NY) fitted with a 1-mL positive displacement tip set to dispense 55 microL per tube The component container was removed from the shaker a magnetic stir-bar was added the container was placed on a magnetic stir plate in the laminar flow hood and the material was stirred gently Filling proceeded until 2200 vials were filled with pipet tip replacement after filling every 200 tubes The filling process consisted of two individuals manually filling each vial within a biosafety cabinet (one physically pipetting each sample the other verifying that each tube was filled) and four individuals tightly closing the lids to each tube on a sterilized and covered bench When all tubes in a rack were filled the rack was moved out of the hood the tubes checked for proper filling their lids closed and the tubes transferred to 100-unit pre-labeled storage boxes As the 100-unit boxes were filled they were transferred to the labeling room where SRM labels were applied by hand and colored inserts were added to the lids red for component A white for component B and blue for component C After all tubes for a given component were labeled the 22 boxes were placed in a locked refrigerator Filling of each of the components took 60 min to 90 min All units were transferred to the refrigerator within 4 hours of beginning the filling process All units of the three materials were allowed to equilibrate for a minimum of two weeks prior to beginning homogeneity analysis Component Homogeneity

The ddPCR measurement protocol described above was used to assess the nDNA homogeneity of the three components One tube from the approximate center of each of the 22 storage boxes was assayed using the NEIF and ND6 assays (see Table 1) A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Three technical replicates were obtained for every tube A total

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

11

of six plates were used two plates per component one with the NEIF assay and one with the ND6 assay Figure 4 details the layout of the samples

Figure 4 Plate Layout Design for Homogeneity Measurements

Component Stability

Stability of the materials was evaluated at 4 degC room temperature (RT asymp 22 degC) and 37 degC Tubes from the homogeneity study were divided among the three temperatures for the stability study Additionally two never-opened tubes (one from box 8 and one from box 12 of each component) were held at each temperature Two tubes at each temperature for each component were opened sampled and resealed 4 times over the course of 8 weeks Measurements were made at week 1 week 3 week 5 and week 8 The measurement on week 8 was from an unopened tube (from box 8) for each of the components at all temperatures The data from the homogeneity study was used as timepoint 0 in the stability study anchored to 4 degC by the storage recommendations A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used for both the NEIF and ND6 assays All temperatures components and both assays fit onto one plate for timepoints 1 to 5 Twelve NTCs placed in rows G and H were included for each assay Figure 5 details the sample layout

Figure 5 Plate Layout Design for Stability Measurements

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

ND6 ND6 ND6

NEIFNEIFNEIFComponent A

Component A

Component B

Component B

Component C

Component C

1 2 3 4 5 6 7 8 9 10 11 12A A-7 A-7 B-7 B-7 C-7 C-7 A-7 A-7 B-7 B-7 C-7 C-7B A-10 A-10 B-10 B-10 C-10 C-10 A-10 A-10 B-10 B-10 C-10 C-10C A-9 A-9 B-9 B-9 C-9 C-9 A-9 A-9 B-9 B-9 C-9 C-9D A-12 A-12 B-12 B-12 C-12 C-12 A-12 A-12 B-12 B-12 C-12 C-12E A-8 A-8 B-8 B-8 C-8 C-8 A-8 A-8 B-8 B-8 C-8 C-8F A-11 A-11 B-11 B-11 C-11 C-11 A-11 A-11 B-11 B-11 C-11 C-11G NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTCH NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

NEIF

4C

RT

37C

ND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

12

Component Certification Measurements

After establishing homogeneity and stability measurements intended to be used to certify nDNA values and inform mtDNA values were started using randomly selected SRM tubes from the boxes listed in Table 7 The ten nDNA assays described in Table 1 and two of the three mtDNA assays described in Table 2 were run simultaneously on the same sample plate A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used in each of the six measurement campaigns

Table 6 Tubes used for Certification Measurements

Component Plate A B C

dd20170711 Box 1 Box 1 Box 1 dd20170712a Box 10 tube a Box 10 tube a Box 10 tube a dd20170712b Box 10 tube b Box 10 tube b Box 10 tube b dd20170713 Box 20 Box 20 Box 20 dd20170714 Box 5 Box 5 Box 5 dd20170801 Box 5 Box 5 Box 5

Every sample was analyzed in duplicate with every assay in all six plates Component A was always located in rows B and C component B in rows D and E component C in rows F and G and NTCs in rows A and H Figure 6 displays the basic layout of the ten nDNA and two mtDNA assays In addition to the plate column assignment scheme shown the three assay groupings were also ordered as Group 2 Group 3 Group 1 and Group 3 Group 1 Group 2 to address potential positional bias

Group 1 Group 2 Group 3 2PR4 POTP NEIF NR4Q D5 ND6 D9 HBB1 ND14 22C3 mtND1 mtBatz 1 2 3 4 5 6 7 8 9 10 11 12 A NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Comp A B Comp A C Comp B D Comp B E Comp C F Comp C G

H NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Figure 6 Basic Plate Layout Design for Certification Measurements

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

13

Mitochondrial DNA Measurements

A ldquoduplexed reactionrdquo was used to determine mtDNAnDNA where 1 μL of one of the nDNA mastermix was added to 49 μL of the mtDNA mastermix resulting in a 1rarr50 volumetric dilution of the nDNA The nDNA was targeted with a FAM-labeled probe while the mtDNA was targeted with a VIC labeled probe The primer and probe concentrations were increased from 5 micromol to 10 micromol in the duplexed reactions Additional testing of the mtDNA assays were performed in monoplex with FAM-labeled probes to determine the additional sources of bias between sampling assay setup and dilution factors Table 7 outlines the different methods examined to determine mtDNAnDNA for the three components

Table 7 Run Parameters for Mitochondrial DNA Droplet

Generation nDNA mtDNA

Plate Component Assay Dilution Reps Assay Dilution Reps dd20170824 A manualDG HBB1 1rarr4 3 mtND1 1rarr200 9 dd20170825 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170829 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170831 AB autoDG HBB1 1rarr3 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr6 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr12 2 mtND1 1rarr10rarr100 3 dd20170907 ABC manualDG HBB1 1rarr3 2 mtBatz 1rarr10rarr100 3 dd20170913 AB manualDG NEIF 1rarr4 3 mtND1 mtBatz mtKav 1rarr10rarr100 3 dd20170913 C manualDG NEIF 1rarr4 3 mtBatz mtKav 1rarr10rarr100 3 Performance in Commercial qPCR Chemistries

Table 8 lists the 11 qPCR chemistries that were commercially available in late 2017 All three of the SRM 2372a components were evaluated with these chemistries using component A of SRM 2372 as the reference standard For each component of SRM 2372a four tubes were pulled from box 22 and combined into one tube A 1rarr5 dilution of all samples was prepared followed by three serial 1rarr2 dilutions A five-point serial dilution of each SRM 2372a component was amplified in triplicate for each chemistry This consisted of the neat material the 1rarr5 dilution then the three serial 1rarr2 dilutions A six-point serial dilution of SRM 2372 Component A was also created for the standard curve for the commercial qPCR chemistries This consisted of the neat material followed by a 1rarr5 dilution then followed by four 1rarr2 dilutions serially The six-point serial dilution series of SRM 2372 Component A was also amplified in triplicate These commercial chemistries were amplified in individual plates on an Applied Biosystems 7500 HID Real-Time PCR Instrument (Foster City CA) following the manufacturersrsquo recommended protocols for reaction mix setup and cycling conditions Figure 7 details the plate layout for the qPCR plates Data was analyzed with the HID Real-Time PCR Analysis Software v12 and further analyzed in a spreadsheet

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

14

Table 8 Commercial qPCR Chemistries Tested with SRM 2372a

Manufacturer Commercial Kit Target Codea

Innogenomics

InnoQuant Long IQh Short IQd

InnoQuantHY Long IQHYh Short IQHYd Y IQHYy

Thermo Fisher Quantifiler Duo Human QFDh Male QFDy

Quantifiler HP Large QFHPh Small QFHPd

Quantifiler Human Human QFHh

Quantifiler Trio Large QFTh Small QFTd Y QFTy

Qiagen

Quantiplex Autosomal QPh

Quantiplex Hyres Autosomal QPHh Y QPHy

Quantiplex Pro Human QPPh Degradation QPPd Y QPPy

Promega

Plexor Autosomal Ph Y Py

PowerQuant Autosomal PQh Degradation PQd Y PQy

a Assays are coded with a short acronym derived from the kit name followed by the following codes for the type of target ldquohrdquo autosomal non-degradation described as ldquoAutosomalrdquo ldquoHumanrdquo or ldquoLongrdquo ldquodrdquo autosomal degradation described as ldquoDegradationrdquo or ldquoShortrdquo ldquoyrdquo Y-chromosome described as ldquoMalerdquo or ldquoYrdquo

1 2 3 4 5 6 7 8 9 10 11 12

A

B 2372-A_57 2372-A_57 2372-A_57 2372a-neat_A

2372a-neat_A

2372a-neat_A

2372a-neat_B

2372a-neat_B

2372a-neat_B

2372a-neat_C

2372a-neat_C

2372a-neat_C

C 2372-A_11

4 2372-A_11

4 2372-A_11

4 2372a-15_A

2372a-15_A

2372a-15_A

2372a-15_B

2372a-15_B

2372a-15_B

2372a-15_C

2372a-15_C

2372a-15_C

D 2372-A_5

7 2372-A_5

7 2372-A_5

7 2372a-110_A

2372a-110_A

2372a-110_A

2372a-110_B

2372a-110_B

2372a-110_B

2372a-110_C

2372a-110_C

2372a-110_C

E 2372-A_2

85 2372-A_2

85 2372-A_2

85 2372a-120_A

2372a-120_A

2372a-120_A

2372a-120_B

2372a-120_B

2372a-120_B

2372a-120_C

2372a-120_C

2372a-120_C

F 2372-A_1

425 2372-A_1

425 2372-A_1

425 2372a-140_A

2372a-140_A

2372a-140_A

2372a-140_B

2372a-140_B

2372a-140_B

2372a-140_C

2372a-140_C

2372a-140_C

G 2372-A_0

7125 2372-A_0

7125 2372-A_0

7125

H NTC NTC NTC

Figure 7 Plate Layout Design for Performance Assessments

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

15

Measurement Results After preparing the SRM 2372a bulk solutions their approximate 50 ngmicroL mass [DNA] were confirmed spectrophotometrically The bulk solutions were then packaged into individual tubes The materials contained in the tubes were characterized for homogeneity stability nDNA copy number [DNA] and mtDNAnDNA ratio The following sections detail the results of these measurements Absorbance Spectrophotometry

By widely accepted convention an aqueous DNA solution having an absorbance of 10 at 260 nm corresponds to a mass [DNA] of 50 ngmicroL for dsDNA and (37 to 40) ngmicroL for ssDNA7 Absorbance at (230 270 280 and 330) nm are traditionally used in the assessment of DNA quality725 Due to minor baseline shifts during acquisition of some spectra it was necessary to bias-adjusted all spectra to have absorbance of 00 at 330 nm Table 9 lists the resulting bias-adjusted values for selected wavelengths and the conventional mass [DNA] conversion of the absorbance at 260 nm The close agreement between the mass concentration values for the nominally dsDNA and the converted-to-ssDNA spectra confirm that the bulk solutions were correctly prepared to be asymp 50 ngμL dsDNA The slight increase in absorbance over the 10 months between acquiring the spectra suggests that like SRM 2372 the tertiary structure of the dsDNA in the SRM 2372a materials may be changing slowly with time Figure 8 displays the spectra

Table 9 Absorbance at Selected Wavelengths

Wavelength nm Form Date Component 230 260 270 280 ngμL dsDNA 2142017 A 0394 0931 0758 0488 466

B 0480 1112 0905 0583 556 C 0451 1006 0815 0524 503

ssDNA 2142017 A 0390 0614 0540 0316 454 B 0457 0723 0634 0372 535 C 0411 0641 0563 0330 474

dsDNA 12142017 A 0402 0956 0775 0503 478 B 0484 1128 0914 0589 564

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

16

Figure 8 Absorbance Spectra of SRM 2372a Components A B and C

dsDNA is denatured to ssDNA by diluting with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to DNA mass concentration

Droplet Volume

Droplet volumes for the two lots of ldquoSupermixrdquo used in the studies described in this report were determined using NISTrsquos Special Test Method 11050S-D as described in Dagata et al19 The lot used for most of the ddPCR analyses produced (07349 plusmn 00085) nL droplets where the ldquoplusmnrdquo value is the combined standard uncertainty The lot used for analyses performed after October 11 2017 produced (07376 plusmn 00085) nL droplets NIST measurements have demonstrated that droplet volumes are not influenced by the presence or absence of DNA in the sample solution Measurements repeated over time on an earlier lot of the Supermix product used in these studies suggest that droplet volumes remain constant over the shelf-life of the lot However droplet volumes for different lots of this product as well as for different Supermix products have been observed to differ by several percent We intend to document these measurements in a future report Other researchers have demonstrated that non-constant droplet volume distributions may contribute to ddPCR measurement imprecision and bias2627 However these effects become significant only with distributions wider than observed using Special Test Method 11050S-D19 and at copydroplet values larger than the λ le 05 copydroplet typically used in our measurements Certification Measurements

Homogeneity Results Table 10 lists the λacute nDNA copy number values adjusted for both droplet volume and sample dilution from the homogeneity measurements The data was analyzed using the three-level Gaussian hierarchical ANOVA model described in Appendix C Table 11 lists the values of the within- and between-tube variance components expressed as percent relative standard deviations At the 95 confidence level all components are homogeneous

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372a-A2141721417-NaOH121417

220 260 300 340

Wavelength nm

2372a-B2141721417-NaOH121517

220 260 300 340

Wavelength nm

2372a-C21417

21417-NaOH

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

17

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 NEIF A-1-01 1632 1529 1511 B-1-01 1790 1909 1885 C-1-01 1493 1385 1510 ND6 1616 1640 1582 NAe 1870 1860 1565 1562 1466

NEIF A-1-02 1531 1525 1549 B-1-02 1712 1719 1704 C-1-02 1516 1513 1577 ND6 1615 1589 1601 1775 1718 1756 1490 1547 1570

NEIF A-1-03 1524 1540 1533 B-1-03 1693 1755 1749 C-1-03 1451 1528 1506 ND6 1582 1152 1555 1824 1483 1896 1485 1505 1471

NEIF A-1-04 1454 1458 1427 B-1-04 1754 1745 1739 C-1-04 1440 1486 1520 ND6 1627 1605 1607 1670 1632 1814 1529 1508 1506

NEIF A-1-05 1465 1540 1504 B-1-05 1757 1687 1735 C-1-05 1523 1472 1483 ND6 1607 1633 1586 1683 1759 1703 1513 1520 1491

NEIF A-1-06 1512 1491 1491 B-1-06 1697 1709 1714 C-1-06 1437 1493 1522 ND6 1511 1483 1524 1693 1723 1722 1492 1481 1461

NEIF A-1-07 1418 1480 1455 B-1-07 1722 1625 1703 C-1-07 1485 1477 1463 ND6 1479 1557 1656 1690 1684 1688 1475 1547 1485

NEIF A-1-08 1496 1560 1501 B-1-08 1884 1802 1772 C-1-08 1433 1405 1478 ND6 1514 1615 1540 1864 1887 1821 1497 1501 1495

NEIF A-1-09 1588 1581 1637 B-1-09 1713 1734 1727 C-1-09 1463 1499 1505 ND6 1474 1438 1454 1709 1756 1758 1503 1491 1516

NEIF A-1-10 1574 1513 1570 B-1-10 1759 1809 1738 C-1-10 1534 1458 1454 ND6 1594 1713 1669 1764 1816 1787 1448 1465 1474

NEIF A-1-11 1557 1508 1633 B-1-11 1749 1782 1740 C-1-11 1533 1514 1555 ND6 1571 1611 1631 1741 1792 1784 1488 1470 1483

NEIF A-1-12 1500 1476 1515 B-1-12 1748 1712 1718 C-1-12 1450 1525 1497 ND6 1626 1586 1608 1797 1614 1790 1496 1483 1519

NEIF A-1-13 1672 1634 1804 B-1-13 1632 1668 1701 C-1-13 1537 1527 1515 ND6 1490 1516 1497 1626 1566 1681 1499 1479 1475

NEIF A-1-14 1551 1568 1538 B-1-14 1752 1535 1673 C-1-14 1515 1490 1494 ND6 1617 1614 1639 1724 1741 1685 1469 1480 1466

NEIF A-1-15b 1793 1722 1733 B-1-15 1871 1835 1887 C-1-15 1511 1568 1594 ND6 1818 1753 1762 1770 1752 1798 1515 1543 1565

NEIF A-1-16 1573 1520 1617 B-1-16 1688 1725 1743 C-1-16 1449 1452 1471 ND6 1634 1627 1575 1744 1697 1710 1543 1448 1491

NEIF A-1-17 1516 1516 1628 B-1-17 1789 1769 1825 C-1-17 1388 1421 1394 ND6 1613 1567 1567 1716 1762 1638 1452 NAe 1435

NEIF A-1-18 1612 1535 1672 B-1-18 1826 1780 1849 C-1-18 1475 1455 1445 ND6 1529 1495 1528 1805 1801 1931 1458 1531 1459

NEIF A-1-19 1598 1514 1596 B-1-19 1751 1759 1880 C-1-19 1468 1525 1446 ND6 1571 1535 1511 1827 1766 1786 1435 1478 1421

NEIF A-1-20 1575 1568 1586 B-1-20 1738 1678 1768 C-1-20 1549 1531 1499 ND6 1652 1545 1627 1752 1747 1768 1499 1465 1468

NEIF A-1-21 1658 1611 1578 B-1-21 1742 1836 1713 C-1-21 1530 1440 1484 ND6 1647 1607 1612 1746 1744 1764 1499 1454 NAe

NEIF A-1-22 1547 1528 1532 B-1-22 1761 1743 1660 C-1-22 1442 1487 1436 ND6 1537 1513 1536 1666 1692 1819 1589 1591 1556

NEIF A-1-15c 1456 1448 1523 ND6 1525 1402 1436

NEIF A-2-15d 1679 1599 1674 ND6 1581 1550 1584

a Tube coded as Component-Tube-Box b These results were identified as a probable error in the preparation or volumetric delivery of the sample

solution into the microfluidic device and were not used in the homogeneity evaluations c Repeat assessment of the solution in sample A-1-15 d Results for a second tube from box 15 analyzed at the same time as the repeat assessment of A-1-15 e Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

18

Table 11 Within- and Between-Tube Relative Standard Deviations

Within-Tube Component NEIF ND6 Between-Tube

A 28 40 33 B 27 38 25 C 23 18 19

Stability Results Table 12 lists the λacute volume- and dilution-adjusted nDNA copy number values from the stability measurements Figure 9 summarizes these results The results for the tubes held at 4 degC room temperature (RT asymp 22 degC) and 37 degC are very similar for all three components The absence of between-temperature differences identifies the between-week differences as attributable to variation in the measurement processes rather than sample instability These results indicate that the SRM 2372a genomic copy number is thermally stable from 4 degC to 37 degC over an extended period

Figure 9 ddPCR Stability Measurements as Function of Time

Each dot-and-bar symbol represents the mean λacute and the approximate 95 expanded uncertainty for all results for a given component at a given storage temperature The thick black horizontal lines represent an approximate 95 confidence interval on the λacute values estimated from the homogeneity data

13

14

15

16

17

18

19

0 1 2 3 4 5 6 7 8

λacute H

aplo

id n

DNA

Copi

es μL

Weeks

A

0 1 2 3 4 5 6 7 8

Weeks

B

4 degC 22 degC 37 degC

0 1 2 3 4 5 6 7 8

Weeks

C

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

19

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Week Temp Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3

1 4 degC NEIF A-10 1568 1441 B-10 1666 1653 C-10 1424 1394 ND6 A-10 1474 1494 B-10 1655 1705 C-10 1461 1394

NEIF A-07 1500 1463 B-07 1718 1640 C-07 1446 1452 ND6 A-07 1479 1479 B-07 1708 1686 C-07 1522 1482

RT (22 degC) NEIF A-12 1674 1608 B-12 1805 1777 C-12 1506 1495 ND6 A-12 1522 1507 B-12 1759 1788 C-12 1478 1461 NEIF A-09 1635 1539 B-09 1791 1793 C-09 1411 NAb ND6 A-09 1543 1519 B-09 1661 1682 C-09 1482 1496

37 degC NEIF A-11 1614 1637 B-11 1748 1744 C-11 1515 1464 ND6 A-11 1494 1502 B-11 1740 1697 C-11 1495 1535 NEIF A-08 1592 1614 B-08 1691 1725 C-08 1528 1500 ND6 A-08 1581 1595 B-08 1727 1699 C-08 1458 1483 3 4 degC NEIF A-10 1520 1706 B-10 NAb 1733 C-10 1427 1473 ND6 A-10 1543 1624 B-10 1670 1755 C-10 1443 1364 NEIF A-07 1537 1570 B-07 1549 1565 C-07 1424 1512 ND6 A-07 1517 1479 B-07 1705 1723 C-07 1483 1439

RT (22 degC) NEIF A-12 1521 1607 B-12 1817 1842 C-12 1566 1471 ND6 A-12 1644 1597 B-12 1845 1786 C-12 1504 1511 NEIF A-09 1432 1626 B-09 1825 1856 C-09 1392 1387 ND6 A-09 1539 1528 B-09 1734 1737 C-09 1535 1465

37 degC NEIF A-11 1651 1685 B-11 1759 1779 C-11 1437 1440 ND6 A-11 1680 1594 B-11 1806 1806 C-11 1471 1509 NEIF A-08 1580 1613 B-08 1740 1791 C-08 1582 1590 ND6 A-08 1592 1524 B-08 1769 1694 C-08 1468 1520 5 4 degC NEIF A-10 1486 1440 B-10 1626 1551 C-10 1381 1330 ND6 A-10 1555 1577 B-10 1673 1625 C-10 1547 1395 NEIF A-07 1387 1429 B-07 1723 1787 C-07 1330 1314 ND6 A-07 1538 1522 B-07 1862 1909 C-07 1452 1458

RT (22 degC) NEIF A-12 1601 1631 B-12 1718 1752 C-12 1546 1541 ND6 A-12 1699 1677 B-12 1772 1721 C-12 1490 1471 NEIF A-09 1381 1484 B-09 1675 1694 C-09 1371 1376 ND6 A-09 1625 1615 B-09 1904 1815 C-09 1468 1472

37 degC NEIF A-11 1456 1517 B-11 1720 1704 C-11 1505 1493 ND6 A-11 1650 1658 B-11 1148 1152 C-11 1527 1600 NEIF A-08 1590 1666 B-08 1745 1723 C-08 1469 1540 ND6 A-08 1235 1244 B-08 1809 1749 C-08 1508 1529 8 4 degC NEIF A-08a 1667 1632 1589 B-08a 1880 1884 1940 C-08a 1525 1475 1542 ND6 A-08a 1599 1593 1590 B-08a 1859 1897 1972 C-08a 1550 1502 1520

RT (22 degC) NEIF A-08b 1693 1666 1634 B-08b 1769 1884 1771 C-08b 1595 1557 1545 ND6 A-08b 1644 1620 1647 B-08b 1862 1738 1811 C-08b 1611 1585 1535

37 degC NEIF A-08c 1620 1576 1608 B-08c 1934 1952 1956 C-08c 1595 1600 1569 ND6 A-08c 1576 1553 1532 B-08c 1889 1844 1889 C-08c 1569 1514 1581

a Tube coded as Component-Tube-Box b Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

20

Ten-Assay Certification Results The original intention was to certify the copy number [DNA] of the three components with results from five independently prepared sets of samples analyzed over four days Each sample of the three components was from a new tube from different boxes and was evaluated with 10 nDNA assays The assay positions were alternated across these five ldquo10-assay certificationrdquo plates The λ results for these five plates are systematically somewhat larger than those for the homogeneity and stability data On review we realized that a different workspace and its suite of pipettes had been used to prepare the 1rarr10 dilution of the 1rarr4 diluted components into the reaction mastermix The ldquohighrdquo workstation was used for stability and homogeneity measurements while the ldquolowrdquo workstation was used for certification measurements Gravimetric testing of the pipettes established that the volumes delivered by these alternate pipettes was sufficiently biased to account for the observed λ differences A sixth ldquocertificationrdquo plate was prepared using the original (and now re-verified) pipettes The λ results or this plate agreed well with the homogeneity and stability data Table 13 lists the λacute volume- and dilution-adjusted nDNA copy number values for all six of the ten-assay datasets Figure 10 displays the extent of agreement among the ten assays While there are small-but-consistent differences between the assays there is only a 4 difference between the assays providing smallest and largest λ

Figure 10 Comparison of Assay Performance

The blue dot-and-bars represent the mean and its combined standard uncertainty for the A B and C components for each assay relative to the over-all grand mean for each component The black dot-and-bars represent combination of three component results for each assay The solid horizontal line represents the median difference of the assay means from the grand mean the dotted horizontal lines bound an approximate 95 confidence region about the median

2PR4POTPNEIFR4Q5 D5ND6 D9 HBB1ND14 22C3-10

-8

-6

-4

-2

0

2

4

6

8

10

Rela

tive

Diffe

renc

e fr

om M

ean

Assays

Component MeansAssay MeansMedianU95

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

21

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

7112017 2PR4 1667 1698 1869 1960 1630 1706 POTP 1655 1685 1850 1781 1596 1576 NEIF 1647 1665 1709 1707 1574 1538 R4Q5 1594 1658 1856 1787 1590 1555 D5 1655 1683 1848 1829 1610 1607 ND6 1629 1647 1828 1826 1589 1548 D9 1655 1652 1861 1858 1569 1572 HBB1 1644 1611 1825 1869 1565 1681 ND14 1645 1707 1871 1773 1564 1492 22C3 1545 1603 1776 1902 1548 1655

7122017 2PR4 1668 1739 2174 2231 1499 1497 POTP 1587 1702 2202 2292 1576 1611 NEIF 1678 1614 2172 2216 1545 1560 R4Q5 1628 1762 NA 2078 1584 1546 D5 1738 1649 2230 2192 1577 1529 ND6 1667 1678 2094 2034 1599 1478 D9 1695 1653 1949 2110 1587 1504 HBB1 1757 1750 1971 2123 1520 1551 ND14 1771 1764 2123 2066 1479 1438 22C3 1762 1759 2107 2046 1512 1494

7122017 2PR4 1638 1667 1810 1806 1452 1424 POTP 1676 1638 1865 1801 1347 1397 NEIF 1616 1624 1806 1786 1395 1350 R4Q5 1559 1587 1878 1768 1442 1459 D5 1608 1652 1824 1790 1448 1602 ND6 1611 1589 1768 1670 1463 1435 D9 1517 1607 1811 1749 1402 1433 HBB1 1557 1582 1763 1741 1474 1481 ND14 1604 1625 1865 1781 1466 1482 22C3 1675 1616 1815 1866 1439 1568

7132017 2PR4 1888 1939 1784 1821 1470 1444 POTP 1868 1815 1735 1813 1411 1427 NEIF 1953 1902 1778 1813 1512 1439 R4Q5 1779 1871 1835 1749 1452 1468 D5 1924 2124 1794 1848 1567 1524 ND6 1861 1787 1700 1711 1480 1424 D9 1824 1844 1799 1779 1442 1459 HBB1 1809 1799 1765 1769 1430 1467 ND14 1864 1857 1754 1725 1413 1411 22C3 1930 1942 1792 1814 1486 1480

7142017 2PR4 1597 1590 1718 1768 1485 1458 POTP 1533 1404 1753 1698 1422 1459 NEIF 1462 1533 1773 1792 1438 1465 R4Q5 1493 1519 1777 1740 1439 1468 D5 1538 1586 1803 1795 1515 1457 ND6 1492 1516 1718 1693 1409 1412 D9 1528 1538 1802 1743 1469 1474 HBB1 1548 1581 1834 1745 1453 1428 ND14 1513 1514 1731 1714 1465 1425 22C3 1562 NA 1842 NA 1363 1393

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

22

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

101217 2PR4 1431 1411 1791 1770 1357 1364 POTP 1407 1395 1746 1820 1447 1361 NEIF 1487 1398 1803 1803 1448 1404 R4Q5 1450 1439 1767 1739 1409 1411 D5 1394 1422 1745 1773 1389 1429 ND6 1396 1393 1824 1729 1415 1417 D9 1414 1390 1752 1744 1431 1380 HBB1 1401 1382 1813 1754 1381 1336 ND14 1412 1460 1648 1671 1403 1387 22C3 1435 1410 1818 1741 1435 1372

Estimation of Copy Number Concentration It was decided that all three sets of results (homogeneity stability and certification) would be used to derive the certified values with the λ of the first five ldquoten-assayrdquo plates adjusted by a data-driven estimate of the volumetric bias This bias was incorporated into the statistical model providing better estimates of the λ means while capturing all of the variance components (homogeneity stability between-assay and between data sets) as random effects using Bayesian MCMC programmed in OpenBUGS28 Table 14 lists the estimated values and their uncertainties See Appendix C for details

Table 14 Measured Copy Number Concentrations

Genomic Copies per Nanoliter of Solution Component Mean a u(Mean) b U95(Mean) c CV95

d A 151 07 14 93 B 175 03 06 34 C 145 05 10 69

a Mean the experimentally determined most probable estimate of the true value b u(Mean) the standard uncertainty of the Mean c Mean plusmnU95(Mean) the interval which is believed with about 95 confidence to contain the true value d CV95 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the counting unit one29 2) the validity of the Poisson endpoint transformation for digital PCR endpoint assays when applied to samples providing le 05 copies per droplet 3) the determination that the copies are dispersed as dsDNA and 4) calibrated mean droplet volume measurements made at NIST during sample dilution and mastermix preparation Conversion of Copy Number to Mass Concentration As described in Reference 18 metrologically valid conversion from copies per nanoliter to nanograms per microliter requires quantitative estimates for the number of nucleotide bases per reference human genome and the average nucleotide mass Using the values estimated in that report and asserting that each target copy corresponds to one HHGE the required transformation is

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

23

[DNA] ngmicroL

= [DNA] HHGE

nLtimes

3301 ngHHGE

The combined standard uncertainty is

119906119906 [DNA] ng

microL = 1199061199062 [DNA] HHGE

nL + 0531100

times[DNA] HHGE

nL 2

Both the measured and the conversion factor uncertainties are associated with ldquolargerdquo degrees of freedom Therefor the approximate 95 confidence uncertainty is estimated

11988011988095 [DNA] ng

microL = 2 times 119906119906

[DNA] ngmicroL

Table 15 lists the transformation of the Table 14 copy number values These values agree well with the conventional 260 nm absorbance values averaging (4 plusmn 5) larger

Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution

ngmicroL Component Value a u(Measurement)b u(Conversion)c U95(Value)d CV95

e A 498 23 03 47 93 B 578 10 03 21 36 C 479 17 03 33 70

a Estimated true value b Standard uncertainty component from the ddPCR measurements c Standard uncertainty estimated from the relative uncertainty of the conversion factor d Value plusmnU95(Value) is believed with about 95 confidence to contain the true value e 100U95(Value)Value the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the measured estimates of copies per nanoliter and 2) determinations made by us from internationally recognized literature resources of the average number of nucleotide bases per HHGE and the mean molecular weights of the bases18 Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)

While not intended to provide certified values mtDNAnDNA was measured as part of 13 experiments over 5 months Measurements were made using dilution-adjusted λ for the mtND1 mtBatz andor mtKav mitochondrial assays referenced against the dilution-adjusted λ of either the NEIF or HBB1 nDNA assays By chance both DNAs in the component C mixture have a binding site mutation that silences the mtND1 probe Figure 11 graphically summarizes the results by assay and component Table 16 lists the data Table 17 summarizes results from a random effects model that accounts for between-assay variability See Appendix C for details

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

24

Due to the limited number of mtDNA assays used in these measurements these results are metrologically traceable only to the assays used

Figure 11 mtDNAnDNA for Components A B and C

Each dot-and-bar symbol represents a mean mtDNAnDNA and its approximate 95 expanded uncertainty The dotted horizontal lines represent approximate 95 confidence intervals on the ratios

Table 16 mtDNAnDNA Ratios

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

20170711a 1 mtND1 1931 1870 2461 2401 1 mtBatz 1845 1764 2555 2608 3201 3299

20170711b 1 mtND1 1810 1906 2229 2200 1 mtBatz 1785 1837 2108 2192 3347 3287

20170712 1 mtND1 1743 1761 2127 2198 1 mtBatz 1829 1847 2455 2506 3217 3384

20170713 1 mtND1 1716 1800 2211 2199 1 mtBatz 1740 1813 2380 2249 3200 3363

20170714 1 mtND1 1923 1835 2105 1943 1 mtBatz 1935 1851 2136 2119 3294 3013

20170814 1 mtND1 1780 1776 1706 2046 2057 2024 1 mtBatz 1804 1751 1738 1998 2043 1952 2927 2776 2762 1 mtKav 1769 1786 1761 2077 2117 2041 3044 2830 2825

20170824 1 mtND1 1947 1928 1922 2 mtND1 1829 1895 1851 3 mtND1 1674 1617

20170825 1 mtBatz 1667 1813 1708 1598 1928 1916 1915 2832 2840 2788 2725 2 mtBatz 1566 1687 1663 1589 1947 1946 1961 1909 2835 2790 2736 2679 3 mtBatz 1643 1668 1604 1573 1953 1952 1941 1881 2803 2773 2757 2752

20170829 1 mtBatz 1719 1716 1670 1666 1865 1863 1828 1828 2599 2508 2551 2609 2 mtBatz 1581 1573 1601 1612 1944 1964 1947 1895 2617 2483 2626 2602 3 mtBatz 1546 1542 1538 1587 1908 1909 1940 1839 2665 2535 2570 2650

20170831 1 mtND1 1617 1599 1688 1783

ND1

ND1

ND1Ka

v

Kav

Kav

Batz

Batz

Batz

A

B

C

160

210

260

mtD

NAn

DNA

Ratio

Assay

Bind

ing s

ite m

utat

ion

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

25

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

2 mtND1 1622 1579 1823 1631 3 mtND1 1575 1528 1920 1887

20170907 1 mtBatz 1584 1544 1890 1886 2501 2522 20170913 1 mtND1 1757 1724 1768 1990 1921 1982

1 mtBatz 1743 1671 1766 1990 1924 1958 2805 2825 2829 1 mtKav 1692 1662 1767 1903 1831 1919 2709 2738 2727

20171016 1 mtND1 1924 1944 1893 2304 2341 2332 1 mtBatz 1786 1752 1743 2048 2078 2067 2797 2642 2792 1 mtKav 1530 1542 1523 2399 2479 2493 2804 2789 2842 2 mtBatz 1580 1557 1592 2715 2606 2687 2 mtKav 2103 2100 2049 2799 2628 2559 3 mtND1 1749 1769 1777 2065 2166 2164 3 mtBatz 1965 1889 1947 2314 2280 2258 2639 2535 2657 3 mtKav 1813 1864 1864 1867 1987 1971 2953 2757 2802

Table 17 mtDNAnDNA Summary Results

mtDNAnDNA Component Value a u(Value)b U95(Value)c CV95

d A 1737 16 38 17 B 2058 22 44 21 C 2791 23 47 17

a Estimated true value b Standard uncertainty c Value plusmnU95(Value) is believed with about 95 confidence to contain the true value d 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) Performance in Commercial qPCR Chemistries

Traditional qPCR assays only quantify relative to a reference standard Thus qPCR assays do not provide absolute [DNA] for the individual materials but they can provide [DNA] of one component relative to another Figure 12 summarizes the BA and CA ratios for the commercial autosomal (Human) targets from the eleven assays evaluated (see Table 8) The Y-chromosome (Male) assays cannot be summarized in this manner as the (all female) component B does not include a Y chromosome Figure 13 summaries the mass [DNA] results for the (all male) component A and (male-female mixture) component C Since this material was prepared as a nominal 13 malefemale mixture the expected malehuman (total) ratio is 14 sjc As displayed in these figures there is intrinsic variability between the SRM 2372a components and between the commercial qPCR chemistries examined This reflects the difference in the targets and amplification chemistries for each of the commercial kits Many of the commercial chemistries on the market to date contain multi-copy targets for both autosomal and Y-chromosomal quantification to increase the sensitivity of the assays

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

26

Figure 12 qPCR Autosomal Quantitation Performance Summary

Open circles denote the BA CA ratios the 17 commercial qPCR assays evaluated each labeled with the code listed in Table 8 The solid black circle represents ratios as estimated by ddPCR The red lines bound the approximate 95 confidence intervals on the ddPCR-estimated ratios

Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary

Open squares denote the ngμL values for SRM 2372a Components A (male) and C (mixture of male and female) for the seven commercial Y-chromosome qPCR assays evaluated each labeled with the code listed in Table 8 The dotted red lines represent malehuman ratios of 13 14 and 15

Appendix D provides a graphical analysis of all the data generated in the performance assessments for each of the targets amplified in the eleven commercial qPCR chemistries and the [DNA] values for all of the assays

dd

IQ_h

IQ_d

IQHY_h

IQHY_d

QFD_h

QFH_h QFHP_h

QFHP_dQFT_h

QFT_d

QPH_h

QPP_hQPP_d

P_h

PQ_h

PQ_d

06

08

10

12

14

07 09 11 13

CA

BA

IQHY_y

QFD_y

QFT_y

QPH_y

QPP_y

P_y

PQ_y

1413

15

10

12

14

16

18

30 40 50 60 70

[DNA

] C n

gμL

[DNA]A nguL

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

27

Qualification of ddPCR as Fit-For-Purpose We have documented the requirements for establishing metrological traceability of ddPCR human genomic assay results expressed as copies per nanoliter of sample in a series of three peer-reviewed papers151617 Through the experimental evidence they are based upon these papers demonstrate that ddPCR can accurately enumerate the number of independently migrating genomic copies in an aqueous extract Our most recent paper establishes the metrological traceability of these ddPCR results when transformed into the nanogram per microliter units used by the forensic and clinical communities18 However our previous studies have not addressed a fundamental requirement for correctly interpreting digital PCR measurements In broad simplification end-point assay systems count the number of independently partitioning entities regardless of the number of copies the entities contain With assays that target just one site per HHGE entities that migrate as ssDNA will yield counts that are twice the value had they migrated as dsDNA30 Should some originally dsDNA be converted over time to ssDNA ddPCR measurements of the number of independently migrating entities would increase although the number of HHGE would be unchanged Measurements of mass [DNA] provided by absorbance spectrophotometry and ddPCR reflect different properties absorbance at 260 nm is sensitive to the spatial configuration of the nucleotides as well as their number while ddPCR counts independently migrating entities The SRM 2372a materials have absorbance spectra shown in Figure 8 consistent with being predominantly dsDNA The SRM 2372 materials now have the absorbance spectra shown in Figure 14 which are consistent with them being entirely ssDNA

Figure 14 Absorbance Spectra of SRM 2372 Components A and B dsDNA is denatured to ssDNA by diluting the source material with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to mass [DNA]

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372-A200720122012-NaOH121417

0

4

6

0

4

6

220 260 300 340

Wavelength nm

2372-B200720122012-NaOH121517

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

28

To ensure the validity of ddPCR-based value assignments it is thus necessary to evaluate the nature of the partitioning entities Should a solution of what is thought to be dsDNA contain an appreciable fraction of ssDNA then the ddPCR results would require adjustment31 To avoid further delay in making SRM 2372a available to the forensic community the following sections detail a series of experiments that have enabled us to conclude with high confidence that the SRM 2372a components have a dsDNA conformation and that for this material ldquoentitiesrdquo can be considered dsDNA ldquocopiesrdquo The remaining stock of the SRM 2372 components also have a dsDNA conformation however the strands in the old materials are much less robustly connected than they are in the new materials Digital Polymerase Chain Reaction Assays

Three nDNA assays were used HBB1 NEIF and ND6 The HBB1 and NEIF assays reproducibly yield very similar results for DNA extracts from many different donors The HBB1 target is at the tip of the short (p) arm of chromosome 11 The NEIF target is in the centromere region of chromosome 2 The ND6 assay was selected for use because it typically yields the lowest results of the ten nDNA assays used to certify the SRM 2372a components (see Figure 10 below) The ND6 target is about halfway down the long (q) arm of chromosome 6 Heat Treatments (Melting of the DNA into Single Strands)

In all experiments the non-heated control and the heat-treated diluted solutions were stored in the same type of PCR reaction tube The starting point for the heat treatment was to determine if there was a difference between snap-cooling on ice and slower cooling at 25 degC Two tubes of the diluted material were placed in a GeneAmp 9700 thermal cycler (Applied Biosystems) and cycled up to 98 degC and held for 15 min After 15 min one tube was removed and placed on ice while the remaining tube was cooled at 1 degCs in the thermal cycler to 25 degC then placed on ice until prepared for ddPCR Additional heat treatments included cycling samples on a GeneAmp 9700 thermal cycler

a) from 98 degC for 5 min then cooling to 25 degC for 2 min During the 25 degC hold a randomly selected tube was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 45 min at 98 degC

b) from 94 degC for 1 min then cooling to 25 degC for 1 min During the 1 min hold at 25 degC a random sample was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 9 min at 94 degC

c) from 90 degC for 1 min then cooling to 25 degC for 1 min for a total of 20 cycles Samples were pulled at 25 degC after 1 min 3 min 5 min 10 min and 20 min and placed on ice

After the above experiments heat treatments were performed using a ProFlex thermal cycler This thermal cycler has six temperature zones that can be le 5 degC apart This enables heating

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

29

tubes at the same time but at different temperatures After treatment tubes were cooled at 5 degCs to 4 degC in the thermal cycler While designed for use with PCR plates a MicroAmp 96-well tray for the VeriFlex block enabled its use with single PCR reaction tubes Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a

As a first step in evaluating whether the DNA in the SRM materials disperse as dsDNA or ssDNA a direct comparison of the five materials was made using the HBB1 and NEIF assays Table 18 reports summary statistics for the ddPCR measurements reported as λ copiesdroplet and the estimated mass [DNA] values from the ddPCR results18 from the conventional dsDNA relationship7 and from the conventional ssDNA relationship after denaturing with 04 molL NaOH89 While the mass [DNA] values differ somewhat between the three methods there is no evidence for the SRM materials segregating exclusively or even mostly as ssDNA The extent of the hypochromic interactions among the nucleotides does not accurately reflect the extent to which the two strands of dsDNA have separated

Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a

λ copies per droplet λ copies per droplet ngmicroL Sample Assay Nrep

a Mean CVb ncmbc Mean CVb ddPCRd dsDNAe ssDNAf

2372-A HBB1 4 03645 19 8 03644 20 566g 524 57 2372-A NEIF 4 03644 23 2372-B HBB1 4 03513 23 8 03487 35 580h 536 61 2372-B NEIF 4 03461 47 2372a-A HBB1 4 02603 27 8 02649 30 474 465 454 2372a-A NEIF 4 02694 24 2372a-B HBB1 4 02908 34 8 02891 37 517 556 534 2372a-B NEIF 4 02873 45 2372a-C HBB1 4 02757 28 8 02711 28 485 502 474 2372a-C NEIF 4 02665 16

a Number of technical replicates for the given assay b CV = 100(standard deviation)mean c Combined number of technical replicates ignoring potential differences between the assays d Estimated from the ddPCR λ measurements listed in this Table18 e Estimated from the conventional relationship between the mass [DNA] of dsDNA and absorbance at

260 nm7 using absorbance spectra acquired shortly after solutions were produced f Estimated from the relationship between the mass [DNA] of ssDNA and absorbance at 260 nm after 11

dilution with 04 molL NaOH89 g Values adjusted to measured (132 plusmn19) volume loss of archived units due to evaporation h Values adjusted to measured (71 plusmn25) volume loss of archived due to evaporation

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

30

Quantitative Evaluation of ssDNA Content

Absorbance spectrophotometry is routinely used to quantitatively assess DNA tertiary structure3233 On storage ldquoat physiological pH the intrastrand repulsion between the negatively charged phosphate groups forces the double helix into more rigid rodlike conformationrdquo34 suggesting a mechanism for the observed slow changes in the SRM 2372 materials As demonstrated in Table 18 absorbance at 260 nm does not necessarily correlate with ddPCR behavior Yet ldquofurthermore the repulsion between phosphate groups on opposite strands tends to separate the complementary strandsrdquo34 Therefore a method that is independent of hypochromism is needed to adequately evaluate the proportion (if any) of ldquotruerdquo (completely disjoint) ssDNA in the SRM materials Circular dichroism spectroscopy is sensitive to the conformational structures of DNA in solution35 and different DNA configurations migrate differentially in agarose gel electrophoresis7 Both techniques have been used to qualitatively discriminate ssDNA from dsDNA in ldquopurerdquo single-sequence materials but neither technique appears to be able to quantify small proportions of ssDNA in a complex multi-chromosome genomic material Wilson and Ellison have proposed a real-time chamber digital PCR (cdPCR) technique that can estimate excess proportions of ssDNA copies from the crossing-threshold distribution31 However their method is insufficiently sensitive for low ssDNA proportions It is well known that heating DNA beyond the melting temperature separates mammalian dsDNA into ssDNA36 albeit we have shown using cdPCR that boiling dsDNA for some minutes also renders some target sites non-amplifiable15 In the following section we document a ddPCR-based approach for estimating the proportion if any of ssDNA in a sample that involves less destructive heating While the method is not yet optimized the results from these preliminary studies demonstrate that the SRM 2372 and 2372a components disperse into the ddPCR droplets as dsDNA ddPCR Measurement of ssDNA Proportion Given two otherwise identical aliquots of a given DNA material one subjected to some treatment and the other untreated the ratio between the average number of copies per droplet of treated material and that of its untreated sibling λtλu quantitatively estimates the change in the number of independently dispersing entities Should the treatment reduce the number of accessible amplifiable targets (AAT) λtλu will be less than 1 Reduction in AAT may result from damage to the target itself andor conformational changes to the fragment containing the target that render the target inaccessible or non-amplifiable16 Assuming no denaturation of dsDNA to ssDNA either because of an ineffective treatment or the material having started as ldquopurerdquo ssDNA if the treatment does not itself damage targets then λtλu will equal 1 Should treatment convert at least some dsDNA to ssDNA without damaging targets the number of AAT will increase and λtλu will be greater than 1 Assuming complete conversion of ldquopurerdquo dsDNA to ldquopurerdquo ssDNA without other damage then λtλu will equal 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

31

Values greater than 1 but less than 2 indicate some conversion of dsDNA to ssDNA but also some combination of (a) incomplete denaturation of dsDNA to ssDNA (b) partial renaturation of ssDNA to dsDNA after treatment but before ddPCR analysis (c) damage to targets during the process or ndash of greatest interest ndash (d) the original sample material was an impure mixture of dsDNA and ssDNA The following discussions graphically summarize sets of λtλu followed by tabular listings of the relevant experimental conditions ddPCR measurements and the numerical λtλu values The information in these tables includes

Plate Index Experiment identifier indicates date performed in YYYYMMDD format Material Sample material SRM 2372-A 2372-B 2372a-A 2372a-B or 2372a-C Assay ddPCR assay HBB1 NEIF or ND6 degC Denaturation temperature in degC Min Length of time the treated material was held at denaturation temperature in

minutes Cool Method used to cool treated material from denaturation temperature Snap

Slow R1 or R5 where ldquoSnaprdquo indicates cooling on ice ldquoSlowrdquo cooling at room temperature to 25 degC ldquoR1rdquo cooling in a thermal cycler to 25 degC at 1 degCs and ldquoR5rdquo cooling in a thermal cycler to 4 degC at 5 degCs

F Dilution factor expressed as (volume Material) per (final volume of diluted solution) 12 14 or 18

n Number of technical replicates Mean Mean of technical replicates CV Coefficient of Variation expressed as a percentage 100 Standard Deviation

Mean

λtλu the copies per droplet result of the treated solution λt divided by the copies per droplet result of its untreated sibling λu

u(λtλu) Standard uncertainty of λtλu 119906119906(120582120582t 120582120582ufrasl ) = 120582120582t 120582120582ufrasl CV(120582120582t)100

2

+ CV(120582120582u)100

2

Proof of Principle Figure 15 summarizes two replicate sets of measurements performed to assess whether the SRM 2372 and 2372a materials respond similarly whether the denatured materials renature if allowed to cool slowly and whether results are (reasonably) repeatable Table 19 details the experimental conditions The λtλu values for the four materials are similar all much greater than 1 but not approaching 2 Most of the ratios are on or close to the ldquoslow cooling is the same as snap coolingrdquo line suggesting that renaturation is not a significant concern While the replicate results for the SRM 2372-A and 2372a-A materials do not overlap as closely as those for SRM 2372-B and 2372a-B similar treatment conditions produce similar results

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

32

Figure 15 Comparison of Snap-Cooling to Slow-Cooling

Symbols mark the means of replicate measurements made using the same dilutions on successive days with bars representing plusmn one standard combined uncertainty Results are labeled by material ldquoArdquo and ldquoBrdquo for SRM 2372 components and ldquoaArdquo ldquoaBrdquo and ldquoaCrdquo for SRM 2372a components The diagonal black line represents equality between the snap- and slow-cooling treatments

Table 19 Comparison of Snap-Cooling to Slow-Cooling

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171218 2372-A HBB1 14 4 03830 29

dd20171218 2372-A HBB1 98 15 Slow 14 2 05644 18 1474 0051 dd20171218 2372-A HBB1 98 15 Snap 14 4 05852 15 1528 0051

dd20171218 2372a-A HBB1 14 4 03029 18

dd20171218 2372a-A HBB1 98 15 Slow 14 4 04430 21 1463 0040 dd20171218 2372a-A HBB1 98 15 Snap 14 4 04454 33 1471 0055

dd20171218 2372-B HBB1 14 4 03594 31

dd20171218 2372-B HBB1 98 15 Slow 14 4 05806 18 1615 0058 dd20171218 2372-B HBB1 98 15 Snap 14 4 05825 13 1620 0055

dd20171218 2372a-B HBB1 14 4 03383 15

dd20171218 2372a-B HBB1 98 15 Slow 14 4 05043 38 1491 0061 dd20171218 2372a-B HBB1 98 15 Snap 14 4 05254 20 1553 0039

dd20171219 2372-A HBB1 14 4 03357 23

dd20171219 2372-A HBB1 98 15 Slow 14 4 05340 16 1591 0046 dd20171219 2372-A HBB1 98 15 Snap 14 4 05340 12 1591 0042

A

aA

B

aB

A

aA

B

aB

14

15

16

17

14 15 16 17

λ tλ

u Sn

ap C

oole

d

λtλu Slow Cooled

Rep 1Rep 2Snap = Slow

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

33

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171219 2372a-A HBB1 14 4 02735 24

dd20171219 2372a-A HBB1 98 15 Slow 14 4 04151 03 1518 0037 dd20171219 2372a-A HBB1 98 15 Snap 14 4 03918 21 1433 0045 dd20171219 2372-B HBB1 14 4 03346 22

dd20171219 2372-B HBB1 98 15 Slow 14 4 05476 17 1637 0045 dd20171219 2372-B HBB1 98 15 Snap 14 4 05480 29 1638 0059 dd20171219 2372a-B HBB1 14 4 03158 19

dd20171219 2372a-B HBB1 98 15 Slow 14 4 04798 16 1519 0038 dd20171219 2372a-B HBB1 98 15 Snap 14 4 04856 17 1538 0039

Time at Denaturation Temperature Figure 16 summarizes λtλu for the SRM 2372-B material as a function of denaturation temperature and the cumulative time the solution is held at that temperature Table 20 details the experimental conditions

Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures

The symbols represent the ratio of the mean result of three technical replicates of the untreated and each of the treated solutions the bars represent combined standard uncertainties The lines denote regression fits to the function λtλu = atimese-bt where t is the cumulative time Values and the standard uncertainties of the coefficients are listed in the legend

12

14

16

18

20

0 5 10 15 20

λ t λ

u

Cummulative Treatment Time min

degC a u (a ) b u (b )90 1804 0039 00198 0000494 1788 0029 00099 0002998 1855 0019 00067 00019

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

34

For all three of the denaturation temperatures studied the relationship between λtλu and cumulative time t can be modeled as

λtλu = ae-bt

where a is the extrapolated value of λtλu at t = 0 and b is the rate of change in AAT per minute As expected the rate of loss due to the treatment increases with increasing treatment temperature and cumulative time at that temperature37 Given the much longer treatment time for the 98 degC series the agreement among the three t = 0 parameter (a) values is remarkable This suggests that the first 1 min of treatment efficiently separates dsDNA into ssDNA The pooled relative standard uncertainty of these regression parameter is 17 somewhat smaller than the 25 pooled relative standard uncertainties of the individual λtλu However the plusmnstandard uncertainty interval of the 1-min λtλu results for the 94 degC and 90 degC both overlap the extrapolated t = 0 value suggesting that reliable results can be obtained using a single treatment of short duration

Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171228 2372-B HBB1 14 3 03650 17 dd20171228 2372-B HBB1 98 5 R1 14 3 06157 12 1687 0035 dd20171228 2372-B HBB1 98 10 R1 14 3 05493 08 1505 0028 dd20171228 2372-B HBB1 98 15 R1 14 3 05019 13 1375 0029 dd20171228 2372-B HBB1 98 20 R1 14 3 04583 21 1256 0034 dd20171228 2372-B HBB1 98 25 R1 14 3 04154 32 1138 0041 dd20171228 2372-B HBB1 98 30 R1 14 3 03784 14 1037 0023 dd20171228 2372-B HBB1 98 35 R1 14 3 03408 10 0934 0019 dd20171228 2372-B HBB1 98 40 R1 14 3 02980 01 0816 0014 dd20171228 2372-B HBB1 98 40 R1 14 3 02815 08 0771 0015

dd20171229 2372-B HBB1 14 3 03708 27

dd20171229 2372-B HBB1 94 1 R1 14 3 06598 26 1779 0056 dd20171229 2372-B HBB1 94 2 R1 14 3 06553 32 1767 0063 dd20171229 2372-B HBB1 94 3 R1 14 3 06301 20 1699 0045 dd20171229 2372-B HBB1 94 4 R1 14 3 06252 14 1686 0037 dd20171229 2372-B HBB1 94 5 R1 14 3 06577 14 1773 0039 dd20171229 2372-B HBB1 94 6 R1 14 3 06156 07 1660 0030 dd20171229 2372-B HBB1 94 7 R1 14 3 06118 10 1650 0033 dd20171229 2372-B HBB1 94 8 R1 14 3 06256 12 1687 0035 dd20171229 2372-B HBB1 94 9 R1 14 3 05999 10 1618 0032

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

35

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180104 2372-B HBB1 14 4 03713 25 dd20180104 2372-B HBB1 90 1 R1 14 3 06916 22 1863 0052 dd20180104 2372-B HBB1 90 3 R1 14 3 06465 06 1741 0032 dd20180104 2372-B HBB1 90 5 R1 14 3 06431 20 1732 0045 dd20180104 2372-B HBB1 90 10 R1 14 3 06014 25 1620 0049 dd20180104 2372-B HBB1 90 15 R1 14 3 06024 10 1623 0032 dd20180104 2372-B HBB1 90 20 R1 14 3 06003 14 1617 0035

Temperature Figure 17 summarizes the effect of denaturation temperature using a single treatment of 05 min on the 2372-B and 2372a-B materials Table 21 details the experimental conditions

Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

The symbols represent the ratio of the mean result of four technical replicates of the untreated and three replicates of the treated solutions the bars represent combined standard uncertainties In all cases the duration of the treatment was 05 min The lines denote regression fits to the linear function λtλu = a +bT where Tis the denaturation temperature

The SRM 2372 and 2372a materials respond very differently to changes in denaturation temperature The older spectrophotometrically ssDNA material suffers somewhat greater loss in λtλu as temperature increases The new spectrophotometrically dsDNA material suffers much greater loss in λtλu as temperature decreases

14

15

16

17

18

19

20

82 86 90 94 98

λ tλ

u

Temperature degC

2372-B

2372a-B

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

36

Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180109 2372-B HBB1 14 4 03577 25 dd20180109 2372-B HBB1 100 05 R5 14 3 06314 16 1765 0051

dd20180104a 2372-B HBB1 14 4 03577 05

dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032 dd20180104a 2372-B HBB1 94 05 R5 14 3 06459 25 1806 0046 dd20180104a 2372-B HBB1 90 05 R5 14 2 06591 08 1842 0017 dd20180104a 2372-B HBB1 88 05 R5 14 3 06589 05 1842 0013 dd20180104a 2372-B HBB1 84 05 R5 14 3 06729 10 1881 0022 dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

We hypothesize that in the older SRM 2372 material the between-strand bond is relatively fragile and can be completely severed at relatively low temperature while avoiding much temperature-dependent AAT loss In contrast the between-strand bond in the SRM 2372a material requires higher temperature to completely disassociate the strands This suggests that there is not a sample-independent ldquooptimumrdquo combination of denaturation temperature and treatment duration However 98 degC for 05 min appears to provide comparable results for these two materials DNA Concentration in the ddPCR Solution Figure 18 summarizes the effect of [DNA] in the ddPCR solution with SRM 2372-B at a denaturation temperature of 98 degC and treatment duration of 05 min Table 22 details the experimental conditions

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

37

Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution

The solid black symbols represent the ratio of the mean result of four technical replicates of the untreated and treated solutions with the λtλu values shown on the left-hand vertical axis the bars represent combined standard uncertainties The solid blue circles represent λu and the open red circles λt with values shown on the right-hand vertical axis at this graphical scale the standard deviations of the technical replicates are covered by the symbol The diagonal blue and red lines represent the log-linear relationships between just the 18 and 14 λ values

The primary effect of [DNA] appears to be the loss of linearity in λt with copiesdroplet values greater than about 1 copiesdroplet There may be some advantage to use somewhat greater dilutions than the routine 14 however the small increase in λtλu may be offset by the increased variability with small λu This suggests the use of dilutions that give λu values of (01 to 04) copiesdroplet

Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180117 2372-B HBB1 12 4 07127 11 dd20180117 2372-B HBB1 98 05 R5 12 4 12331 14 1730 0031

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 3 06151 15 1843 0040

dd20180117 2372-B HBB1 18 4 01549 18 dd20180117 2372-B HBB1 98 05 R5 18 4 02897 20 1870 0051

12 14 18

170

175

180

185

190

02

04

06

08

101214

λ T

arge

t D

ropl

et

λ tλ

u

Dilution Factor

λ₊λᵤλᵤλ₊

λtλu

λu

λt

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

38

Equivalence of Assays Figure 19 summarizes the agreement between the HBB1 NEIF and ND6 assays at a denaturation temperature of 98 degC and treatment duration of 05 min Table 23 details the experimental conditions

Figure 19 Comparison of Assays

The symbols compared the λtλu results from the HBB1 and either the NEIF or ND6 assays Each λtλu value is the ratio of the mean result of four technical replicates of an untreated and three or four replicates of the treated solutions The bars represent combined standard uncertainties The diagonal line represents equality between the assays

There is little or no difference between the λtλu results from the HBB1 NEIF and ND6 assays suggesting that the chromosomal location of the assay target has little influence Since the NEIFHBB1 comparison includes results for both the SRM 2372 and 2372a materials this equivalence is not material-specific

14

16

18

20

14 16 18 20

λ tλ

u NE

IF o

r ND6

λtλu HBB1

NEIFHBB1

ND6HBB1

2nd Assay = HBB1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

39

Table 23 Comparison of Assays

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu)

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045 dd20180105 2372a-B ND6 94 05 R5 14 3 05526 36 1711 0074 dd20180105 2372a-B ND6 90 05 R5 14 3 05326 18 1649 0051 dd20180105 2372a-B ND6 88 05 R5 14 3 05077 22 1572 0052 dd20180105 2372a-B ND6 86 05 R5 14 3 05038 23 1560 0053 dd20180105 2372a-B ND6 84 05 R5 14 3 04906 10 1519 0041

dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

40

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials

Figure 20 summarizes the currently available λtλu results obtained by denaturing at 98 degC for 05 min rapidly cooling in the thermal cycler with dilutions that provide asymp 03 copiesdroplet The figure also displays the consensus result and its 95 confidence region estimated using the Hierarchical Bayes method provided in the NIST Consensus Builder (NICOB) system3839 Table 24 details the experimental conditions

Figure 20 Evaluation of Limiting Ratio

The symbols represent the ratio of the mean result of three or four technical replicates of the untreated and treated solutions the bars represent combined standard uncertainties The thick horizontal line represents a consensus value covering all displayed results the dotted horizontal lines bracket the 95 confidence interval about the consensus value

The Hierarchical Bayes consensus result for these values is 183 with a 95 confidence interval of plusmn003 All but three of the 16 λtλu values are within one standard uncertainty of this consensus value with most of the λtλu means lying within the 95 confidence interval This suggests that there are no statistically significant differences between the SRM 2372 and 2372a materials with regard to ssDNA proportion

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

17

18

19

20

λ tλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

41

Table 24 Evaluation of Limiting Ratio

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180104a 2372-B HBB1 14 4 03577 05 dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032

dd20180108 2372-B HBB1 14 3 03463 09 dd20180108 2372-B HBB1 98 05 R5 14 3 06562 14 1895 0031

dd20180109 2372-B HBB1 14 3 03577 25 dd20180109 2372-B HBB1 98 05 R5 14 3 06495 21 1816 0059

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 4 06151 15 1843 0040

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

42

The agreement among the measured λtλu is more directly visualized using estimated unilateral ldquoDegrees of Equivalencerdquo d Each d is the absolute difference between the measured and consensus values

119889119889 = |120582120582119905119905 120582120582119906119906frasl minus consensus|

and is associated with a 95 confidence interval U95(d) that combines the uncertainty estimates of the measured and consensus values The U95(d) calculation details are specific to the consensus estimation method39 Values of d plusmn U95(d) that include 0 indicate that the λtλu result does not differ from the majority at about a 95 level of confidence Figure 21 displays the d for the λtλu results for the Hierarchical Bayes consensus value39 By this figure of merit all the λtλu can be considered as members of a single population

Figure 21 Degrees of Equivalence

The symbols represent the Degree of Equivalence d for the λtλu results relative to the Hierarchical Bayes consensus value The bars represent 95 level of confidence uncertainty intervals d plusmn U95(d) The horizontal line marks the d = 0 ldquono differencerdquo value Results with d plusmn U95(d) that include this line are not significantly different from the majority

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

-024

-018

-012

-006

000

006

012

018

024

Degr

ees o

f Equ

ival

ence

λtλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

43

Conclusions and Recommendations With high confidence all tubes of the SRM 2372a solution have the same nDNA copy number content within measurement repeatability Additionally within measurement repeatability all tubes of the SRM 2372a have the same mtDNAnDNA With high confidence the nDNA copy number content of the SRM 2372a solution is thermally stable from 4 degC to 37 degC over an extended period The solution should not be shipped or stored below 4 degC With high confidence the ddPCR measurements used to establish the entity and mass concentrations and the mtDNAnDNA ratios are adequately precise and unbiased With high confidence the DNA in the SRM 2372a solutions is now dsDNA and can be expected to remain dsDNA for at least ten years when stored at 4 degC Recommended Certification Values

Since the component materials are believed to be homogeneous are of very similar composition and were prepared and processed similarly the observed differences in the measured entity concentration uncertainties are likely artifacts of the measurement process To limit over-interpretation of these differences we recommend that the copy per nanoliter values of the three components be assigned same 95 relative uncertainty interval (CV95) based on the largest of the measured CV95 values (93 ) rounded up to 10 Since the ngμL transformation does not appreciably inflate this relative uncertainty we recommend that the mass concentration values also be assigned CV95 of 10 Similarly we recommend that the mtDNAnDNA ratios be rounded to integer values all three components be assigned the same CV95 and that this uncertainty be based on the largest (21 ) of the measured CV95 rounded up to 25 Table 25 lists the values that we recommend be used in the Certificate of Analysis

Table 25 Recommended Values and 95 Uncertainties for Certification

Component Units Value U95(Value) a A Copies per nanoliter 151 15 B Copies per nanoliter 175 18 C Copies per nanoliter 145 15 A ngμL 498 50

B ngμL 578 58 C ngμL 479 48 A mtDNAnDNA 174 4

B mtDNAnDNA 206 5 C mtDNAnDNA 279 7

a The interval Value plusmn U95(Value) is believed with about 95 confidence to contain the true value of the measurand

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

44

Use of SRM 2372a for Value-Assigning In-House Reference Materials

The within-component and between-assay variability observed for the commercial qPCR assays is not unexpected since different qPCR methods exploit qualitatively different molecular targets Given the intrinsic variability of the human genome the prospect of developing a human DNA source material that responds equally to all current and future qPCR assays is improbable All three components should be used to elucidate potential bias when using these materials to value assign in-house DNA reference and control materials for qPCR assays

Certificate of Analysis In accordance with ISO Guide 31 2000 a NIST SRM certificate is a document containing the name description and intended purpose of the material the logo of the US Department of Commerce the name of NIST as a certifying body instructions for proper use and storage of the material certified property value(s) with associated uncertainty(ies) method(s) used to obtain property values the period of validity if appropriate and any other technical information deemed necessary for its proper use A Certificate is issued for an SRM certified for one or more specific physical or engineering performance properties and may contain NIST reference information or both values in addition to certified values A Certificate of Analysis is issued for an SRM certified for one or more specific chemical properties Note ISO Guide 31 is updated periodically check with ISO for the latest version [httpswwwnistgovsrmsrm-definitions] For the most current version of the Certificate of Analysis for NIST SRM 2372a Human DNA Quantification Standard please visit httpswww-snistgovsrmorsview_detailcfmsrm=2372a

References

1 National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research Department of Health Education and Welfare (1978) Belmont Report Ethical Principles and Guidelines for the Protection of Human Subjects of Research DHEW Publication No (OS) 78-0012 httpvideocastnihgovpdfohrp_belmont_reportpdf

2 CDCNIH Biosafety in Microbiological and Biomedical Laboratories 5th ed HHS publication No (CDC) 21-1112 Chosewood LC Wilson DE Eds US Government Printing Office Washington DC (2009) httpwwwcdcgovbiosafetypublicationsbmbl5indexhtm

3 Duewer DL Kline MC Redman JW Newall PJ Reeder DJ NIST Mixed Stain Studies 1 and 2 Interlaboratory Comparison of DNA Quantification Practice and Short Tandem Repeat Multiplex Performance with Multiple-Source Samples J Forensic Sci 200146(5)1199-1210 PMID 11569565

4 Kline MC Duewer DL Redman JW Butler JM NIST Mixed Stain Study 3 DNA quantitation accuracy and its influence on short tandem repeat multiplex signal intensity Anal Chem 200375(10)2463-2469 httpsdoiorg101021ac026410i

5 Kline MC Duewer DL Redman JW Butler JM Results from the NIST 2004 DNA Quantitation Study J Forensic Sci 200550(3)571-578 PMID 15932088

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

45

6 Kline MC Duewer DL Travis JC Smith MV Redman JW Vallone PM Decker AE Butler JM Production and Certification of Standard Reference Material 2372 Human DNA Quantitation Standard Anal Bioanal Chem 2009394(4)1183-1192 httpsdoiorg101007s00216-009-2782-0

7 Sambrook J and Russell DW (2001) Molecular Cloning a Laboratory Manual Cold Spring Harbor Laboratory Press Cold Spring Harbor New York

8 EU-JRC Protocol NK603 ndash Community Reference Laboratory Event-specific method for the quantification of maize line NK603 using real-time PCR Section 33 Spectrophotometric Measurement of DNA Concentration httpgmo-crljrceceuropaeusummariesNK603-WEB-ProtocolValidationpdf

9 ISO 215712005(E) Foodstuffs ndash Methods of analysis for the detection of genetically modified organisms and derived products ndash Nucleic acid extraction Annex B Methods for quantitation of extracted DNA pp 34-36 International Organization for Standardization Geneva (2005)

10 Certificate of Analysis Standard Reference Material 2372 Human DNA Quantitation Standard (2013) Gaithersburg MD USA Standard Reference Materials Program National Institute of Standards and Technology httpswww-snistgovsrmorsview_certcfmsrm=2372

11 Vogelstein B Kinzler KW Digital PCR Proc Natl Acad Sci USA 1999969236-9241 12 Hindson BJ Ness KD Masquelier DA Belgrader P Heredia NJ Makarewicz AJ Bright IJ Lucero

MY Hiddessen AL Legler TC Kitano TK Hodel MR Petersen JF Wyatt PW Steenblock ER Shah PH Bousse LJ Troup CB Mellen JC Wittmann DK Erndt NG Cauley TH Koehler RT So AP Dube S Rose KA Montesclaros L Wang S Stumbo DP Hodges SP Romine S Milanovich FP White HE Regan JF Karlin-Neumann GA Hindson CM Saxonov S Colston BW High-Throughput Droplet Digital PCR System for Absolute Quantitation of DNA Copy Number Anal Chem 2011838604-8610 httpsdoiorg101021ac202028g

13 Pinheiro LB Coleman VA Hindson CM Herrmann J Hindson BJ Bhat S Emslie KR Evaluation of a Droplet Digital Polymerase Chain Reaction Format for DNA Copy Number Quantification AnalChem 2012841003-1011 httpsdoiorg101021ac202578x

14 Corbisier P Pinheiro L Mazoua S Kortekaas AM Chung PY Gerganova T Roebben G Emons H Emslie K DNA copy number concentration measured by digital and droplet digital quantitative PCR using certified reference materials Anal Bioanal Chem 2015407(7)1831-1840 httpsdoiorg101007s00216-015-8458-z

15 Duewer DL Kline MC Romsos EL Real-time cdPCR opens a window into events occurring in the first few PCR amplification cycles Anal Bioanal Chem 2015407(30)9061-9069 httpsdoiorg101007s00216-015-9073-8

16 Kline MC Romsos EL Duewer DL Evaluating Digital PCR for the Quantification of Human Genomic DNA Accessible Amplifiable Targets Anal Chem 201688(4)2132-2139 httpsdoiorg101021acsanalchem5b03692

17 Kline MC Duewer DL Evaluating Droplet Digital Polymerase Chain Reaction for the Quantification of Human Genomic DNA Lifting the Traceability Fog Anal Chem 201789(8)4648-4654 httpsdoiorg101021acsanalchem7b00240

18 Duewer DL Kline MC Romsos EL Toman B Evaluating Droplet Digital PCR for the quantification of human genomic DNA Conversion from Copies per Nanoliter to Nanogram Genomic DNA per Microliter Anal Bioanal Chem 2018in press httpsdoiorg101007s00216-018-0982-1

19 Dagata JA Farkas N Kramer JA Method for measuring the volume of nominally 100 μm diameter spherical water in oil emulsion droplets NIST Special Publication 260-184 2016 httpsdoiorg106028NISTSP260-184

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

20 ISO 170342016 General requirements for the competence of reference material producers International Organization for Standardization Geneva (2016)

21 NCBI Standard Nucleotide BLAST httpsblastncbinlmnihgovBlastcgiPAGE_TYPE=BlastSearch

22 Timken MD Swango KL Orrego C Buoncristiani MR A Duplex Real-Time qPCR Assay for the Quantification of Human Nuclear and Mitochondrial DNA in Forensic Samples Implications for Quantifying DNA in Degraded Samples J Forensic Sci 200550(5)1044-60 PMID 16225209

23 Walker JA Hedges DJ Perodeau BP Landry KE Stoilova N Laborde ME Shewale J Sinha SK Batzer MA Multiplex polymerase chain reaction for simultaneous quantitation of human nuclear mitochondrial and male Y-chromosome DNA application in human identification Anal Biochem 200533789ndash97 httpsdoiorg101016jab200409036

24 Kavlick MF Lawrence HS Merritt RT Fisher C Isenberg A Robertson JM Budowle B Quantification of Human Mitochondrial DNA Using Synthesized DNA Standards J Forensic Sci 201156(6) 1457ndash1463 httpsdoiorg101111j1556-4029201101871x

25 Stulnig TM Amberger A Exposing contaminating phenol in nucleic acid preparations BioTechniques 199416(3) 403-404

26 Majumdar N Wessel T Marks J Digital PCR Modeling for Maximal Sensitivity Dynamic Range and Measurement Precision PLoS ONE 201510(3)e0118833 httpsdoi101371journalpone0118833

27 Tellinghuisen J Partition Volume Variability in Digital Polymerase Chain Reaction Methods Polydispersity Causes Bias but Can Improve Precision Anal Chem 2016881283-12187 httpsdoiorg101021acsanalchem6b03139

28 Lunn DJ Spiegelhalter D Thomas A Best N (2009) Statistics in Medicine 283049ndash3082 29 De Biegravevre P Dybkaer R Fajgelj A Hibbert DB Metrological traceability of measurement results in

chemistry Concepts and implementation Pure Appl Chem 201183(10)1873-1935 httpsdoiorg101007s00769-011-0805-y

30 Huggett JF Foy CA Benes V Emslie K Garson JA Haynes R Hellemans J Kubista M Mueller RD Nolan T Pfaffl MW Shipley GL Vandesompele J Wittwer CT Bustin SA The Digital MIQE Guidelines Minimum Information for Publication of Quantitative Digital PCR Experiments Clin Chem 201359(6)892-902 httpsdoiorg101373clinchem2013206375

31 Wilson PJ Ellison SLR Extending digital PCR analysis by modelling quantification cycle data BMC Bioinfo 201617421 httpsdoiorg101186s12859-016-1275-3

32 Wang X Son A Effects of pretreatment on the denaturation and fragmentation of genomic DNA for DNA hybridization Environ Sci Processes Impacts 2013152204 httpsdoiorg101039c3em00457k

33 Nwokeoji AO Kilby PM Portwood DE Dickman MJ Accurate Quantification of Nucleic Acids Using Hypochromicity Measurements in Conjunction with UV Spectrophotometry Anal Chem 201789(24)13567-13574 https101021acsanalchem7b04000

34 Farkas DH Kaul KL Wiedbrauk DL Kiechle FL Specimen collection and storage for diagnostic molecular pathology investigation Arch Pathol Lab Med 1996120(6)591-6

35 Kypr J Kejnovskaacute I Renčiuk D Vorličkovaacute M Circular dichroism and conformational polymorphism of DNA Nucleic Acids Res 200937(6)1713-1725 httpsdoiorg101093nargkp026

36 Strachan T Read AP (1999) Human Molecular Genetics 2nd Ed John Wiley amp Sons NY

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

37 Bhat S McLaughlin JL Emslie KR Effect of sustained elevated temperature prior to amplification on template copy number estimation using digital polymerase chain reaction Analyst 2011136724ndash32 httpsdoiorg101039c0an00484g

38 NIST Consensus Builder (NICOB) httpsconsensusnistgov 39 Koepke A Lafarge T Toman B Possolo A NIST Consensus Builder Userrsquos Manual

httpsconsensusnistgovNISTConsensusBuilder-UserManualpdf

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

Appendix A DNA Extraction Method The following method has been adapted from Miller et al (1988)A1 Procedure

1 Transfer approximately 5 mL of buffy coat into a 50 mL sterile polypropylene centrifuge tube and add 30 mL Red Blood Cell Lysis Buffer

2 Mix by gentle inversion and let stand at room temperature for 15 min

3 Centrifuge at approximately 209 rads (2000 rpm) for 20 min

4 Discard supernatant

5 Add 9 mL of Nucleic Lysis Buffer and re-suspend the pelleted cells by agitating with a plastic transfer pipette

6 Add 300 microL of 20 sodium dodecyl sulfate (SDS) and 18 mL of Protease K Solution (freshly prepared 2 mg to 5 mg Protease K per 1 mL Protease K Buffer)

7 Mix by inversion followed by incubation overnight or longer at 37 ordmC to 40 ordmC in a shaking water bath The length of incubation is determined by the time required to solubilize the cells present in the tube

8 Add 4 mL saturated ammonium acetate solution (in deionized water stored at 2 ordmC to 8 ordmC) Shake vigorously for 15 s

9 Centrifuge at approximately 471 rads (4500 rpm) for 10 min

10 Pipette supernatant into a 50 mL sterile polypropylene centrifuge tube and add two volumes of room temperature absolute ethanol Mix by gentle inversion until the DNA aggregates and floats to the top of the tube

11 Spool the aggregated DNA from the current tube into a fresh tube of 80 ethanol using a sterile inoculation loop At this step multiple tubes of precipitated DNA may be pooled to make a large lot of the purified material Mix by gentle inversion

12 Transfer the aggregated DNA to a fresh dry 50 mL polypropylene centrifuge tube Remove as much ethanol as possible by squeezing the DNA against the wall of the tube with a sterile inoculation loop Air-dry the material in a laminar flow hood

13 Resolubilize DNA in TE-4 buffer and store material at 4 ordmC Reagents

Red Blood Cell Lysis Buffer 144 mmolL ammonium chloride 1 mmolL sodium bicarbonate in deionized (DI) water Store at 2 ordmC to 8 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

Nucleic Lysis Buffer 10 mmolL Tris-HCl 400 mmolL sodium chloride 20 mmolL disodium ethylenediaminetetraacetic acid in DI Water adjust pH to 82 Store at 2 ordmC to 8 ordmC

Protease K Buffer

20 mmolL disodium ethylenediaminetetraacetic acid 10 (mass per volume) Sodium Dodecyl Sulfate (SDS) in DI water Store at 2 ordmC to 8 ordmC

Protease K

Sigma-Aldrich cat P6556 Tritirachium album lyophilized powder Store at ndash20 ordmC

Ammonium Acetate

192 molL ammonium acetate (saturated) in DI water Store at 2 ordmC to 8 ordmC

Absolute Ethanol (200 proof)

Store at room temperature Reference

A1 Miller SA Dykes DD Polesky HF A simple salting out procedure for extracting DNA from human nucleated cells Nucleic Acids Res 198816(3)1215 httpsdoiorg101093nar1631215

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

48

Appendix B Spectrophotometric Measurement Protocol The following is for use with a scanning recording spectrophotometer The procedures include the spectrophotometer calibration measurement of the ldquonativerdquo dsDNA and measurement of the NaOH denatured ssDNA The procedure for evaluating solutions after treatment with NaOH is adapted from references89 Spectrophotometer Calibration

Establish Baseline Set the spectrophotometer to acquire data from 220 nm to 345 nm in steps of 01 nm Zero the instrument with nothing (air) in the sample and reference cells Acquire an air-air baseline of the instrument with nothing (air) in the sample and reference cells If the trace has significant structure or excessive noise take corrective action If the trace is acceptable proceed with data collection Enable Verification of Wavelength Calibration SRM 2034 is a solution of holmium oxide in 10 perchloric acid contained in a sealed quartz cuvette delivering an intrinsic UVVis wavelength standard for absorption spectrophotometryB1 Evaluation of the location of peak maxima in the spectra of this solution enables verification of the spectrophotometers wavelength calibration Place SRM 2034 in the sample cell Create and name a file to hold the verification spectra Perform three replicate SRM 2034-to-air scans saving each scan to the file Remove SRM 2034 from the spectrophotometer and store appropriately Enable Verification of Absorbance Calibration SRM 2031 consists of three metal-on-fused silica filters held in cuvette-sized filter holders plus an empty filter holder The three filters are certified for transmittance and transmittance density (-log10(transmittance) effectively absorbance) at several wavelengths with nominal wavelength-independent transmittances of 10 30 and 90 Evaluation of the absorbance of each of these three filters at the certified 250 nm 280 nm and 340 nm wavelengths enables verification of the spectrophotometers absorbance calibration Place the empty SRM 2031 filter holder in the reference cell Remove the protective covers from the 10 transmittance filter and place in the sample cell Perform three replicate scans saving each scan to file Remove the filter from the spectrophotometer replace the protective covers and store appropriately Repeat for the 30 and 90 filters Close the file containing the verification spectra For ldquonativerdquo dsDNA Samples

Verify Optical Balance of Cuvettes The measurements of the DNA-containing samples are made using a 100 microL quartz cuvette in the sample cell compared to an optically matched cuvette in the reference cell The optical equivalence of these two cuvettes can be verified by filling both cuvettes with appropriate volumes of a sample solution matrix (TE-4 buffer) that does not contain DNA Fill two dry

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

49

clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes followed by a methanol rinse in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquonativerdquo dsDNA remove and clean both the reference cuvette and the sample cuvette For NaOH denatured ssDNA Samples

Prepare a 04 molL NaOH Stock Solution To a 15 mL disposable centrifuge tube add 016 g NaOH pellets add deionized (DI) water to the 10 mL line and mix until all the NaOH pellets have dissolved Centrifuge at 524 rads (5000 rpm) for 5 min For every 9 samples to be analyzed aliquot 1 mL of this 04 molL NaOH stock solution into a 15 mL screw top tube Use these tubes to prepare the NaOH blanks and DNA samples In a 15 mL screw cap vial mix 200 microL TE-4 buffer with 200 microL 04 molL sodium hydroxide solution to achieve a final NaOH concentration of 02 molL Vortex 5 s then centrifuge at 1257 rads (12 000 rpm) for 1 min Fill two dry clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Prepare the Sample Solutions For each unit of material to be measured (Tube or vial)

a) Vortex the unit for 5 s b) Centrifuge at 1257 rads (12 000 rpm) for 1 min c) Remove a 100 microL aliquot and place in a new labeled vial d) Add 100 microL of the freshly prepared 04 molL NaOH solution (with the same pipet) e) Vortex 5 s f) Centrifuge at 1257 rads (12 000 rpm) for 1 min

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

50

Measure Sample Solutions Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes rinse with methanol in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the NaOH denatured DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquodenaturedrdquo ssDNA close the file holding the spectra for this series Remove and clean both the reference cuvette and the sample cuvette After All Sample Measurements are Complete

Enable Verification of Spectrophotometer Stability The stability of the spectrophotometer over the course of the data collection can be evaluated by comparing spectra for SRM 2034 and 2031 taken at the start of data collection with spectra taken at the end of data collection Repeat the wavelength and absorbance verification measurements described above collecting the spectra in a suitably named file Clean Up Transfer all relevant data files from the spectrophotometers storage to a transfer device Clean the spectrophotometer work area as necessary and restore the SRMs and cuvettes to their appropriate storage locations Evaluation

Export the files to a spreadsheet Average the replicate spectra Subtract the average absorbance at A320 from the absorbance at A260 resulting in the corrected absorbance at A260 The conventional spectrophotometric estimate of the mass concentration of the dsDNA in the sample is achieved by multiplying the corrected absorbance by 50 The conventional estimate for the NaOH-treated ssDNA is achieved by multiplying the corrected absorbance by 37times2 where the ldquotimes2rdquo adjusts for the 1rarr2 volumetric dilution of the sample Note reliable spectrophotometric measurements require corrected A260 to be greater than 005 Reference

B1 Travis JC Acosta JC Andor G Bastie J Blattner P Chunnilall CJ Crosson SC Duewer DL Early EA Hengstberger F Kim C-S Liedquist L Manoocheri F Mercader F Metzdorf J Mito A Monardh LAG Nevask S Nilsson M Noeumll M Rodriguez AC Ruiacutez A Schirmacher A Smith MV Valencia G van Tonder N Zwinkels J Intrinsic Wavelength Standard Absorption Bands in Holmium Oxide Solution for UVvisible Molecular Absorption Spectrophotometry J Phys Chem Ref Data 200534(1)41-56

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

51

Appendix C SRM 2372a Statisticianrsquos report Extracted from Blaza Toman

November 30 2017 1 Introduction Candidate material for SRM 2372a is composed of three components labeled A B and C All components are human genomic DNA and prepared at NIST from buffy coat white blood cells from single-source anonymous donors Component A and B are single source (malefemale respectively) and Component C is a 1 male to 3 female mixture of two single-source donors Droplet digital PCR (ddPCR) assays were used to obtain measurements of the concentration of copies per microliter of sample Ten NIST-developed human nuclear DNA and three literature-derived human mitochondrial DNA PCR assays were used in this study

11 OpenBUGS Analysis All parameters in the following sections were estimated using Markov Chain Monte Carlo (MCMC) implemented in the OpenBUGS freewareC1 OpenBUGS is a software application for the Bayesian analysis of complex statistical models

12 OpenBUGS Exemplar Output All following sections list the OpenBUGs model (ldquocoderdquo) initialization values (ldquoinitsrdquo) if they are needed and the data used to evaluate the various data sets discussed in the body of this report homogeneity stability nDNA certification and mtDNAnDNA value assignment The provided information in each of the following sections is sufficient to produce estimates but the estimates may not exactly reproduce the exemplar results provided The MCMC paradigm does not produce ldquoanalyticalrdquo values but approaches them asymptotically as the number o model updates increases All exemplar results presented here are based upon the following bull 10-fold ldquothinningrdquo to minimize autocorrelation there were 10 MCMC estimates for every

model update used in parameter estimation bull a ldquoburn inrdquo of 5 000 updates before beginning parameter estimation and bull 10 000 model updates

The following statistics are provided in the tables listing the exemplar results

mean arithmetic mean of the parameterrsquos model update values sd standard deviation of the parameterrsquos model update values MC_error Monte Carlo standard deviation of the mean meanradic(number of model updates) val25pc 25th percentile of the model update values median 50th percentile of the model update values val975pc 975th percentile of the model update values startnumber of ldquoburn inrdquo updates sample number of model update values

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

52

2 Homogeneity assessment 21 Data

Samples of the material of all three components were drawn from 22 different boxes Two assays (NEIF ND6) were used to obtain 3 replicates of the measurement in copies per nanoliter for each sample and component For the complete homogeneity data see

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

22 Model The data were analyzed using a three-level Gaussian hierarchical ANOVA model

119910119910119894119894119894119894119894119894~119873119873120572120572119894119894119894119894119894119894 120590120590119908119908119894119894119905119905ℎ119894119894119894119894 119887119887119887119887119887119887 1198941198941198941198942

where ~N() denotes ldquodistributed as a normal distribution of some mean μ and variancerdquo 119894119894 denotes component (1 for A 2 for B 3 for C) 119895119895 denotes assay (1 for NEIF 2 for ND6) 119896119896 denotes box (1hellip22) and 120572120572119894119894119894119894119894119894~119873119873micro119894119894119894119894120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119887119887119887119887119887119887119894119894

2

All variance components were given a Half Cauchy prior as described in Reference C2 Note The Half Cauchy prior is specified by ldquo~dnorm(0 00016)I(0001)rdquo The ldquoIrdquo notation is for censoring It cuts off any random draws from the Normal distribution that are lower than 0001 An informative prior is needed when there are less than four data per distribution and the half Cauchy is a good general-purpose prior

23 OpenBUGS code and data Homogeneity code ModelBegin for(i in 13)xiNsb[i] ~ dnorm(0 00016)I(0001) chSqNsb[i] ~ dgamma(0505) sigb[i] lt- xiNsb[i]sqrt(chSqNsb[i]) for(i in 13)for(j in 12)mu[ij]~dnorm(010E-5) for(i in 13) muA[i]lt-mean(mu[i])sigbP[i]lt-100(muA[i]sqrt(sigb[i])) for(i in 13)for(j in 12)xiNs[ij] ~ dnorm(0 00016)I(0001) chSqNs[ij]~dgamma(0505) sigt[ij]lt-xiNs[ij]sqrt(chSqNs[ij]) sigtP[ij]lt-100(mu[ij]sqrt(sigt[ij])) for(i in 13)for(j in 12)for(k in 122)alpha[ijk]~dnorm(mu[ij]sigb[i]) for(idx in 1393)lamb[idx]~dnorm(alpha[cmp[idx]asy[idx]box[idx]]sigt[cmp[idx]asy[idx]]) ModelEnd Inits None required

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

53

Homogeneity Data lamb[] asy[] box[] cmp[] rep[] 1632 1 1 1 1 1529 1 1 1 2 1511 1 1 1 3 1616 2 1 1 1 1640 2 1 1 2 1582 2 1 1 3 1531 1 2 1 1 1525 1 2 1 2 1549 1 2 1 3 1615 2 2 1 1 1589 2 2 1 2 1601 2 2 1 3 1524 1 3 1 1 1540 1 3 1 2 1533 1 3 1 3 1582 2 3 1 1 1152 2 3 1 2 1555 2 3 1 3 1454 1 4 1 1 1458 1 4 1 2 1427 1 4 1 3 1627 2 4 1 1 1605 2 4 1 2 1607 2 4 1 3 1465 1 5 1 1 1540 1 5 1 2 1504 1 5 1 3 1607 2 5 1 1 1633 2 5 1 2 1586 2 5 1 3 1512 1 6 1 1 1491 1 6 1 2 1491 1 6 1 3 1511 2 6 1 1 1483 2 6 1 2 1524 2 6 1 3 1418 1 7 1 1 1480 1 7 1 2 1455 1 7 1 3 1479 2 7 1 1 1557 2 7 1 2 1656 2 7 1 3 1496 1 8 1 1 1560 1 8 1 2 1501 1 8 1 3 1514 2 8 1 1 1615 2 8 1 2 1540 2 8 1 3 1588 1 9 1 1 1581 1 9 1 2 1637 1 9 1 3 1474 2 9 1 1 1438 2 9 1 2 1454 2 9 1 3 1574 1 10 1 1 1513 1 10 1 2 1570 1 10 1 3 1594 2 10 1 1 1713 2 10 1 2 1669 2 10 1 3

1557 1 11 1 1 1508 1 11 1 2 1633 1 11 1 3 1571 2 11 1 1 1611 2 11 1 2 1631 2 11 1 3 1500 1 12 1 1 1476 1 12 1 2 1515 1 12 1 3 1626 2 12 1 1 1586 2 12 1 2 1608 2 12 1 3 1672 1 13 1 1 1634 1 13 1 2 1804 1 13 1 3 1490 2 13 1 1 1516 2 13 1 2 1497 2 13 1 3 1551 1 14 1 1 1568 1 14 1 2 1538 1 14 1 3 1617 2 14 1 1 1614 2 14 1 2 1639 2 14 1 3 1456 1 15 1 1 1448 1 15 1 2 1523 1 15 1 3 1525 2 15 1 1 1402 2 15 1 2 1436 2 15 1 3 1573 1 16 1 1 1520 1 16 1 2 1617 1 16 1 3 1634 2 16 1 1 1627 2 16 1 2 1575 2 16 1 3 1516 1 17 1 1 1516 1 17 1 2 1628 1 17 1 3 1613 2 17 1 1 1567 2 17 1 2 1567 2 17 1 3 1612 1 18 1 1 1535 1 18 1 2 1672 1 18 1 3 1529 2 18 1 1 1495 2 18 1 2 1528 2 18 1 3 1598 1 19 1 1 1514 1 19 1 2 1596 1 19 1 3 1571 2 19 1 1 1535 2 19 1 2 1511 2 19 1 3 1575 1 20 1 1 1568 1 20 1 2 1586 1 20 1 3 1652 2 20 1 1 1545 2 20 1 2 1627 2 20 1 3 1658 1 21 1 1

1611 1 21 1 2 1578 1 21 1 3 1647 2 21 1 1 1607 2 21 1 2 1612 2 21 1 3 1547 1 22 1 1 1528 1 22 1 2 1532 1 22 1 3 1537 2 22 1 1 1513 2 22 1 2 1536 2 22 1 3 1790 1 1 2 1 1909 1 1 2 2 1885 1 1 2 3 1870 2 1 2 2 1860 2 1 2 3 1712 1 2 2 1 1719 1 2 2 2 1704 1 2 2 3 1775 2 2 2 1 1718 2 2 2 2 1756 2 2 2 3 1693 1 3 2 1 1755 1 3 2 2 1749 1 3 2 3 1824 2 3 2 1 1483 2 3 2 2 1896 2 3 2 3 1754 1 4 2 1 1745 1 4 2 2 1739 1 4 2 3 1670 2 4 2 1 1632 2 4 2 2 1814 2 4 2 3 1757 1 5 2 1 1687 1 5 2 2 1735 1 5 2 3 1683 2 5 2 1 1759 2 5 2 2 1703 2 5 2 3 1697 1 6 2 1 1709 1 6 2 2 1714 1 6 2 3 1693 2 6 2 1 1723 2 6 2 2 1722 2 6 2 3 1722 1 7 2 1 1625 1 7 2 2 1703 1 7 2 3 1690 2 7 2 1 1684 2 7 2 2 1688 2 7 2 3 1884 1 8 2 1 1802 1 8 2 2 1772 1 8 2 3 1864 2 8 2 1 1887 2 8 2 2 1821 2 8 2 3 1713 1 9 2 1 1734 1 9 2 2 1727 1 9 2 3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

54

1709 2 9 2 1 1756 2 9 2 2 1758 2 9 2 3 1759 1 10 2 1 1809 1 10 2 2 1738 1 10 2 3 1764 2 10 2 1 1816 2 10 2 2 1787 2 10 2 3 1749 1 11 2 1 1782 1 11 2 2 1740 1 11 2 3 1741 2 11 2 1 1792 2 11 2 2 1784 2 11 2 3 1748 1 12 2 1 1712 1 12 2 2 1718 1 12 2 3 1797 2 12 2 1 1614 2 12 2 2 1790 2 12 2 3 1632 1 13 2 1 1668 1 13 2 2 1701 1 13 2 3 1626 2 13 2 1 1566 2 13 2 2 1681 2 13 2 3 1752 1 14 2 1 1535 1 14 2 2 1673 1 14 2 3 1724 2 14 2 1 1741 2 14 2 2 1685 2 14 2 3 1871 1 15 2 1 1835 1 15 2 2 1887 1 15 2 3 1770 2 15 2 1 1752 2 15 2 2 1798 2 15 2 3 1688 1 16 2 1 1725 1 16 2 2 1743 1 16 2 3 1744 2 16 2 1 1697 2 16 2 2 1710 2 16 2 3 1789 1 17 2 1 1769 1 17 2 2 1825 1 17 2 3 1716 2 17 2 1 1762 2 17 2 2 1638 2 17 2 3 1826 1 18 2 1 1780 1 18 2 2 1849 1 18 2 3 1805 2 18 2 1 1801 2 18 2 2 1931 2 18 2 3 1751 1 19 2 1 1759 1 19 2 2 1880 1 19 2 3 1827 2 19 2 1 1766 2 19 2 2

1786 2 19 2 3 1738 1 20 2 1 1678 1 20 2 2 1768 1 20 2 3 1752 2 20 2 1 1747 2 20 2 2 1768 2 20 2 3 1742 1 21 2 1 1836 1 21 2 2 1713 1 21 2 3 1746 2 21 2 1 1744 2 21 2 2 1764 2 21 2 3 1761 1 22 2 1 1743 1 22 2 2 1660 1 22 2 3 1666 2 22 2 1 1692 2 22 2 2 1819 2 22 2 3 1493 1 1 3 1 1385 1 1 3 2 1510 1 1 3 3 1565 2 1 3 1 1562 2 1 3 2 1466 2 1 3 3 1516 1 2 3 1 1513 1 2 3 2 1577 1 2 3 3 1490 2 2 3 1 1547 2 2 3 2 1570 2 2 3 3 1451 1 3 3 1 1528 1 3 3 2 1506 1 3 3 3 1485 2 3 3 1 1505 2 3 3 2 1471 2 3 3 3 1440 1 4 3 1 1486 1 4 3 2 1520 1 4 3 3 1529 2 4 3 1 1508 2 4 3 2 1506 2 4 3 3 1523 1 5 3 1 1472 1 5 3 2 1483 1 5 3 3 1513 2 5 3 1 1520 2 5 3 2 1491 2 5 3 3 1437 1 6 3 1 1493 1 6 3 2 1522 1 6 3 3 1492 2 6 3 1 1481 2 6 3 2 1461 2 6 3 3 1485 1 7 3 1 1477 1 7 3 2 1463 1 7 3 3 1475 2 7 3 1 1547 2 7 3 2 1485 2 7 3 3 1433 1 8 3 1

1405 1 8 3 2 1478 1 8 3 3 1497 2 8 3 1 1501 2 8 3 2 1495 2 8 3 3 1463 1 9 3 1 1499 1 9 3 2 1505 1 9 3 3 1503 2 9 3 1 1491 2 9 3 2 1516 2 9 3 3 1534 1 10 3 1 1458 1 10 3 2 1454 1 10 3 3 1448 2 10 3 1 1465 2 10 3 2 1474 2 10 3 3 1533 1 11 3 1 1514 1 11 3 2 1555 1 11 3 3 1488 2 11 3 1 1470 2 11 3 2 1483 2 11 3 3 1450 1 12 3 1 1525 1 12 3 2 1497 1 12 3 3 1496 2 12 3 1 1483 2 12 3 2 1519 2 12 3 3 1537 1 13 3 1 1527 1 13 3 2 1515 1 13 3 3 1499 2 13 3 1 1479 2 13 3 2 1475 2 13 3 3 1515 1 14 3 1 1490 1 14 3 2 1494 1 14 3 3 1469 2 14 3 1 1480 2 14 3 2 1466 2 14 3 3 1511 1 15 3 1 1568 1 15 3 2 1594 1 15 3 3 1515 2 15 3 1 1543 2 15 3 2 1565 2 15 3 3 1449 1 16 3 1 1452 1 16 3 2 1471 1 16 3 3 1543 2 16 3 1 1448 2 16 3 2 1491 2 16 3 3 1388 1 17 3 1 1421 1 17 3 2 1394 1 17 3 3 1452 2 17 3 1 1435 2 17 3 3 1475 1 18 3 1 1455 1 18 3 2 1445 1 18 3 3 1458 2 18 3 1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

55

1531 2 18 3 2 1459 2 18 3 3 1468 1 19 3 1 1525 1 19 3 2 1446 1 19 3 3 1435 2 19 3 1 1478 2 19 3 2 1421 2 19 3 3 1549 1 20 3 1

1531 1 20 3 2 1499 1 20 3 3 1499 2 20 3 1 1465 2 20 3 2 1468 2 20 3 3 1530 1 21 3 1 1440 1 21 3 2 1484 1 21 3 3 1499 2 21 3 1

1454 2 21 3 2 1442 1 22 3 1 1487 1 22 3 2 1436 1 22 3 3 1589 2 22 3 1 1591 2 22 3 2 1556 2 22 3 3 END

Table C-1 Example OpenBUGS Homogeneity Summary

mean sd MC_error val25pc median val975pc start sample mu[11] 15460 0122 000123 1522 1546 1570 5001 10000 mu[12] 15640 0137 000141 1537 1564 1591 5001 10000 mu[21] 17490 0113 000118 1727 1749 1772 5001 10000 mu[22] 17480 0129 000125 1722 1748 1774 5001 10000 mu[31] 14870 0073 000073 1473 1487 1501 5001 10000 mu[32] 14960 0069 000063 1482 1496 1509 5001 10000 muA[1] 15550 0092 000092 1536 1555 1573 5001 10000 muA[2] 17490 0085 000086 1732 1749 1765 5001 10000 muA[3] 14910 0050 000046 1481 1491 1501 5001 10000 sigbP[1] 3269 04985 000514 2376 3244 4324 5001 10000 sigbP[2] 2548 04385 000461 1733 2533 3472 5001 10000 sigbP[3] 1850 0292 000260 1329 1832 2477 5001 10000

sigtP[11] 2774 0298 000260 2268 2750 3442 5001 10000 sigtP[12] 4036 0429 000438 3290 4005 4986 5001 10000 sigtP[21] 2664 0301 000321 2155 2640 3313 5001 10000 sigtP[22] 3807 0395 000377 3116 3772 4662 5001 10000 sigtP[31] 2313 0250 000209 1880 2291 2873 5001 10000 sigtP[32] 1812 0195 000170 1473 1797 2243 5001 10000

24 Results The material appears to be homogeneous as can be seen in Figure C-1 Based on these results the between-tube variability is of the same order as within-tube variability

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

56

Component A Assay NEIF Component A Assay ND6 box plot gamma[11]

[111][112][113]

[114]

[115][116]

[117]

[118]

[119]

[1110][1111]

[1112]

[1113]

[1114]

[1115]

[1116][1117]

[1118]

[1119][1120]

[1121]

[1122]

114

box plot gamma[12]

[121][122]

[123]

[124][125]

[126]

[127][128]

[129]

[1210]

[1211][1212]

[1213]

[1214]

[1215]

[1216]

[1217]

[1218][1219]

[1220][1221]

[1222]

115

Component B Assay NEIF Component B Assay ND6

box plot gamma[21]

[211]

[212]

[213][214]

[215][216]

[217]

[218]

[219]

[2110][2111]

[2112]

[2113][2114]

[2115]

[2116]

[2117]

[2118]

[2119]

[2120]

[2121]

[2122]

129

box plot gamma[22]

[221]

[222][223]

[224][225][226][227]

[228]

[229]

[2210][2211]

[2212]

[2213]

[2214]

[2215]

[2216][2217]

[2218]

[2219]

[2220][2221][2222]

128

Component C Assay NEIF Component C Assay ND6

box plot gamma[31]

[311]

[312]

[313][314]

[315][316]

[317]

[318]

[319][3110]

[3111]

[3112]

[3113]

[3114]

[3115]

[3116]

[3117]

[3118]

[3119]

[3120]

[3121]

[3122]109

box plot gamma[32]

[321][322]

[323]

[324][325]

[326]

[327][328][329]

[3210]

[3211]

[3212]

[3213][3214]

[3215]

[3216]

[3217]

[3218]

[3219]

[3220][3221]

[3222]

11

Figure C-1 Homogeneity Assessment

Note These results were obtained using the 1-4 gravimetric dilution-adjusted λ but uncorrected for droplet volume

or the 1-10 volumetric dilution That is the vertical axis in each panel should be multiplied by 07349times10 = 7349 This constant difference in scale does not impact the homogeneity analysis Error bars span approximate 95 confidence intervals

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

57

3 Certification measurements 31 Data

Originally the experiment consisted of 5 plates containing the 3 components (A B C) and 12 assays Ten of the assays were for genomic DNA identification (for the certified value) and two of the assays were for mitochondrial detection which is used for informational values (and are not included in this data set) Each of the plates contained samples from a new vial from different boxes (noted in the ldquovial usedrdquo portion the number references the box) The assay positions were alternated across the five plates as well For the plate design and which tubes were used see

Figure 6 Basic Plate Layout Design for Certification Measurements It was later determined that the homogeneity and stability data sets were obtained using a different set of pipettes than the certification data set and since all data was to be combined for certification another plate was obtained using the original pipette For the complete certification data see

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component Stability data was measured at different temperatures over several weeks using assays NEIF and ND6 Only the 4degC data was also used for certification For the complete stability data see

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

32 Model It was decided that all three sets of data should be used to derive the certified values In this way the uncertainty of the certified values would better reflect any potential heterogeneity in the samples as well as any additional unknown factors that may have effected the results Because two different pipettes were used to obtain different sets of data there was a significant bias in some of the measurements Specifically the ldquocertificationrdquo data from the first five plates was obtained using pipette set ldquo1rdquo and the remaining measurements were obtained using pipette set ldquo2rdquo This bias was introduced into the statistical model as part of the mean Possible effects of different assays and type of data (heterogeneity certification and stability) were accounted for in the statistical model in terms of a random effect Specifically the response for each component (A B C) was represented as

119910119910119894119894119894119894119894119894~1198731198731205741205741198941198941198941198941198941198941205901205901198941198942 where ~N() denotes ldquodistributed as a normal distribution of some mean and variancerdquo i = 123 denotes type of measurement heterogeneity certification and stability j = 1hellip10 denotes assay 2PR4 POTP NEIF R4Q5 D5 ND6 D9 HBB1 ND14 22C3 k = 1 2 denotes pipette set 120574120574119894119894119894119894119894119894 = 120572120572119894119894119894119894 + 120573120573119894119894 120572120572119894119894119894119894~119873119873(micro1198941198941205901205901205721205722) 120573120573119894119894 = 120573120573 for data type 1 and 120573120573119894119894 = 0 for data types 2 and 3 For a given 2372a component the variance 1205901205901205721205722 measures the between assay variability and 1205901205901198941198942 measures the repeatability variability for each data type The data was analyzed using Bayesian MCMC programmed in OpenBUGSC2 with non-informative Gaussian priors for the means and Gamma priors for the variance components

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

58

33 OpenBUGS code inits and data Certification code ModelBegin for(i in 13)mu[i]~dnorm(010E-5) sig[i]~dgamma(10E-510E-5) sigalph~dgamma(10E-510E-5) for(i in 110)alpha[1i]~dnorm(mu[1] sigalph) for(i in 12)alpha[2i]~dnorm(mu[2] sigalph) alpha[3i]~dnorm(mu[3]sigalph) b~dnorm(010E-5) beta[1]lt-b beta[2]lt-00 for(i in 1Ntot)mean[i]lt-alpha[typ[i]asy[i]]+beta[pip[i]] for(i in 1Ntot)lamb[i]~dnorm(mean[i]sig[typ[i]]) ModelEnd Certification Inits list(sig=c(111) sigalph=1) Certification Data for Component A list(Ntot=275) lamb[] asy[] cmp[] typ[] pip[] 1667 1 1 1 1 1655 2 1 1 1 1647 3 1 1 1 1594 4 1 1 1 1655 5 1 1 1 1629 6 1 1 1 1655 7 1 1 1 1644 8 1 1 1 1645 9 1 1 1 1545 10 1 1 1 1698 1 1 1 1 1685 2 1 1 1 1665 3 1 1 1 1658 4 1 1 1 1683 5 1 1 1 1647 6 1 1 1 1652 7 1 1 1 1611 8 1 1 1 1707 9 1 1 1 1603 10 1 1 1 1668 1 1 1 1 1587 2 1 1 1 1678 3 1 1 1 1628 4 1 1 1 1738 5 1 1 1 1667 6 1 1 1 1695 7 1 1 1 1757 8 1 1 1 1771 9 1 1 1 1762 10 1 1 1 1739 1 1 1 1 1702 2 1 1 1 1614 3 1 1 1 1762 4 1 1 1 1649 5 1 1 1 1678 6 1 1 1 1653 7 1 1 1

175 8 1 1 1 1764 9 1 1 1 1759 10 1 1 1 1638 1 1 1 1 1676 2 1 1 1 1616 3 1 1 1 1559 4 1 1 1 1608 5 1 1 1 1611 6 1 1 1 1517 7 1 1 1 1557 8 1 1 1 1604 9 1 1 1 1675 10 1 1 1 1667 1 1 1 1 1638 2 1 1 1 1624 3 1 1 1 1587 4 1 1 1 1652 5 1 1 1 1589 6 1 1 1 1607 7 1 1 1 1582 8 1 1 1 1625 9 1 1 1 1616 10 1 1 1 1888 1 1 1 1 1868 2 1 1 1 1953 3 1 1 1 1779 4 1 1 1 1924 5 1 1 1 1861 6 1 1 1 1824 7 1 1 1 1809 8 1 1 1 1864 9 1 1 1 193 10 1 1 1 1939 1 1 1 1 1815 2 1 1 1 1902 3 1 1 1 1871 4 1 1 1 2124 5 1 1 1

1787 6 1 1 1 1844 7 1 1 1 1799 8 1 1 1 1857 9 1 1 1 1942 10 1 1 1 1597 1 1 1 1 1533 2 1 1 1 1462 3 1 1 1 1493 4 1 1 1 1538 5 1 1 1 1492 6 1 1 1 1528 7 1 1 1 1548 8 1 1 1 1513 9 1 1 1 1562 10 1 1 1 159 1 1 1 1 1404 2 1 1 1 1533 3 1 1 1 1519 4 1 1 1 1586 5 1 1 1 1516 6 1 1 1 1538 7 1 1 1 1581 8 1 1 1 1514 9 1 1 1 1431 1 1 1 2 1407 2 1 1 2 1487 3 1 1 2 145 4 1 1 2 1394 5 1 1 2 1396 6 1 1 2 1414 7 1 1 2 1401 8 1 1 2 1412 9 1 1 2 1435 10 1 1 2 1411 1 1 1 2 1395 2 1 1 2 1398 3 1 1 2 1439 4 1 1 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

59

1422 5 1 1 2 1393 6 1 1 2 139 7 1 1 2 1382 8 1 1 2 146 9 1 1 2 141 10 1 1 2 1632 1 1 2 2 1531 1 1 2 2 1524 1 1 2 2 1454 1 1 2 2 1465 1 1 2 2 1512 1 1 2 2 1418 1 1 2 2 1496 1 1 2 2 1588 1 1 2 2 1574 1 1 2 2 1557 1 1 2 2 15 1 1 2 2 1672 1 1 2 2 1551 1 1 2 2 1573 1 1 2 2 1516 1 1 2 2 1612 1 1 2 2 1598 1 1 2 2 1575 1 1 2 2 1658 1 1 2 2 1547 1 1 2 2 1529 1 1 2 2 1525 1 1 2 2 154 1 1 2 2 1458 1 1 2 2 154 1 1 2 2 1491 1 1 2 2 148 1 1 2 2 156 1 1 2 2 1581 1 1 2 2 1513 1 1 2 2 1508 1 1 2 2 1476 1 1 2 2 1634 1 1 2 2 1568 1 1 2 2 152 1 1 2 2 1516 1 1 2 2 1535 1 1 2 2 1514 1 1 2 2 1568 1 1 2 2 1611 1 1 2 2 1528 1 1 2 2 1511 1 1 2 2 1549 1 1 2 2 1533 1 1 2 2 1427 1 1 2 2 1504 1 1 2 2 1491 1 1 2 2 1455 1 1 2 2

1501 1 1 2 2 1637 1 1 2 2 157 1 1 2 2 1633 1 1 2 2 1515 1 1 2 2 1804 1 1 2 2 1538 1 1 2 2 1617 1 1 2 2 1628 1 1 2 2 1672 1 1 2 2 1596 1 1 2 2 1586 1 1 2 2 1578 1 1 2 2 1532 1 1 2 2 1616 2 1 2 2 1615 2 1 2 2 1582 2 1 2 2 1627 2 1 2 2 1607 2 1 2 2 1511 2 1 2 2 1479 2 1 2 2 1514 2 1 2 2 1474 2 1 2 2 1594 2 1 2 2 1571 2 1 2 2 1626 2 1 2 2 149 2 1 2 2 1617 2 1 2 2 1634 2 1 2 2 1613 2 1 2 2 1529 2 1 2 2 1571 2 1 2 2 1652 2 1 2 2 1647 2 1 2 2 1537 2 1 2 2 164 2 1 2 2 1589 2 1 2 2 1152 2 1 2 2 1605 2 1 2 2 1633 2 1 2 2 1483 2 1 2 2 1557 2 1 2 2 1615 2 1 2 2 1438 2 1 2 2 1713 2 1 2 2 1611 2 1 2 2 1586 2 1 2 2 1516 2 1 2 2 1614 2 1 2 2 1627 2 1 2 2 1567 2 1 2 2 1495 2 1 2 2 1535 2 1 2 2 1545 2 1 2 2 1607 2 1 2 2

1513 2 1 2 2 1582 2 1 2 2 1601 2 1 2 2 1555 2 1 2 2 1607 2 1 2 2 1586 2 1 2 2 1524 2 1 2 2 1656 2 1 2 2 154 2 1 2 2 1454 2 1 2 2 1669 2 1 2 2 1631 2 1 2 2 1608 2 1 2 2 1497 2 1 2 2 1639 2 1 2 2 1575 2 1 2 2 1567 2 1 2 2 1528 2 1 2 2 1511 2 1 2 2 1627 2 1 2 2 1612 2 1 2 2 1536 2 1 2 2 1568 1 1 3 2 152 1 1 3 2 1486 1 1 3 2 1441 1 1 3 2 1706 1 1 3 2 144 1 1 3 2 15 1 1 3 2 1537 1 1 3 2 1387 1 1 3 2 1463 1 1 3 2 157 1 1 3 2 1429 1 1 3 2 1667 1 1 3 2 1632 1 1 3 2 1589 1 1 3 2 1474 2 1 3 2 1543 2 1 3 2 1555 2 1 3 2 1599 2 1 3 2 1494 2 1 3 2 1624 2 1 3 2 1577 2 1 3 2 1593 2 1 3 2 1479 2 1 3 2 1517 2 1 3 2 1538 2 1 3 2 159 2 1 3 2 1479 2 1 3 2 1479 2 1 3 2 1522 2 1 3 2 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

60

Table C-2 Example OpenBUGS Certification Summary for Component A

mean sd MC_error val25pc median val975pc start sample b 2584 02948 0005376 2001 2588 3158 100001 100000 mu[1] 1417 02706 0005366 1364 1416 147 100001 100000 mu[2] 1559 008368 4776E-4 1543 1559 1576 100001 100000 mu[3] 1533 01504 0001419 1504 1533 1563 100001 100000

Certification Data for Component B list(Ntot=278) lamb[] asy[] cmp[] typ[] pip[] 1869 1 2 1 1 185 2 2 1 1 1709 3 2 1 1 1856 4 2 1 1 1848 5 2 1 1 1828 6 2 1 1 1861 7 2 1 1 1825 8 2 1 1 1871 9 2 1 1 1776 10 2 1 1 196 1 2 1 1 1781 2 2 1 1 1707 3 2 1 1 1787 4 2 1 1 1829 5 2 1 1 1826 6 2 1 1 1858 7 2 1 1 1869 8 2 1 1 1773 9 2 1 1 1902 10 2 1 1 2174 1 2 1 1 2202 2 2 1 1 2172 3 2 1 1 223 5 2 1 1 2094 6 2 1 1 1949 7 2 1 1 1971 8 2 1 1 2123 9 2 1 1 2107 10 2 1 1 2231 1 2 1 1 2292 2 2 1 1 2216 3 2 1 1 2078 4 2 1 1 2192 5 2 1 1 2034 6 2 1 1 211 7 2 1 1 2123 8 2 1 1 2066 9 2 1 1 2046 10 2 1 1 181 1 2 1 1 1865 2 2 1 1 1806 3 2 1 1 1878 4 2 1 1 1824 5 2 1 1 1768 6 2 1 1 1811 7 2 1 1 1763 8 2 1 1 1865 9 2 1 1 1815 10 2 1 1

1806 1 2 1 1 1801 2 2 1 1 1786 3 2 1 1 1768 4 2 1 1 179 5 2 1 1 167 6 2 1 1 1749 7 2 1 1 1741 8 2 1 1 1781 9 2 1 1 1866 10 2 1 1 1784 1 2 1 1 1735 2 2 1 1 1778 3 2 1 1 1835 4 2 1 1 1794 5 2 1 1 17 6 2 1 1 1799 7 2 1 1 1765 8 2 1 1 1754 9 2 1 1 1792 10 2 1 1 1821 1 2 1 1 1813 2 2 1 1 1813 3 2 1 1 1749 4 2 1 1 1848 5 2 1 1 1711 6 2 1 1 1779 7 2 1 1 1769 8 2 1 1 1725 9 2 1 1 1814 10 2 1 1 1718 1 2 1 1 1753 2 2 1 1 1773 3 2 1 1 1777 4 2 1 1 1803 5 2 1 1 1718 6 2 1 1 1802 7 2 1 1 1834 8 2 1 1 1731 9 2 1 1 1842 10 2 1 1 1768 1 2 1 1 1698 2 2 1 1 1792 3 2 1 1 174 4 2 1 1 1795 5 2 1 1 1693 6 2 1 1 1743 7 2 1 1 1745 8 2 1 1 1714 9 2 1 1 1791 1 2 1 2

1746 2 2 1 2 1803 3 2 1 2 1767 4 2 1 2 1745 5 2 1 2 1824 6 2 1 2 1752 7 2 1 2 1813 8 2 1 2 1648 9 2 1 2 1818 10 2 1 2 177 1 2 1 2 182 2 2 1 2 1803 3 2 1 2 1739 4 2 1 2 1773 5 2 1 2 1729 6 2 1 2 1744 7 2 1 2 1754 8 2 1 2 1671 9 2 1 2 1741 10 2 1 2 179 1 2 2 2 1712 1 2 2 2 1693 1 2 2 2 1754 1 2 2 2 1757 1 2 2 2 1697 1 2 2 2 1722 1 2 2 2 1884 1 2 2 2 1713 1 2 2 2 1759 1 2 2 2 1749 1 2 2 2 1748 1 2 2 2 1632 1 2 2 2 1752 1 2 2 2 1871 1 2 2 2 1688 1 2 2 2 1789 1 2 2 2 1826 1 2 2 2 1751 1 2 2 2 1738 1 2 2 2 1742 1 2 2 2 1761 1 2 2 2 1909 1 2 2 2 1719 1 2 2 2 1755 1 2 2 2 1745 1 2 2 2 1687 1 2 2 2 1709 1 2 2 2 1625 1 2 2 2 1802 1 2 2 2 1734 1 2 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

61

1809 1 2 2 2 1782 1 2 2 2 1712 1 2 2 2 1668 1 2 2 2 1535 1 2 2 2 1835 1 2 2 2 1725 1 2 2 2 1769 1 2 2 2 178 1 2 2 2 1759 1 2 2 2 1678 1 2 2 2 1836 1 2 2 2 1743 1 2 2 2 1885 1 2 2 2 1704 1 2 2 2 1749 1 2 2 2 1739 1 2 2 2 1735 1 2 2 2 1714 1 2 2 2 1703 1 2 2 2 1772 1 2 2 2 1727 1 2 2 2 1738 1 2 2 2 174 1 2 2 2 1718 1 2 2 2 1701 1 2 2 2 1673 1 2 2 2 1887 1 2 2 2 1743 1 2 2 2 1825 1 2 2 2 1849 1 2 2 2 188 1 2 2 2 1768 1 2 2 2 1713 1 2 2 2 166 1 2 2 2 1775 2 2 2 2 1824 2 2 2 2 167 2 2 2 2 1683 2 2 2 2 1693 2 2 2 2 169 2 2 2 2 1864 2 2 2 2 1709 2 2 2 2 1764 2 2 2 2

1741 2 2 2 2 1797 2 2 2 2 1626 2 2 2 2 1724 2 2 2 2 177 2 2 2 2 1744 2 2 2 2 1716 2 2 2 2 1805 2 2 2 2 1827 2 2 2 2 1752 2 2 2 2 1746 2 2 2 2 1666 2 2 2 2 187 2 2 2 2 1718 2 2 2 2 1483 2 2 2 2 1632 2 2 2 2 1759 2 2 2 2 1723 2 2 2 2 1684 2 2 2 2 1887 2 2 2 2 1756 2 2 2 2 1816 2 2 2 2 1792 2 2 2 2 1614 2 2 2 2 1566 2 2 2 2 1741 2 2 2 2 1752 2 2 2 2 1697 2 2 2 2 1762 2 2 2 2 1801 2 2 2 2 1766 2 2 2 2 1747 2 2 2 2 1744 2 2 2 2 1692 2 2 2 2 186 2 2 2 2 1756 2 2 2 2 1896 2 2 2 2 1814 2 2 2 2 1703 2 2 2 2 1722 2 2 2 2 1688 2 2 2 2 1821 2 2 2 2 1758 2 2 2 2 1787 2 2 2 2

1784 2 2 2 2 179 2 2 2 2 1681 2 2 2 2 1685 2 2 2 2 1798 2 2 2 2 171 2 2 2 2 1638 2 2 2 2 1931 2 2 2 2 1786 2 2 2 2 1768 2 2 2 2 1764 2 2 2 2 1819 2 2 2 2 1666 1 2 3 2 1626 1 2 3 2 188 1 2 3 2 1653 1 2 3 2 1733 1 2 3 2 1551 1 2 3 2 1884 1 2 3 2 1718 1 2 3 2 1549 1 2 3 2 1723 1 2 3 2 194 1 2 3 2 164 1 2 3 2 1565 1 2 3 2 1787 1 2 3 2 1655 2 2 3 2 167 2 2 3 2 1673 2 2 3 2 1705 2 2 3 2 1755 2 2 3 2 1625 2 2 3 2 1708 2 2 3 2 1705 2 2 3 2 1862 2 2 3 2 1686 2 2 3 2 1723 2 2 3 2 1909 2 2 3 2 1859 2 2 3 2 1897 2 2 3 2 1972 2 2 3 2 END

Table C-3 Example OpenBUGS Certification Summary for Component B

mean sd MC_error val25pc median val975pc start sample b 09337 03387 0007504 02695 09371 1592 100001 100000 mu[1] 1764 03101 0007483 1704 1763 1825 100001 100000 mu[2] 1748 008014 4719E-4 1733 1748 1764 100001 100000 mu[3] 1735 02338 0003093 1688 1736 1781 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

62

Certification data for Component C list(Ntot=280) lamb[] asy[] cmp[] typ[] pip[] 163 1 3 1 1 1596 2 3 1 1 1574 3 3 1 1 159 4 3 1 1 161 5 3 1 1 1589 6 3 1 1 1569 7 3 1 1 1565 8 3 1 1 1564 9 3 1 1 1548 10 3 1 1 1706 1 3 1 1 1576 2 3 1 1 1538 3 3 1 1 1555 4 3 1 1 1607 5 3 1 1 1548 6 3 1 1 1572 7 3 1 1 1681 8 3 1 1 1492 9 3 1 1 1655 10 3 1 1 1499 1 3 1 1 1576 2 3 1 1 1545 3 3 1 1 1584 4 3 1 1 1577 5 3 1 1 1599 6 3 1 1 1587 7 3 1 1 152 8 3 1 1 1479 9 3 1 1 1512 10 3 1 1 1497 1 3 1 1 1611 2 3 1 1 156 3 3 1 1 1546 4 3 1 1 1529 5 3 1 1 1478 6 3 1 1 1504 7 3 1 1 1551 8 3 1 1 1438 9 3 1 1 1494 10 3 1 1 1452 1 3 1 1 1347 2 3 1 1 1395 3 3 1 1 1442 4 3 1 1 1448 5 3 1 1 1463 6 3 1 1 1402 7 3 1 1 1474 8 3 1 1 1466 9 3 1 1 1439 10 3 1 1 1424 1 3 1 1 1397 2 3 1 1 135 3 3 1 1 1459 4 3 1 1 1602 5 3 1 1 1435 6 3 1 1 1433 7 3 1 1 1481 8 3 1 1 1482 9 3 1 1

1568 10 3 1 1 147 1 3 1 1 1411 2 3 1 1 1512 3 3 1 1 1452 4 3 1 1 1567 5 3 1 1 148 6 3 1 1 1442 7 3 1 1 143 8 3 1 1 1413 9 3 1 1 1486 10 3 1 1 1444 1 3 1 1 1427 2 3 1 1 1439 3 3 1 1 1468 4 3 1 1 1524 5 3 1 1 1424 6 3 1 1 1459 7 3 1 1 1467 8 3 1 1 1411 9 3 1 1 148 10 3 1 1 1485 1 3 1 1 1422 2 3 1 1 1438 3 3 1 1 1439 4 3 1 1 1515 5 3 1 1 1409 6 3 1 1 1469 7 3 1 1 1453 8 3 1 1 1465 9 3 1 1 1363 10 3 1 1 1458 1 3 1 1 1459 2 3 1 1 1465 3 3 1 1 1468 4 3 1 1 1457 5 3 1 1 1412 6 3 1 1 1474 7 3 1 1 1428 8 3 1 1 1425 9 3 1 1 1393 10 3 1 1 1357 1 3 1 2 1447 2 3 1 2 1448 3 3 1 2 1409 4 3 1 2 1389 5 3 1 2 1415 6 3 1 2 1431 7 3 1 2 1381 8 3 1 2 1403 9 3 1 2 1435 10 3 1 2 1364 1 3 1 2 1361 2 3 1 2 1404 3 3 1 2 1411 4 3 1 2 1429 5 3 1 2 1417 6 3 1 2 138 7 3 1 2 1336 8 3 1 2 1387 9 3 1 2

1372 10 3 1 2 1493 1 3 2 2 1516 1 3 2 2 1451 1 3 2 2 144 1 3 2 2 1523 1 3 2 2 1437 1 3 2 2 1485 1 3 2 2 1433 1 3 2 2 1463 1 3 2 2 1534 1 3 2 2 1533 1 3 2 2 145 1 3 2 2 1537 1 3 2 2 1515 1 3 2 2 1511 1 3 2 2 1449 1 3 2 2 1388 1 3 2 2 1475 1 3 2 2 1468 1 3 2 2 1549 1 3 2 2 153 1 3 2 2 1442 1 3 2 2 1385 1 3 2 2 1513 1 3 2 2 1528 1 3 2 2 1486 1 3 2 2 1472 1 3 2 2 1493 1 3 2 2 1477 1 3 2 2 1405 1 3 2 2 1499 1 3 2 2 1458 1 3 2 2 1514 1 3 2 2 1525 1 3 2 2 1527 1 3 2 2 149 1 3 2 2 1568 1 3 2 2 1452 1 3 2 2 1421 1 3 2 2 1455 1 3 2 2 1525 1 3 2 2 1531 1 3 2 2 144 1 3 2 2 1487 1 3 2 2 151 1 3 2 2 1577 1 3 2 2 1506 1 3 2 2 152 1 3 2 2 1483 1 3 2 2 1522 1 3 2 2 1463 1 3 2 2 1478 1 3 2 2 1505 1 3 2 2 1454 1 3 2 2 1555 1 3 2 2 1497 1 3 2 2 1515 1 3 2 2 1494 1 3 2 2 1594 1 3 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

63

1471 1 3 2 2 1394 1 3 2 2 1445 1 3 2 2 1446 1 3 2 2 1499 1 3 2 2 1484 1 3 2 2 1436 1 3 2 2 1565 2 3 2 2 149 2 3 2 2 1485 2 3 2 2 1529 2 3 2 2 1513 2 3 2 2 1492 2 3 2 2 1475 2 3 2 2 1497 2 3 2 2 1503 2 3 2 2 1448 2 3 2 2 1488 2 3 2 2 1496 2 3 2 2 1499 2 3 2 2 1469 2 3 2 2 1515 2 3 2 2 1543 2 3 2 2 1452 2 3 2 2 1458 2 3 2 2 1435 2 3 2 2 1499 2 3 2 2 1499 2 3 2 2 1589 2 3 2 2 1562 2 3 2 2 1547 2 3 2 2 1505 2 3 2 2 1508 2 3 2 2 152 2 3 2 2

1481 2 3 2 2 1547 2 3 2 2 1501 2 3 2 2 1491 2 3 2 2 1465 2 3 2 2 147 2 3 2 2 1483 2 3 2 2 1479 2 3 2 2 148 2 3 2 2 1543 2 3 2 2 1448 2 3 2 2 1531 2 3 2 2 1478 2 3 2 2 1465 2 3 2 2 1454 2 3 2 2 1591 2 3 2 2 1466 2 3 2 2 157 2 3 2 2 1471 2 3 2 2 1506 2 3 2 2 1491 2 3 2 2 1461 2 3 2 2 1485 2 3 2 2 1495 2 3 2 2 1516 2 3 2 2 1474 2 3 2 2 1483 2 3 2 2 1519 2 3 2 2 1475 2 3 2 2 1466 2 3 2 2 1565 2 3 2 2 1491 2 3 2 2 1435 2 3 2 2 1459 2 3 2 2

1421 2 3 2 2 1468 2 3 2 2 1556 2 3 2 2 1424 1 3 3 2 1427 1 3 3 2 1381 1 3 3 2 1394 1 3 3 2 1473 1 3 3 2 133 1 3 3 2 1446 1 3 3 2 1424 1 3 3 2 133 1 3 3 2 1452 1 3 3 2 1512 1 3 3 2 1314 1 3 3 2 1525 1 3 3 2 1475 1 3 3 2 1542 1 3 3 2 1461 2 3 3 2 1443 2 3 3 2 1547 2 3 3 2 1394 2 3 3 2 1364 2 3 3 2 1395 2 3 3 2 1522 2 3 3 2 1483 2 3 3 2 1452 2 3 3 2 1482 2 3 3 2 1439 2 3 3 2 1458 2 3 3 2 155 2 3 3 2 1502 2 3 3 2 152 2 3 3 2 END

Table C-4 Example OpenBUGS Certification Summary for Component C

mean sd MC_error val25pc median val975pc start sample b 09626 01701 0002428 06285 09633 1293 100001 100000 mu[1] 1399 01566 0002413 1369 1399 143 100001 100000 mu[2] 1491 005701 2386E-4 148 1491 1503 100001 100000 mu[3] 1449 0132 0001181 1423 1449 1475 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

64

34 Results Table C-5 lists the results in CopiesnL for the three components

Table C-5 Parameter Values for Components A B and C

Posterior Distribution Summary Statistics Component Parameter Mean SD 25 50 975

A

β 260 029 203 260 318 μ1 1415 027 1362 1415 1467 μ2 1559 010 1539 1559 158 μ3 1533 016 1501 1533 1565

B

β 095 034 029 095 162 μ1 1762 031 1701 1762 1822 μ2 1748 010 1728 1748 1769 μ3 1735 024 1687 1735 1783

C

β 097 017 064 097 130 μ1 1399 016 1368 1399 1429 μ2 1491 008 1476 1491 1507 μ3 1449 014 1421 1449 1477

Since in each case the three data set types represent independent measurements the three estimates can be combined to obtain a consensus value In this way a between data set type variability is included in the final standard uncertainty Table C-6 lists the results in Copies per Nanoliter obtained using the NIST Consensus Builder (NICOB)C3

Table C-6 Certified Values for Components A B and C

Component Mean u(Mean) Uk=2(Mean) A 151 07 14 B 175 03 06 C 145 05 10

4 Mito-to-nuclear DNA ratios (mtDNAnDNA)

41 Data Measurements were also obtained on the same samples using three mitochondrial detection assays mtND1 mtBatz and mtKav and a single nDNA assay (genomic) to compute the ratios mtDNAnDNA For the complete mtDNAnDNA data see

Table 16 mtDNAnDNA Ratios

42 Model For each component the data was analyzed using the following random effects model to account for between assay variability

119903119903119894119894~119873119873120572120572119894119894 1205901205901198861198861198861198861198861198861198861198861198861198862

where 119894119894 denotes assay (for A and B 1 = mtND1 2 = mtBatz 3 = mtKav for C 1 = mtBatz 2 = mtKav) and 120572120572119894119894~119873119873micro120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119886119886119886119886119886119886119886119886119886119886

2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

65

43 OpenBUGS code inits and data Two versions of the same model are required one for the results from the three mtDNA assays available for components A and B mtND1 mtBatz and mtKav and the other for the results from the two mtDNA assays available for component C mtBatz and mtKav mtDNAnDNA code for components A and B ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 13)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for components A and B list(sigt=c(111)) Data for Component A list(Ndata=102) set[] rep[] asy[] rat[] 1 1 1 1931 1 2 1 187 1 1 2 1845 1 2 2 1764 2 1 1 181 2 2 1 1906 2 1 2 1785 2 2 2 1837 3 1 1 1743 3 2 1 1761 3 1 2 1829 3 2 2 1847 4 1 1 1716 4 2 1 180 4 1 2 174 4 2 2 1813 5 1 1 1923 5 2 1 1835 5 1 2 1935 5 2 2 1851 6 1 1 1757 6 2 1 1724 6 3 1 1768 6 1 2 1743 6 2 2 1671 6 3 2 1766 6 1 3 1692 6 2 3 1662 6 3 3 1767 7 1 1 1924 7 2 1 1944 7 3 1 1893 7 1 2 1786

7 2 2 1752 7 3 2 1743 7 1 3 153 7 2 3 1542 7 3 3 1523 7 1 2 158 7 2 2 1557 7 3 2 1592 7 1 3 2103 7 2 3 210 7 3 3 2049 7 1 1 1749 7 2 1 1769 7 3 1 1777 7 1 2 1965 7 2 2 1889 7 3 2 1947 7 1 3 1813 7 2 3 1864 7 3 3 1864 8 1 1 178 8 2 1 1776 8 3 1 1706 8 1 2 1804 8 2 2 1751 8 3 2 1738 8 1 3 1769 8 2 3 1786 8 3 3 1761 9 1 2 1584 9 2 2 1544 10 1 1 1947 10 2 1 1928 10 3 1 1922 10 1 1 1829

10 2 1 1895 10 3 1 1851 10 1 1 1674 10 2 1 1617 11 1 2 1667 11 2 2 1813 11 3 2 1708 11 4 2 1598 11 1 2 1566 11 2 2 1687 11 3 2 1663 11 4 2 1589 11 1 2 1643 11 2 2 1668 11 3 2 1604 11 4 2 1573 12 1 2 1719 12 2 2 1716 12 3 2 167 12 4 2 1666 12 1 2 1581 12 2 2 1573 12 3 2 1601 12 4 2 1612 12 1 2 1546 12 2 2 1542 12 3 2 1538 12 4 2 1587 13 1 1 1617 13 2 1 1599 13 1 1 1622 13 2 1 1579 13 1 1 1575 13 2 1 1528 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

66

Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A

mean sd MC_error val25pc median val975pc start sample mu 1730 1443 004419 1703 1730 1759 100001 100000

data for Component B list(Ndata=87) set[] rep[] asy[] rat[] 1 1 1 2461 1 2 1 2401 1 1 2 2555 1 2 2 2608 2 1 1 2229 2 2 1 2200 2 1 2 2108 2 2 2 2192 3 1 1 2127 3 2 1 2198 3 1 2 2455 3 2 2 2506 4 1 1 2211 4 2 1 2199 4 1 2 2380 4 2 2 2249 5 1 1 2105 5 2 1 1943 5 1 2 2136 5 2 2 2119 6 1 1 1990 6 2 1 1921 6 3 1 1982 6 1 2 1990 6 2 2 1924 6 3 2 1958 6 1 3 1903 6 2 3 1831

6 3 3 1919 7 1 1 2304 7 2 1 2341 7 3 1 2332 7 1 2 2048 7 2 2 2078 7 3 2 2067 7 1 3 2399 7 2 3 2479 7 3 3 2493 7 1 1 2065 7 2 1 2166 7 3 1 2164 7 1 2 2314 7 2 2 2280 7 3 2 2258 7 1 3 1867 7 2 3 1987 7 3 3 1971 8 1 1 2046 8 2 1 2057 8 3 1 2024 8 1 2 1998 8 2 2 2043 8 3 2 1952 8 1 3 2077 8 2 3 2117 8 3 3 2041 9 1 2 1890 9 2 2 1886

10 1 2 1928 10 2 2 1916 10 3 2 1915 10 1 2 1947 10 2 2 1946 10 3 2 1961 10 4 2 1909 10 1 2 1953 10 2 2 1952 10 3 2 1941 10 4 2 1881 11 1 2 1865 11 2 2 1863 11 3 2 1828 11 4 2 1828 11 1 2 1944 11 2 2 1964 11 3 2 1947 11 4 2 1895 11 1 2 1908 11 2 2 1909 11 3 2 1940 11 4 2 1839 12 1 1 1688 12 2 1 1783 12 1 1 1823 12 2 1 1631 12 1 1 1920 12 2 1 1887 END

Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B

mean sd MC_error val25pc median val975pc start sample mu 2045 2246 0111 1999 2045 2088 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

67

mtDNAnDNA code for component C ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 12)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for component C list(sigt=c(11)) mtDNAnDNA data for component C list(Ndata=66) set[] rep[] asy[] rat[] 1 1 1 3201 1 2 1 3299 2 1 1 3347 2 2 1 3287 3 1 1 3217 3 2 1 3384 4 1 1 3200 4 2 1 3363 5 1 1 3294 5 2 1 3013 6 1 1 2805 6 2 1 2825 6 3 1 2829 6 1 2 2709 6 2 2 2738 6 3 2 2727 7 1 1 2797 7 2 1 2642 7 3 1 2792 7 1 2 2804 7 2 2 2789

7 3 2 2842 7 1 1 2715 7 2 1 2606 7 3 1 2687 7 1 2 2799 7 2 2 2628 7 3 2 2559 7 1 1 2639 7 2 1 2535 7 3 1 2657 7 1 2 2953 7 2 2 2757 7 3 2 2802 8 1 1 2927 8 2 1 2776 8 3 1 2762 8 1 2 3044 8 2 2 2830 8 3 2 2825 9 1 1 2501 9 2 1 2522 10 1 1 2832 10 2 1 2840

10 3 1 2788 10 4 1 2725 10 1 1 2835 10 2 1 2790 10 3 1 2736 10 4 1 2679 10 1 1 2803 10 2 1 2773 10 3 1 2757 10 4 1 2752 11 1 1 2599 11 2 1 2508 11 3 1 2551 11 4 1 2609 11 1 1 2617 11 2 1 2483 11 3 1 2626 11 4 1 2602 11 1 1 2665 11 2 1 2535 11 3 1 2570 11 4 1 2650 END

Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C

mean sd MC_error val25pc median val975pc start sample mu 2774 2367 01176 2729 2774 2821 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

68

Table C-10 Results for mtDNAnDNA Posterior Distribution Summary Statistics

Component Mean Std Dev 25 50 975 A 1730 14 1703 1730 1758 B 2045 23 1994 2047 2087 C 2774 23 2729 2774 2821

5 ReferencesC1 Lunn DJ Spiegelhalter D Thomas A Best N Statistics in Medicine 2009283049ndash3082C2 Polson NG Scott JG Bayesian Analysis 20127887ndash902C3 NIST Consensus Builder Software and userrsquos guide freely available at

httpsconsensusnistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

69

Appendix D Performance in Commercial qPCR Chemistries The performance of the SRM 2372a materials in commercial qPCR chemistries was assessed using manufacturerrsquos recommended protocols Each panel in Figures D-1 to D-4 displays the average cycle threshold (Ct) as a function of DNA concentration for the 5-point dilutions of SRM 2372a components In all cases the external calibrant was SRM 2372 component A Error bars represent approximate 95 confidence intervals Table D-1 lists the estimated [DNA] for the three SRM 2372a components for all commercial assays evaluated This Table also lists the Component BComponent A (BA) and Component CComponent A (CA) ratios calculated from the [DNA] estimates

Figure D-1 Results for the Innogenomics Commercial qPCR assays

20

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantLong(IQ_h)

ABCStd

[A] = (384 plusmn 26) ngμL[B] = (441 plusmn 24) ngμL[C] = (516 plusmn 24) ngμL

Rsup2 = 0999 se(Ct) = 01016

17

18

19

20

21

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantShort(IQ_d)

ABCStd

[A] = (435 plusmn 28) ngμL[B] = (543 plusmn 12) ngμL[C] = (510 plusmn 28) ngμL

Rsup2 = 0999 se(Ct) = 007

24

25

26

27

28

29

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYLong(IQHY_h)

ABCStd

[A] = (387 plusmn 24) ngμL[B] = (448 plusmn 21) ngμL[C] = (559 plusmn 31) ngμL

Rsup2 = 0998 se(Ct) = 01314

16

18

20

22

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYShort(IQHY_d)

ABCStd

[A] = (353 plusmn 23) ngμL[B] = (267 plusmn 15) ngμL[C] = (390 plusmn 18) ngμL

Rsup2 = 1000 se(Ct) = 00420

22

24

26

28

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYY(IQHY_y)

ABCStd

[A] = (572 plusmn 13) ngμL[B] = 0 ngμL[C] = (137 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

70

Figure D-2 Results for the Thermo Fisher Commercial qPCR assays

21

22

23

24

25

26

27

28

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HumanHuman(QFH_h)

ABCStd

[A] = (464 plusmn 13) ngμL[B] = (529 plusmn 19) ngμL[C] = (472 plusmn 22) ngμL

Rsup2 = 1000 se(Ct) = 004

22

23

24

25

26

27

28

29

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoHuman(QFD_h)

ABCStd

[A] = (701 plusmn 53) ngμL[B] = (561 plusmn 11) ngμL[C] = (472 plusmn 11) ngμL

Rsup2 = 0994 se(Ct) = 01922

24

26

28

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoMale(QFD_y)

ABCStd

[A] = (706 plusmn 49) ngμL[B] = 0 ngμL[C] = (122 plusmn 02) ngμL

Rsup2 = 0999 se(Ct) = 007

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPLarge(QFHP_h)

ABCStd

[A] = (340 plusmn 18) ngμL[B] = (432 plusmn 05) ngμL[C] = (342 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 00720

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPSmall(QFHP_d)

ABCStd

[A] = (434 plusmn 15) ngμL[B] = (514 plusmn 10) ngμL[C] = (466 plusmn 07) ngμL

Rsup2 = 1000 se(Ct) = 005

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioLarge(QFT_h)

ABCStd

[A] = (362 plusmn 23) ngμL[B] = (430 plusmn 06) ngμL[C] = (378 plusmn 07) ngμL

Rsup2 = 0999 se(Ct) = 00920

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioSmall(QFT_d)

ABCStd

[A] = (468 plusmn 08) ngμL[B] = (505 plusmn 08) ngμL[C] = (523 plusmn 14) ngμL

Rsup2 = 1000 se(Ct) = 00318

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioY(QFT_y)

ABCStd

[A] = (372 plusmn 11) ngμL[B] = 0 ngμL[C] = (110 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 003

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

71

Figure D-3 Results for the Qiagen Commercial qPCR assays

Figure D-4 Results for the Promega Commercial qPCR assays

19

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresAutosomal(QPH_h)

ABCStd

[A] = (554 plusmn 21) ngμL[B] = (479 plusmn 05) ngμL[C] = (381 plusmn 09) ngμL

Rsup2 = 1000 se(Ct) = 00418

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresY(QPH_y)

ABCStd

[A] = (358 plusmn 10) ngμL[B] = 0 ngμL[C] = (118 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 007

16

17

18

19

20

21

22

23

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProHuman(QPP_h)

ABCStd

[A] = (525 plusmn 49) ngμL[B] = (480 plusmn 42) ngμL[C] = (467 plusmn 45) ngμL

Rsup2 = 0999 se(Ct) = 01019

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProDegradation(QPP_d)

ABCStd

[A] = (567 plusmn 18) ngμL[B] = (558 plusmn 08) ngμL[C] = (484 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 00410

15

20

25

30

35

-50 -40 -30 -20 -10 00 10

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProY(QPP_y)

ABCStd

[A] = (538 plusmn 27) ngμL[B] = (00 plusmn 00) ngμL[C] = (162 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 003

16

18

20

22

24

-05 00 05 10 15 20

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorAutosomal(P_h)

ABCStd

[A] = (759 plusmn 64) ngμL[B] = (775 plusmn 24) ngμL[C] = (1056 plusmn 17) ngμL

Rsup2 = 0995 se(Ct) = 01818

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorY(P_y)

ABCStd

[A] = (596 plusmn 51) ngμL[B] = 0 ngμL[C] = (117 plusmn 01) ngμL

Rsup2 = 0994 se(Ct) = 020

18

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantAutosomal(PQ_h)

ABCStd

[A] = (436 plusmn 13) ngμL[B] = (606 plusmn 20) ngμL[C] = (504 plusmn 10) ngμL

Rsup2 = 1000 se(Ct) = 00218

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantDegradation(PQ_d)

ABCStd

[A] = (420 plusmn 18) ngμL[B] = (529 plusmn 16) ngμL[C] = (530 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 00516

18

20

22

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantY(PQ_y)

ABCStd

[A] = (410 plusmn 20) ngμL[B] = 0 ngμL[C] = (139 plusmn 05) ngμL

Rsup2 = 0999 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

72

Table D-1 Summary [DNA] Results for the Commercial qPCR Assays Estimated [DNA] ngμL Ratios

A B C BA CA Assaya xb u(x)c xb u(x)c xb u(x)c yd u(y)e yd u(y)e

IQd 435 28 543 12 510 28 125 008 117 010 IQh 384 26 441 24 516 24 115 010 134 011

IQHYd 353 23 267 15 390 18 076 006 111 009 IQHYh 387 24 448 21 559 31 116 009 144 012

Ph 759 64 775 24 1056 16 102 009 139 012 PQd 420 18 529 16 530 06 126 007 126 006 PQh 436 13 606 20 504 10 139 006 116 004

QFDh 701 53 561 11 472 11 080 006 067 005 QFHh 464 13 529 19 472 22 114 005 102 006

QFHPd 434 15 514 10 466 07 118 005 107 004 QFHPh 340 18 432 05 342 03 127 007 101 005

QFTd 468 08 505 08 523 14 108 003 112 004 QFTh 362 23 430 06 378 07 118 008 104 007 QPHh 554 21 479 05 380 09 086 003 069 003 QPPd 567 18 558 08 484 11 098 003 085 003 QPPh 525 49 480 42 467 45 092 012 089 012

IQHYy 572 13 NDf 137 06

NRg

024 001 Py 596 51 NDf 117 01 020 002

PQy 410 20 NDf 139 05 034 002 QFDy 706 49 NDf 122 02 017 001 QFTy 372 11 NDf 110 06 030 002 QPHy 358 10 NDf 118 03 033 001 QPPy 538 27 0002 0001 162 11 030 003

a Assay codes are listed in Table 8 b Mean of dilution-adjusted results from five dilutions of the SRM 2372a component c Standard uncertainty of the mean estimated from the stdanrd deviation of the five dilutions d Ratio of the estimated [DNA] of component B or C relative to that of component A e Standard uncertainty of the ratio estimated from the standard uncertainties of the individual components f Not detected Component B was prepared using tissue from only female donors it is not expected to

respond to male-specific assays g Not relevant

  • Abstract
  • Keywords
  • Technical Information Contact for this SRM
  • Table of Contents
  • Table of Tables
  • Table of Figures
  • Purpose and Description
  • Warning SRM 2372a is a Human Source Material
  • Storage and Use
  • History and Background
    • Figure 1 Sales History of SRM 2372
      • Experimental Methods
        • Polymerase Chain Reaction Assays
          • Table 1 NIST-Developed Human Nuclear DNA Assays
          • Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays
          • Table 2 NIST-Optimized Human Mitochondrial DNA Assays
          • Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays
            • Droplet Digital Polymerase Chain Reaction Measurements
              • Nuclear DNA (nDNA)
                • Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays
                  • Mitochondrial DNA (mtDNA)
                    • Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays
                    • Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays
                        • Sample Preparation
                        • Absorbance Measurements
                        • SRM Unit Production
                        • Component Homogeneity
                          • Figure 4 Plate Layout Design for Homogeneity Measurements
                            • Component Stability
                              • Figure 5 Plate Layout Design for Stability Measurements
                                • Component Certification Measurements
                                  • Table 6 Tubes used for Certification Measurements
                                  • Figure 6 Basic Plate Layout Design for Certification Measurements
                                    • Mitochondrial DNA Measurements
                                      • Table 7 Run Parameters for Mitochondrial DNA
                                        • Performance in Commercial qPCR Chemistries
                                          • Table 8 Commercial qPCR Chemistries Tested with SRM 2372a
                                          • Figure 7 Plate Layout Design for Performance Assessments
                                              • Measurement Results
                                                • Absorbance Spectrophotometry
                                                  • Table 9 Absorbance at Selected Wavelengths
                                                  • Figure 8 Absorbance Spectra of SRM 2372a Components A B and C
                                                    • Droplet Volume
                                                    • Certification Measurements
                                                      • Homogeneity Results
                                                        • Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component
                                                        • Table 11 Within- and Between-Tube Relative Standard Deviations
                                                          • Stability Results
                                                            • Figure 9 ddPCR Stability Measurements as Function of Time
                                                            • Table 12 Stability Data as λacute Copy Number per Nanoliter of Component
                                                              • Ten-Assay Certification Results
                                                                • Figure 10 Comparison of Assay Performance
                                                                • Table 13 Certification Data as λacute Copy Number per Nanoliter of Component
                                                                  • Estimation of Copy Number Concentration
                                                                    • Table 14 Measured Copy Number Concentrations
                                                                      • Conversion of Copy Number to Mass Concentration
                                                                        • Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution
                                                                            • Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)
                                                                              • Figure 11 mtDNAnDNA for Components A B and C
                                                                              • Table 16 mtDNAnDNA Ratios
                                                                              • Table 17 mtDNAnDNA Summary Results
                                                                                • Performance in Commercial qPCR Chemistries
                                                                                  • Figure 12 qPCR Autosomal Quantitation Performance Summary
                                                                                  • Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary
                                                                                      • Qualification of ddPCR as Fit-For-Purpose
                                                                                        • Figure 14 Absorbance Spectra of SRM 2372 Components A and B
                                                                                        • Digital Polymerase Chain Reaction Assays
                                                                                        • Heat Treatments (Melting of the DNA into Single Strands)
                                                                                        • Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a
                                                                                          • Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a
                                                                                            • Quantitative Evaluation of ssDNA Content
                                                                                              • ddPCR Measurement of ssDNA Proportion
                                                                                              • Proof of Principle
                                                                                                • Table 19 Comparison of Snap-Cooling to Slow-Cooling
                                                                                                  • Time at Denaturation Temperature
                                                                                                    • Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures
                                                                                                      • Temperature
                                                                                                        • Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a
                                                                                                          • DNA Concentration in the ddPCR Solution
                                                                                                            • Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution
                                                                                                              • Equivalence of Assays
                                                                                                                • Table 23 Comparison of Assays
                                                                                                                    • ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials
                                                                                                                      • Table 24 Evaluation of Limiting Ratio
                                                                                                                          • Conclusions and Recommendations
                                                                                                                            • Recommended Certification Values
                                                                                                                              • Table 25 Recommended Values and 95 Uncertainties for Certification
                                                                                                                                • Use of SRM 2372a for Value-Assigning In-House Reference Materials
                                                                                                                                  • Certificate of Analysis
                                                                                                                                  • References
                                                                                                                                  • Appendix A DNA Extraction Method
                                                                                                                                    • Procedure
                                                                                                                                    • Reagents
                                                                                                                                    • Reference
                                                                                                                                      • Appendix B Spectrophotometric Measurement Protocol
                                                                                                                                        • Spectrophotometer Calibration
                                                                                                                                          • Establish Baseline
                                                                                                                                          • Enable Verification of Wavelength Calibration
                                                                                                                                          • Enable Verification of Absorbance Calibration
                                                                                                                                            • For ldquonativerdquo dsDNA Samples
                                                                                                                                              • Verify Optical Balance of Cuvettes
                                                                                                                                                • For NaOH denatured ssDNA Samples
                                                                                                                                                  • Prepare a 04 molL NaOH Stock Solution
                                                                                                                                                  • Prepare the Sample Solutions
                                                                                                                                                  • Measure Sample Solutions
                                                                                                                                                    • After All Sample Measurements are Complete
                                                                                                                                                      • Enable Verification of Spectrophotometer Stability
                                                                                                                                                      • Clean Up
                                                                                                                                                        • Evaluation
                                                                                                                                                        • Reference
                                                                                                                                                          • Appendix C SRM 2372a Statisticianrsquos report
                                                                                                                                                            • 1 Introduction
                                                                                                                                                            • 11 OpenBUGS Analysis
                                                                                                                                                            • 12 OpenBUGS Exemplar Output
                                                                                                                                                            • 2 Homogeneity assessment
                                                                                                                                                            • 21 Data
                                                                                                                                                            • 22 Model
                                                                                                                                                            • 23 OpenBUGS code and data
                                                                                                                                                              • Table C-1 Example OpenBUGS Homogeneity Summary
                                                                                                                                                                • 24 Results
                                                                                                                                                                  • Figure C-1 Homogeneity Assessment
                                                                                                                                                                    • 3 Certification measurements
                                                                                                                                                                    • 31 Data
                                                                                                                                                                    • 32 Model
                                                                                                                                                                    • 33 OpenBUGS code inits and data
                                                                                                                                                                      • Table C-2 Example OpenBUGS Certification Summary for Component A
                                                                                                                                                                      • Table C-3 Example OpenBUGS Certification Summary for Component B
                                                                                                                                                                      • Table C-4 Example OpenBUGS Certification Summary for Component C
                                                                                                                                                                        • 34 Results
                                                                                                                                                                          • Table C-5 Parameter Values for Components A B and C
                                                                                                                                                                          • Table C-6 Certified Values for Components A B and C
                                                                                                                                                                            • 4 Mito-to-nuclear DNA ratios (mtDNAnDNA)
                                                                                                                                                                            • 41 Data
                                                                                                                                                                            • 42 Model
                                                                                                                                                                            • 43 OpenBUGS code inits and data
                                                                                                                                                                              • Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A
                                                                                                                                                                              • Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B
                                                                                                                                                                              • Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C
                                                                                                                                                                              • Table C-10 Results for mtDNAnDNA
                                                                                                                                                                                • 5 References
                                                                                                                                                                                  • Appendix D Performance in Commercial qPCR Chemistries
                                                                                                                                                                                    • Table D-1 Summary [DNA] Results for the Commercial qPCR Assays
Page 11: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

3

Figure 1 Sales History of SRM 2372

Each small black dot represents the sale of one unit of SRM 2372 The red lines summarize the average rate of sales as originally certified (2007 to 2012) and after re-certification (2013 to 2017) The magenta labels state the average yearly sales rate during these two periods

The nDNA copy number concentration human haploid genome equivalents (HHGE) per nanoliter component solution and symbolized here as λacute is calculated as

copy number [DNA] = λ = λ(VF) HHGEnL

where V is the average droplet volume and F is the total dilution factor the volume fraction of the component solution in the ddPCR reaction mixture The manufacturer of the ddPCR platform used a Bio-Rad QX200 Droplet Digital PCR System (Hercules CA) system does not provide metrologically traceable droplet volume information This lack of information prevented metrologically traceable conversion of the number per partition measurements to the desired number per reaction volume units In consequence NIST developed a measurement method that provides metrologically traceable droplet volumes19 We show elsewhere that the mass [DNA] nanograms of human DNA per microliter component can be traceably estimated as

mass [DNA] = 3031timesλacute ngmicroL

where the constant is the product of the number of nucleotide basepairs (bp) in the human reference genome the average bp molecular weight Avogadrorsquos number of bp per mole and

2372

2372(Recertif ied)

050

010

00

1713 unitsyear

1416 unitsyear

2007 2009 2011 2013 2015 2017

Uni

ts S

old

Calendar Year

Human DNA Quantitation StandardSRM 2372

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

4

unit scaling factors18 The three components of SRM 2372a are certified for both copy number and mass [DNA] Mitochondrial DNA (mtDNA) analysis within the forensic community is useful when the nDNA may be degraded or is in low-to-undetectable quantity The challenges of mtDNA testing include contamination of the standard due to routine laboratory use degradation of plasmids or synthetic oligomers or improper commutability of the standard relative to casework samples Commonly cell line materials are used as standards for qPCR with human genomic DNA which may lead to an incorrect mtDNA-to-nDNA ratio (mtDNAnDNA) for unknown samples Non-certified values of mtDNAnDNA for the three components are provided to bridge the current gap between well characterized mitochondrial DNA quantification assays and the availability of commercial standards With the understanding that the forensic DNA community primarily uses qPCR as a method for DNA quantification the performance of the three SRM 2372a materials was evaluated in eleven commercial qPCR chemistries currently used employed within the forensic DNA community

Experimental Methods Production and value-assignment of any certified reference material (CRM) is a complex process In addition to acquiring processing and packaging the materials included in the CRM the materials must be sufficiently homogenous across the units stable over time within its packaging and well-characterized to deliver results that are fit-for-purpose within the intended measurement community20 The process has been particularly complicated for SRM 2372a because we needed to determine the metrological properties of the required ddPCR measurement systems before we could trust the results they provide The following sections describe the methods and processes used to produce and characterize the SRM 2372a materials including sample preparation SRM unit production PCR assays ddPCR measurement processes spectrophotometric absorbance measurements homogeneity testing stability testing ddPCR nDNA and mtDNAnDNA measurements and performance in commercial qPCR assessments Many of these processes were mutually dependent therefore the sections are presented in a ldquodefinitionalrdquo order that sequentially introduces terminology Polymerase Chain Reaction Assays

Ten PCR assays were developed for characterization of SRM 2372a Each assay was confirmed to target one locus per HHGE using the National Center for Biotechnology Informationrsquos BLASTblastn system21 Primers were purchased from Eurofins Operon (Huntsville AL) TaqMan probes labeled with 6-carboxyfluorescein (FAM) were purchased from Thermo Fisher (Waltham MA) FAM-labeled Blackhole Quencher and Blackhole Quencher+ probes were purchased from LGC Biosearch Technologies (Novato CA) Table 1 details the ten nDNA assays used to certify copy number and mass [DNA] of the SRM 2372a components FAM-labeled TaqMan probes were purchased from Thermo Fisher

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

5

and used in monoplex reactions VIC (a proprietary dye)-labeled probes were purchased from the same source and were used when nDNA and mtDNA were analyzed in the same reaction mixture Figure 2 displays exemplar droplet patterns for the assays All assays were developed at NIST and have been optimized for ddPCR18 Table 2 details three mitochondrial DNA (mtDNA) assays used to estimate mtDNAnDNA in the SRM 2372a components Figure 3 displays exemplar droplet patterns for the assays These assays were optimized for ddPCR from real-time quantitative PCR (qPCR) assays from the literature All primers probes were diluted with TE-4 buffer purchased from Affymetrix-USB Products (Affymetrix Inc Cleveland OH) to a nominal 5 micromol working solution for all single-plex assays (individual nDNA and mtDNA assays) For duplex assays (combined nDNA and mtDNA) primer and probe working solutions were diluted to a nominal 10 micromol

Table 1 NIST-Developed Human Nuclear DNA Assays

Assay Target

Chromosome Band Accession Primers and Probe a

Amplicon Length bp

NEIF Gene EIF5B

Chr 2 p111-q111 NC_00000212

F gccaaacttcagccttctcttc R ctctggcaacatttcacactaca PB+ tcatgcagttgtcagaagctg

67

2PR4 Gene RPS27A

Chr 2 p16 NC_00000212

F cgggtttgggttcaggtctt R tgctacaatgaaaacattcagaagtct PB tttgtctaccacttgcaaagctggccttt

97

POTP STR TPOX

Chr 2 p253 NC_00000212

F ccaccttcctctgcttcacttt R acatgggtttttgcctttgg PT caccaactgaaatatg

60

NR4Q Gene DCK

Chr 4 q133-q211 NC_00000412

F tggtgggaatgttcttcagatga R tcgactgagacaggcatatgtt PB+ tgtatgagaaacctgaacgatggt

83

D5 STR D5S2500

Chr 5 q112 NC_00000510

F ttcatacaggcaagcaatgcat R cttaaagggtaaatgtttgcagtaatagat PT ataatatcagggtaaacaggg

75

ND6 STR D6S474

Chr 6 q21-22 NC_00000612

F gcatggctgagtctaaagttcaaag R gcagcctcagggttctcaa PB+ cccagaaccaaggaagatggt

82

D9 STR D9S2157

Chr 9 q342 NC_00000912

F ggctttgctgggtactgctt R ggaccacagcacatcagtcact PT cagggcacatgaat

60

HBB1 Gene HBB

Chr 11 p155 NC_00001110

F gctgagggtttgaagtccaactc R ggtctaagtgatgacagccgtacct PT agccagtgccagaagagccaagga

76

ND14 STR D14S1434

Chr 14 q3213C NC_0000149

F tccaccactgggttctatagttc R ggctgggaagtcccacaatc PB+ tcagactgaatcacaccatcag

109

22C3 Gene PMM1

Chr 22 q132 NC_00002210

F cccctaagaggtctgttgtgttg R aggtctggtggcttctccaat PB caaatcacctgaggtcaaggccagaaca

78

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

6

a F Forward primer R Reverse primer PB Blackhole quencher probe (FAM labeled) PB+ Blackhole Plus quencher probe (FAM labeled) PT TaqMan MGB probe (FAM labeled)

Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays

Table 2 NIST-Optimized Human Mitochondrial DNA Assays

Assay Location Primers and Probe a Amplicon Length bp Ref

mtND1 34853504 F ccctaaaacccgccacatct

69 22 35333553 R gagcgatggtgagagctaaggt 35063522 PT ccatcaccctctacatc

mtBatz 84468473 F aatattaaacacaaactaccacctacct

79 23 85038524 R tggttctcagggtttgttataa 84758491 PT cctcaccaaagcccata

mtKav 1318813308 F ggcatcaaccaaccacaccta

105 24 1336913392 R attgttaaggttgtggatgatgga 1331013325 PT cattcctgcacatctg

a F Forward primer R Reverse primer PT TaqMan MGB probe (FAM labeled)

2PR4

POTP

NEI

F

R4Q

5

D5

ND6

D9

HBB1

ND1

4

22C3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

7

Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays

Droplet Digital Polymerase Chain Reaction Measurements

Nuclear DNA (nDNA) All nDNA ddPCR measurements used in this study followed the same measurement protocol

bull Droplets were generated on the Auto Droplet Generator using the Bio-Rad ldquoddPCR ldquoSupermix for Probes (No dUTP)rdquo (Bio-Rad cat 186-3024 Control 64079080) and ldquoDroplet Generation Oil for Probesrdquo (Bio-Rad cat 186-4110 Control 64046051) Note desoxyuridine 5-triphosphatase (dUTP) is an enzyme sometimes included in PCR reaction mixtures to minimize PCR carryover contamination

bull Table 3 details the composition of the ddPCR mastermix setup for a given assay and sample where ldquomastermix refers to reaction mixture of manufacturerrsquos proprietary PCR reagent ldquoSupermixrdquo forward and reverse PCR primers for a given assay probe for that assay the sample DNA and sometimes additional water to produce the desired value dilution factor F This setup was used for all assays and samples

bull Non-Template Controls (NTCs) for each assay were included in every analysis bull Droplets were thermal cycled on a ProFlex thermal cycler (Applied Biosystems) for

95 degC for 10 min followed by 60 cycles of 94 degC for 05 min and 61 degC for 1 min then 98 degC for 10 min before a 4 degC hold until the plate was removed

bull Droplets were read on the QX200 Droplet Reader and analysis was performed using the QuantaSoft Analysis Software version 1740917

mtND1 mtBatz mtKav

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

8

bull The numbers of positive and negative droplets at the end of 60 cycles were determined and were exported into a spreadsheet for further analysis Assay-specific intensity thresholds were determined by visual inspection for each assay for each measurement session

Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays

Concentration nDNA Mastermix Microliter

per Reaction 2X ldquoSupermix for Probes (no dUTPs)rdquo 125

5 μmolL Forward Primer 1875 5 μmolL Reverse Primer 1875 5 μmolL Probe (FAM) 125

PCR Grade Water 50 14 Diluted Sample 25 Total Volume 250

Mitochondrial DNA (mtDNA) All the mtDNA measurements used in this study followed the nDNA protocol described above but with different mastermix setups The duplex mtDNA measurements used the setup described in Table 4 The monoplex mtDNA measurements used the setup described in Table 5

Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 125

10 μmolL Forward Primer 09375 10 μmolL Reverse Primer 09375 10 μmolL Probe (VIC) 0625

PCR Grade Water 95 nDNA Mastermix from Table 3 05 Total Volume 250

Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays

Concentration mtDNA Mastermix Microliter

per Reaction 2X Supermix for Probes (no dUTPs) 120

10 μmolL Forward Primer 09 10 μmolL Reverse Primer 09 10 μmolL Probe (FAM) 06

PCR Grade Water 72 Diluted Sample (See Table 7) 24

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

9

Total Volume 240 Sample Preparation

The buffy coat white blood cells used to prepare SRM 2372a materials were purchased from Interstate Blood Bank These materials were from four single-source anonymous individuals supplied to us labeled as Female1 Female2 Male1 and Male2 In our laboratories each donor buffy coat bag was aliquoted (5 mL per aliquot) into sterile 50 mL conical tubes and stored at 4 degC prior to extraction The modified salt-out manual extraction protocol described in Appendix A was performed for each aliquot with rehydration of the DNA in TE-4 buffer From 150 mL to 250 mL of solubilized DNA for each component was accumulated through the extraction of the 5 mL aliquots The DNA was stored in perfluoroalkoxy polymer (PFA) containers that had been bleach sterilized thoroughly rinsed with deionized water rinsed with ethanol and air-dried The minimum volume for production of 2000 units was 120 mL Preliminary estimates of the mass [DNA] of the re-solubilized components were obtained using a Nanodrop microvolume spectrophotometer Dilutions of the component materials were made until the absorbance at 260 nm was approximately 1 corresponding to a nominal mass [DNA] of 50 ngμL of dsDNA7 After addition of the TE-4 buffer the containers were placed on an orbital shaker and gently rotated for several hours prior to refrigerated storage at 4 ordmC The four single-source components were initially screened with the nDNA ddPCR measurement process using four nDNA assays 2PR4 NR4Q ND6 and 22C3 Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate The gravimetric 1rarr4 dilution was prepared by weighing an empty PFA vial and recording the weight then pipetting 30 microL TE-4 into the vial and recording the weight then pipetting 10 microL DNA and recording the final weight The dilution factor was calculated in a spreadsheet These screening measurements verified that the individual component materials were fit-for-purpose The four single-source components were evaluated with the nDNA ddPCR measurement process using the ten nDNA assays Samples of the four single-source materials were gravimetrically prepared as 1rarr4 dilutions of the bulk materials followed by a 1rarr10 volumetric dilution and run in duplicate Quantitative information was derived from the average of the ten nDNA assays to prepare the component C mixture Thirty milliliters of sample Male2 (estimated at 452 ngμL mass [DNA]) were combined with 80 mL of sample Female2 (estimated at 507 ngμL mass [DNA]) and 10 mL of TE-4 buffer to create a 1 male nDNA to 3 female nDNA producing a 14 malehuman mixture After addition of the TE-4 buffer the component C container was gently rotated on the orbital shaker for several hours prior to refrigerated storage at 4 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

10

Absorbance Measurements

In February 2017 absorbance spectra of the recently prepared fully diluted nominally dsDNA and ssDNA solutions of the three components were acquired with a BioCary 100 spectrophotometer using the measurement protocol described in Appendix B In December 2017 additional spectra were acquired for the remaining bulk solution of components A and B and for SRM 2372 components A and B These solutions had been stored at 4 degC in well-sealed PFA containers There was no remaining bulk component C solution SRM Unit Production

In the morning of the day a component was transferred into the sterile component tubes (Sarstedt Biosphere SC micro tube-05 mL Newton NC) used to package the SRM the container holding the diluted component was removed from the refrigerator and placed on a slow orbital shaker for 2 h The tubes were opened and placed under the laminar flow hood in rows of 5 within 80-hole tube racks for a total of 30 tubes per rack Tubes were placed in every second row of the racks to facilitate filling with an Eppendorf Repeater Xstream pipette (Eppendorf North America Inc Hauppauge NY) fitted with a 1-mL positive displacement tip set to dispense 55 microL per tube The component container was removed from the shaker a magnetic stir-bar was added the container was placed on a magnetic stir plate in the laminar flow hood and the material was stirred gently Filling proceeded until 2200 vials were filled with pipet tip replacement after filling every 200 tubes The filling process consisted of two individuals manually filling each vial within a biosafety cabinet (one physically pipetting each sample the other verifying that each tube was filled) and four individuals tightly closing the lids to each tube on a sterilized and covered bench When all tubes in a rack were filled the rack was moved out of the hood the tubes checked for proper filling their lids closed and the tubes transferred to 100-unit pre-labeled storage boxes As the 100-unit boxes were filled they were transferred to the labeling room where SRM labels were applied by hand and colored inserts were added to the lids red for component A white for component B and blue for component C After all tubes for a given component were labeled the 22 boxes were placed in a locked refrigerator Filling of each of the components took 60 min to 90 min All units were transferred to the refrigerator within 4 hours of beginning the filling process All units of the three materials were allowed to equilibrate for a minimum of two weeks prior to beginning homogeneity analysis Component Homogeneity

The ddPCR measurement protocol described above was used to assess the nDNA homogeneity of the three components One tube from the approximate center of each of the 22 storage boxes was assayed using the NEIF and ND6 assays (see Table 1) A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Three technical replicates were obtained for every tube A total

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

11

of six plates were used two plates per component one with the NEIF assay and one with the ND6 assay Figure 4 details the layout of the samples

Figure 4 Plate Layout Design for Homogeneity Measurements

Component Stability

Stability of the materials was evaluated at 4 degC room temperature (RT asymp 22 degC) and 37 degC Tubes from the homogeneity study were divided among the three temperatures for the stability study Additionally two never-opened tubes (one from box 8 and one from box 12 of each component) were held at each temperature Two tubes at each temperature for each component were opened sampled and resealed 4 times over the course of 8 weeks Measurements were made at week 1 week 3 week 5 and week 8 The measurement on week 8 was from an unopened tube (from box 8) for each of the components at all temperatures The data from the homogeneity study was used as timepoint 0 in the stability study anchored to 4 degC by the storage recommendations A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used for both the NEIF and ND6 assays All temperatures components and both assays fit onto one plate for timepoints 1 to 5 Twelve NTCs placed in rows G and H were included for each assay Figure 5 details the sample layout

Figure 5 Plate Layout Design for Stability Measurements

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12A A-1 A-1 A-1 A-9 A-9 A-9 A-17 A-17 A-17 A B-1 B-1 B-1 B-9 B-9 B-9 B-17 B-17 B-17 A C-1 C-1 C-1 C-9 C-9 C-9 C-17 C-17 C-17B A-2 A-2 A-2 A-10 A-10 A-10 A-18 A-18 A-18 B B-2 B-2 B-2 B-10 B-10 B-10 B-18 B-18 B-18 B C-2 C-2 C-2 C-10 C-10 C-10 C-18 C-18 C-18C A-3 A-3 A-3 A-11 A-11 A-11 A-19 A-19 A-19 C B-3 B-3 B-3 B-11 B-11 B-11 B-19 B-19 B-19 C C-3 C-3 C-3 C-11 C-11 C-11 C-19 C-19 C-19D A-4 A-4 A-4 A-12 A-12 A-12 A-20 A-20 A-20 D B-4 B-4 B-4 B-12 B-12 B-12 B-20 B-20 B-20 D C-4 C-4 C-4 C-12 C-12 C-12 C-20 C-20 C-20E A-5 A-5 A-5 A-13 A-13 A-13 A-21 A-21 A-21 E B-5 B-5 B-5 B-13 B-13 B-13 B-21 B-21 B-21 E C-5 C-5 C-5 C-13 C-13 C-13 C-21 C-21 C-21F A-6 A-6 A-6 A-14 A-14 A-14 A-22 A-22 A-22 F B-6 B-6 B-6 B-14 B-14 B-14 B-22 B-22 B-22 F C-6 C-6 C-6 C-14 C-14 C-14 C-22 C-22 C-22G A-7 A-7 A-7 A-15 A-15 A-15 NTC NTC NTC G B-7 B-7 B-7 B-15 B-15 B-15 NTC NTC NTC G C-7 C-7 C-7 C-15 C-15 C-15 NTC NTC NTCH A-8 A-8 A-8 A-16 A-16 A-16 NTC NTC NTC H B-8 B-8 B-8 B-16 B-16 B-16 NTC NTC NTC H C-8 C-8 C-8 C-16 C-16 C-16 NTC NTC NTC

ND6 ND6 ND6

NEIFNEIFNEIFComponent A

Component A

Component B

Component B

Component C

Component C

1 2 3 4 5 6 7 8 9 10 11 12A A-7 A-7 B-7 B-7 C-7 C-7 A-7 A-7 B-7 B-7 C-7 C-7B A-10 A-10 B-10 B-10 C-10 C-10 A-10 A-10 B-10 B-10 C-10 C-10C A-9 A-9 B-9 B-9 C-9 C-9 A-9 A-9 B-9 B-9 C-9 C-9D A-12 A-12 B-12 B-12 C-12 C-12 A-12 A-12 B-12 B-12 C-12 C-12E A-8 A-8 B-8 B-8 C-8 C-8 A-8 A-8 B-8 B-8 C-8 C-8F A-11 A-11 B-11 B-11 C-11 C-11 A-11 A-11 B-11 B-11 C-11 C-11G NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTCH NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

NEIF

4C

RT

37C

ND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

12

Component Certification Measurements

After establishing homogeneity and stability measurements intended to be used to certify nDNA values and inform mtDNA values were started using randomly selected SRM tubes from the boxes listed in Table 7 The ten nDNA assays described in Table 1 and two of the three mtDNA assays described in Table 2 were run simultaneously on the same sample plate A sample from each tube was gravimetrically diluted 1rarr4 with TE-4 followed by a volumetric 1rarr10 dilution into the mastermix Two technical replicates were used in each of the six measurement campaigns

Table 6 Tubes used for Certification Measurements

Component Plate A B C

dd20170711 Box 1 Box 1 Box 1 dd20170712a Box 10 tube a Box 10 tube a Box 10 tube a dd20170712b Box 10 tube b Box 10 tube b Box 10 tube b dd20170713 Box 20 Box 20 Box 20 dd20170714 Box 5 Box 5 Box 5 dd20170801 Box 5 Box 5 Box 5

Every sample was analyzed in duplicate with every assay in all six plates Component A was always located in rows B and C component B in rows D and E component C in rows F and G and NTCs in rows A and H Figure 6 displays the basic layout of the ten nDNA and two mtDNA assays In addition to the plate column assignment scheme shown the three assay groupings were also ordered as Group 2 Group 3 Group 1 and Group 3 Group 1 Group 2 to address potential positional bias

Group 1 Group 2 Group 3 2PR4 POTP NEIF NR4Q D5 ND6 D9 HBB1 ND14 22C3 mtND1 mtBatz 1 2 3 4 5 6 7 8 9 10 11 12 A NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Comp A B Comp A C Comp B D Comp B E Comp C F Comp C G

H NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC NTC

Figure 6 Basic Plate Layout Design for Certification Measurements

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

13

Mitochondrial DNA Measurements

A ldquoduplexed reactionrdquo was used to determine mtDNAnDNA where 1 μL of one of the nDNA mastermix was added to 49 μL of the mtDNA mastermix resulting in a 1rarr50 volumetric dilution of the nDNA The nDNA was targeted with a FAM-labeled probe while the mtDNA was targeted with a VIC labeled probe The primer and probe concentrations were increased from 5 micromol to 10 micromol in the duplexed reactions Additional testing of the mtDNA assays were performed in monoplex with FAM-labeled probes to determine the additional sources of bias between sampling assay setup and dilution factors Table 7 outlines the different methods examined to determine mtDNAnDNA for the three components

Table 7 Run Parameters for Mitochondrial DNA Droplet

Generation nDNA mtDNA

Plate Component Assay Dilution Reps Assay Dilution Reps dd20170824 A manualDG HBB1 1rarr4 3 mtND1 1rarr200 9 dd20170825 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170829 ABC manualDG HBB1 1rarr4 4 mtBatz 1rarr10rarr100 4 dd20170831 AB autoDG HBB1 1rarr3 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr6 2 mtND1 1rarr10rarr100 3 dd20170831 AB autoDG HBB1 1rarr12 2 mtND1 1rarr10rarr100 3 dd20170907 ABC manualDG HBB1 1rarr3 2 mtBatz 1rarr10rarr100 3 dd20170913 AB manualDG NEIF 1rarr4 3 mtND1 mtBatz mtKav 1rarr10rarr100 3 dd20170913 C manualDG NEIF 1rarr4 3 mtBatz mtKav 1rarr10rarr100 3 Performance in Commercial qPCR Chemistries

Table 8 lists the 11 qPCR chemistries that were commercially available in late 2017 All three of the SRM 2372a components were evaluated with these chemistries using component A of SRM 2372 as the reference standard For each component of SRM 2372a four tubes were pulled from box 22 and combined into one tube A 1rarr5 dilution of all samples was prepared followed by three serial 1rarr2 dilutions A five-point serial dilution of each SRM 2372a component was amplified in triplicate for each chemistry This consisted of the neat material the 1rarr5 dilution then the three serial 1rarr2 dilutions A six-point serial dilution of SRM 2372 Component A was also created for the standard curve for the commercial qPCR chemistries This consisted of the neat material followed by a 1rarr5 dilution then followed by four 1rarr2 dilutions serially The six-point serial dilution series of SRM 2372 Component A was also amplified in triplicate These commercial chemistries were amplified in individual plates on an Applied Biosystems 7500 HID Real-Time PCR Instrument (Foster City CA) following the manufacturersrsquo recommended protocols for reaction mix setup and cycling conditions Figure 7 details the plate layout for the qPCR plates Data was analyzed with the HID Real-Time PCR Analysis Software v12 and further analyzed in a spreadsheet

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

14

Table 8 Commercial qPCR Chemistries Tested with SRM 2372a

Manufacturer Commercial Kit Target Codea

Innogenomics

InnoQuant Long IQh Short IQd

InnoQuantHY Long IQHYh Short IQHYd Y IQHYy

Thermo Fisher Quantifiler Duo Human QFDh Male QFDy

Quantifiler HP Large QFHPh Small QFHPd

Quantifiler Human Human QFHh

Quantifiler Trio Large QFTh Small QFTd Y QFTy

Qiagen

Quantiplex Autosomal QPh

Quantiplex Hyres Autosomal QPHh Y QPHy

Quantiplex Pro Human QPPh Degradation QPPd Y QPPy

Promega

Plexor Autosomal Ph Y Py

PowerQuant Autosomal PQh Degradation PQd Y PQy

a Assays are coded with a short acronym derived from the kit name followed by the following codes for the type of target ldquohrdquo autosomal non-degradation described as ldquoAutosomalrdquo ldquoHumanrdquo or ldquoLongrdquo ldquodrdquo autosomal degradation described as ldquoDegradationrdquo or ldquoShortrdquo ldquoyrdquo Y-chromosome described as ldquoMalerdquo or ldquoYrdquo

1 2 3 4 5 6 7 8 9 10 11 12

A

B 2372-A_57 2372-A_57 2372-A_57 2372a-neat_A

2372a-neat_A

2372a-neat_A

2372a-neat_B

2372a-neat_B

2372a-neat_B

2372a-neat_C

2372a-neat_C

2372a-neat_C

C 2372-A_11

4 2372-A_11

4 2372-A_11

4 2372a-15_A

2372a-15_A

2372a-15_A

2372a-15_B

2372a-15_B

2372a-15_B

2372a-15_C

2372a-15_C

2372a-15_C

D 2372-A_5

7 2372-A_5

7 2372-A_5

7 2372a-110_A

2372a-110_A

2372a-110_A

2372a-110_B

2372a-110_B

2372a-110_B

2372a-110_C

2372a-110_C

2372a-110_C

E 2372-A_2

85 2372-A_2

85 2372-A_2

85 2372a-120_A

2372a-120_A

2372a-120_A

2372a-120_B

2372a-120_B

2372a-120_B

2372a-120_C

2372a-120_C

2372a-120_C

F 2372-A_1

425 2372-A_1

425 2372-A_1

425 2372a-140_A

2372a-140_A

2372a-140_A

2372a-140_B

2372a-140_B

2372a-140_B

2372a-140_C

2372a-140_C

2372a-140_C

G 2372-A_0

7125 2372-A_0

7125 2372-A_0

7125

H NTC NTC NTC

Figure 7 Plate Layout Design for Performance Assessments

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

15

Measurement Results After preparing the SRM 2372a bulk solutions their approximate 50 ngmicroL mass [DNA] were confirmed spectrophotometrically The bulk solutions were then packaged into individual tubes The materials contained in the tubes were characterized for homogeneity stability nDNA copy number [DNA] and mtDNAnDNA ratio The following sections detail the results of these measurements Absorbance Spectrophotometry

By widely accepted convention an aqueous DNA solution having an absorbance of 10 at 260 nm corresponds to a mass [DNA] of 50 ngmicroL for dsDNA and (37 to 40) ngmicroL for ssDNA7 Absorbance at (230 270 280 and 330) nm are traditionally used in the assessment of DNA quality725 Due to minor baseline shifts during acquisition of some spectra it was necessary to bias-adjusted all spectra to have absorbance of 00 at 330 nm Table 9 lists the resulting bias-adjusted values for selected wavelengths and the conventional mass [DNA] conversion of the absorbance at 260 nm The close agreement between the mass concentration values for the nominally dsDNA and the converted-to-ssDNA spectra confirm that the bulk solutions were correctly prepared to be asymp 50 ngμL dsDNA The slight increase in absorbance over the 10 months between acquiring the spectra suggests that like SRM 2372 the tertiary structure of the dsDNA in the SRM 2372a materials may be changing slowly with time Figure 8 displays the spectra

Table 9 Absorbance at Selected Wavelengths

Wavelength nm Form Date Component 230 260 270 280 ngμL dsDNA 2142017 A 0394 0931 0758 0488 466

B 0480 1112 0905 0583 556 C 0451 1006 0815 0524 503

ssDNA 2142017 A 0390 0614 0540 0316 454 B 0457 0723 0634 0372 535 C 0411 0641 0563 0330 474

dsDNA 12142017 A 0402 0956 0775 0503 478 B 0484 1128 0914 0589 564

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

16

Figure 8 Absorbance Spectra of SRM 2372a Components A B and C

dsDNA is denatured to ssDNA by diluting with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to DNA mass concentration

Droplet Volume

Droplet volumes for the two lots of ldquoSupermixrdquo used in the studies described in this report were determined using NISTrsquos Special Test Method 11050S-D as described in Dagata et al19 The lot used for most of the ddPCR analyses produced (07349 plusmn 00085) nL droplets where the ldquoplusmnrdquo value is the combined standard uncertainty The lot used for analyses performed after October 11 2017 produced (07376 plusmn 00085) nL droplets NIST measurements have demonstrated that droplet volumes are not influenced by the presence or absence of DNA in the sample solution Measurements repeated over time on an earlier lot of the Supermix product used in these studies suggest that droplet volumes remain constant over the shelf-life of the lot However droplet volumes for different lots of this product as well as for different Supermix products have been observed to differ by several percent We intend to document these measurements in a future report Other researchers have demonstrated that non-constant droplet volume distributions may contribute to ddPCR measurement imprecision and bias2627 However these effects become significant only with distributions wider than observed using Special Test Method 11050S-D19 and at copydroplet values larger than the λ le 05 copydroplet typically used in our measurements Certification Measurements

Homogeneity Results Table 10 lists the λacute nDNA copy number values adjusted for both droplet volume and sample dilution from the homogeneity measurements The data was analyzed using the three-level Gaussian hierarchical ANOVA model described in Appendix C Table 11 lists the values of the within- and between-tube variance components expressed as percent relative standard deviations At the 95 confidence level all components are homogeneous

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372a-A2141721417-NaOH121417

220 260 300 340

Wavelength nm

2372a-B2141721417-NaOH121517

220 260 300 340

Wavelength nm

2372a-C21417

21417-NaOH

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

17

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 NEIF A-1-01 1632 1529 1511 B-1-01 1790 1909 1885 C-1-01 1493 1385 1510 ND6 1616 1640 1582 NAe 1870 1860 1565 1562 1466

NEIF A-1-02 1531 1525 1549 B-1-02 1712 1719 1704 C-1-02 1516 1513 1577 ND6 1615 1589 1601 1775 1718 1756 1490 1547 1570

NEIF A-1-03 1524 1540 1533 B-1-03 1693 1755 1749 C-1-03 1451 1528 1506 ND6 1582 1152 1555 1824 1483 1896 1485 1505 1471

NEIF A-1-04 1454 1458 1427 B-1-04 1754 1745 1739 C-1-04 1440 1486 1520 ND6 1627 1605 1607 1670 1632 1814 1529 1508 1506

NEIF A-1-05 1465 1540 1504 B-1-05 1757 1687 1735 C-1-05 1523 1472 1483 ND6 1607 1633 1586 1683 1759 1703 1513 1520 1491

NEIF A-1-06 1512 1491 1491 B-1-06 1697 1709 1714 C-1-06 1437 1493 1522 ND6 1511 1483 1524 1693 1723 1722 1492 1481 1461

NEIF A-1-07 1418 1480 1455 B-1-07 1722 1625 1703 C-1-07 1485 1477 1463 ND6 1479 1557 1656 1690 1684 1688 1475 1547 1485

NEIF A-1-08 1496 1560 1501 B-1-08 1884 1802 1772 C-1-08 1433 1405 1478 ND6 1514 1615 1540 1864 1887 1821 1497 1501 1495

NEIF A-1-09 1588 1581 1637 B-1-09 1713 1734 1727 C-1-09 1463 1499 1505 ND6 1474 1438 1454 1709 1756 1758 1503 1491 1516

NEIF A-1-10 1574 1513 1570 B-1-10 1759 1809 1738 C-1-10 1534 1458 1454 ND6 1594 1713 1669 1764 1816 1787 1448 1465 1474

NEIF A-1-11 1557 1508 1633 B-1-11 1749 1782 1740 C-1-11 1533 1514 1555 ND6 1571 1611 1631 1741 1792 1784 1488 1470 1483

NEIF A-1-12 1500 1476 1515 B-1-12 1748 1712 1718 C-1-12 1450 1525 1497 ND6 1626 1586 1608 1797 1614 1790 1496 1483 1519

NEIF A-1-13 1672 1634 1804 B-1-13 1632 1668 1701 C-1-13 1537 1527 1515 ND6 1490 1516 1497 1626 1566 1681 1499 1479 1475

NEIF A-1-14 1551 1568 1538 B-1-14 1752 1535 1673 C-1-14 1515 1490 1494 ND6 1617 1614 1639 1724 1741 1685 1469 1480 1466

NEIF A-1-15b 1793 1722 1733 B-1-15 1871 1835 1887 C-1-15 1511 1568 1594 ND6 1818 1753 1762 1770 1752 1798 1515 1543 1565

NEIF A-1-16 1573 1520 1617 B-1-16 1688 1725 1743 C-1-16 1449 1452 1471 ND6 1634 1627 1575 1744 1697 1710 1543 1448 1491

NEIF A-1-17 1516 1516 1628 B-1-17 1789 1769 1825 C-1-17 1388 1421 1394 ND6 1613 1567 1567 1716 1762 1638 1452 NAe 1435

NEIF A-1-18 1612 1535 1672 B-1-18 1826 1780 1849 C-1-18 1475 1455 1445 ND6 1529 1495 1528 1805 1801 1931 1458 1531 1459

NEIF A-1-19 1598 1514 1596 B-1-19 1751 1759 1880 C-1-19 1468 1525 1446 ND6 1571 1535 1511 1827 1766 1786 1435 1478 1421

NEIF A-1-20 1575 1568 1586 B-1-20 1738 1678 1768 C-1-20 1549 1531 1499 ND6 1652 1545 1627 1752 1747 1768 1499 1465 1468

NEIF A-1-21 1658 1611 1578 B-1-21 1742 1836 1713 C-1-21 1530 1440 1484 ND6 1647 1607 1612 1746 1744 1764 1499 1454 NAe

NEIF A-1-22 1547 1528 1532 B-1-22 1761 1743 1660 C-1-22 1442 1487 1436 ND6 1537 1513 1536 1666 1692 1819 1589 1591 1556

NEIF A-1-15c 1456 1448 1523 ND6 1525 1402 1436

NEIF A-2-15d 1679 1599 1674 ND6 1581 1550 1584

a Tube coded as Component-Tube-Box b These results were identified as a probable error in the preparation or volumetric delivery of the sample

solution into the microfluidic device and were not used in the homogeneity evaluations c Repeat assessment of the solution in sample A-1-15 d Results for a second tube from box 15 analyzed at the same time as the repeat assessment of A-1-15 e Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

18

Table 11 Within- and Between-Tube Relative Standard Deviations

Within-Tube Component NEIF ND6 Between-Tube

A 28 40 33 B 27 38 25 C 23 18 19

Stability Results Table 12 lists the λacute volume- and dilution-adjusted nDNA copy number values from the stability measurements Figure 9 summarizes these results The results for the tubes held at 4 degC room temperature (RT asymp 22 degC) and 37 degC are very similar for all three components The absence of between-temperature differences identifies the between-week differences as attributable to variation in the measurement processes rather than sample instability These results indicate that the SRM 2372a genomic copy number is thermally stable from 4 degC to 37 degC over an extended period

Figure 9 ddPCR Stability Measurements as Function of Time

Each dot-and-bar symbol represents the mean λacute and the approximate 95 expanded uncertainty for all results for a given component at a given storage temperature The thick black horizontal lines represent an approximate 95 confidence interval on the λacute values estimated from the homogeneity data

13

14

15

16

17

18

19

0 1 2 3 4 5 6 7 8

λacute H

aplo

id n

DNA

Copi

es μL

Weeks

A

0 1 2 3 4 5 6 7 8

Weeks

B

4 degC 22 degC 37 degC

0 1 2 3 4 5 6 7 8

Weeks

C

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

19

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Week Temp Assay Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3 Tubea Rep1 Rep2 Rep3

1 4 degC NEIF A-10 1568 1441 B-10 1666 1653 C-10 1424 1394 ND6 A-10 1474 1494 B-10 1655 1705 C-10 1461 1394

NEIF A-07 1500 1463 B-07 1718 1640 C-07 1446 1452 ND6 A-07 1479 1479 B-07 1708 1686 C-07 1522 1482

RT (22 degC) NEIF A-12 1674 1608 B-12 1805 1777 C-12 1506 1495 ND6 A-12 1522 1507 B-12 1759 1788 C-12 1478 1461 NEIF A-09 1635 1539 B-09 1791 1793 C-09 1411 NAb ND6 A-09 1543 1519 B-09 1661 1682 C-09 1482 1496

37 degC NEIF A-11 1614 1637 B-11 1748 1744 C-11 1515 1464 ND6 A-11 1494 1502 B-11 1740 1697 C-11 1495 1535 NEIF A-08 1592 1614 B-08 1691 1725 C-08 1528 1500 ND6 A-08 1581 1595 B-08 1727 1699 C-08 1458 1483 3 4 degC NEIF A-10 1520 1706 B-10 NAb 1733 C-10 1427 1473 ND6 A-10 1543 1624 B-10 1670 1755 C-10 1443 1364 NEIF A-07 1537 1570 B-07 1549 1565 C-07 1424 1512 ND6 A-07 1517 1479 B-07 1705 1723 C-07 1483 1439

RT (22 degC) NEIF A-12 1521 1607 B-12 1817 1842 C-12 1566 1471 ND6 A-12 1644 1597 B-12 1845 1786 C-12 1504 1511 NEIF A-09 1432 1626 B-09 1825 1856 C-09 1392 1387 ND6 A-09 1539 1528 B-09 1734 1737 C-09 1535 1465

37 degC NEIF A-11 1651 1685 B-11 1759 1779 C-11 1437 1440 ND6 A-11 1680 1594 B-11 1806 1806 C-11 1471 1509 NEIF A-08 1580 1613 B-08 1740 1791 C-08 1582 1590 ND6 A-08 1592 1524 B-08 1769 1694 C-08 1468 1520 5 4 degC NEIF A-10 1486 1440 B-10 1626 1551 C-10 1381 1330 ND6 A-10 1555 1577 B-10 1673 1625 C-10 1547 1395 NEIF A-07 1387 1429 B-07 1723 1787 C-07 1330 1314 ND6 A-07 1538 1522 B-07 1862 1909 C-07 1452 1458

RT (22 degC) NEIF A-12 1601 1631 B-12 1718 1752 C-12 1546 1541 ND6 A-12 1699 1677 B-12 1772 1721 C-12 1490 1471 NEIF A-09 1381 1484 B-09 1675 1694 C-09 1371 1376 ND6 A-09 1625 1615 B-09 1904 1815 C-09 1468 1472

37 degC NEIF A-11 1456 1517 B-11 1720 1704 C-11 1505 1493 ND6 A-11 1650 1658 B-11 1148 1152 C-11 1527 1600 NEIF A-08 1590 1666 B-08 1745 1723 C-08 1469 1540 ND6 A-08 1235 1244 B-08 1809 1749 C-08 1508 1529 8 4 degC NEIF A-08a 1667 1632 1589 B-08a 1880 1884 1940 C-08a 1525 1475 1542 ND6 A-08a 1599 1593 1590 B-08a 1859 1897 1972 C-08a 1550 1502 1520

RT (22 degC) NEIF A-08b 1693 1666 1634 B-08b 1769 1884 1771 C-08b 1595 1557 1545 ND6 A-08b 1644 1620 1647 B-08b 1862 1738 1811 C-08b 1611 1585 1535

37 degC NEIF A-08c 1620 1576 1608 B-08c 1934 1952 1956 C-08c 1595 1600 1569 ND6 A-08c 1576 1553 1532 B-08c 1889 1844 1889 C-08c 1569 1514 1581

a Tube coded as Component-Tube-Box b Value not available due to a technical failure by either the droplet generator or droplet reader

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

20

Ten-Assay Certification Results The original intention was to certify the copy number [DNA] of the three components with results from five independently prepared sets of samples analyzed over four days Each sample of the three components was from a new tube from different boxes and was evaluated with 10 nDNA assays The assay positions were alternated across these five ldquo10-assay certificationrdquo plates The λ results for these five plates are systematically somewhat larger than those for the homogeneity and stability data On review we realized that a different workspace and its suite of pipettes had been used to prepare the 1rarr10 dilution of the 1rarr4 diluted components into the reaction mastermix The ldquohighrdquo workstation was used for stability and homogeneity measurements while the ldquolowrdquo workstation was used for certification measurements Gravimetric testing of the pipettes established that the volumes delivered by these alternate pipettes was sufficiently biased to account for the observed λ differences A sixth ldquocertificationrdquo plate was prepared using the original (and now re-verified) pipettes The λ results or this plate agreed well with the homogeneity and stability data Table 13 lists the λacute volume- and dilution-adjusted nDNA copy number values for all six of the ten-assay datasets Figure 10 displays the extent of agreement among the ten assays While there are small-but-consistent differences between the assays there is only a 4 difference between the assays providing smallest and largest λ

Figure 10 Comparison of Assay Performance

The blue dot-and-bars represent the mean and its combined standard uncertainty for the A B and C components for each assay relative to the over-all grand mean for each component The black dot-and-bars represent combination of three component results for each assay The solid horizontal line represents the median difference of the assay means from the grand mean the dotted horizontal lines bound an approximate 95 confidence region about the median

2PR4POTPNEIFR4Q5 D5ND6 D9 HBB1ND14 22C3-10

-8

-6

-4

-2

0

2

4

6

8

10

Rela

tive

Diffe

renc

e fr

om M

ean

Assays

Component MeansAssay MeansMedianU95

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

21

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

7112017 2PR4 1667 1698 1869 1960 1630 1706 POTP 1655 1685 1850 1781 1596 1576 NEIF 1647 1665 1709 1707 1574 1538 R4Q5 1594 1658 1856 1787 1590 1555 D5 1655 1683 1848 1829 1610 1607 ND6 1629 1647 1828 1826 1589 1548 D9 1655 1652 1861 1858 1569 1572 HBB1 1644 1611 1825 1869 1565 1681 ND14 1645 1707 1871 1773 1564 1492 22C3 1545 1603 1776 1902 1548 1655

7122017 2PR4 1668 1739 2174 2231 1499 1497 POTP 1587 1702 2202 2292 1576 1611 NEIF 1678 1614 2172 2216 1545 1560 R4Q5 1628 1762 NA 2078 1584 1546 D5 1738 1649 2230 2192 1577 1529 ND6 1667 1678 2094 2034 1599 1478 D9 1695 1653 1949 2110 1587 1504 HBB1 1757 1750 1971 2123 1520 1551 ND14 1771 1764 2123 2066 1479 1438 22C3 1762 1759 2107 2046 1512 1494

7122017 2PR4 1638 1667 1810 1806 1452 1424 POTP 1676 1638 1865 1801 1347 1397 NEIF 1616 1624 1806 1786 1395 1350 R4Q5 1559 1587 1878 1768 1442 1459 D5 1608 1652 1824 1790 1448 1602 ND6 1611 1589 1768 1670 1463 1435 D9 1517 1607 1811 1749 1402 1433 HBB1 1557 1582 1763 1741 1474 1481 ND14 1604 1625 1865 1781 1466 1482 22C3 1675 1616 1815 1866 1439 1568

7132017 2PR4 1888 1939 1784 1821 1470 1444 POTP 1868 1815 1735 1813 1411 1427 NEIF 1953 1902 1778 1813 1512 1439 R4Q5 1779 1871 1835 1749 1452 1468 D5 1924 2124 1794 1848 1567 1524 ND6 1861 1787 1700 1711 1480 1424 D9 1824 1844 1799 1779 1442 1459 HBB1 1809 1799 1765 1769 1430 1467 ND14 1864 1857 1754 1725 1413 1411 22C3 1930 1942 1792 1814 1486 1480

7142017 2PR4 1597 1590 1718 1768 1485 1458 POTP 1533 1404 1753 1698 1422 1459 NEIF 1462 1533 1773 1792 1438 1465 R4Q5 1493 1519 1777 1740 1439 1468 D5 1538 1586 1803 1795 1515 1457 ND6 1492 1516 1718 1693 1409 1412 D9 1528 1538 1802 1743 1469 1474 HBB1 1548 1581 1834 1745 1453 1428 ND14 1513 1514 1731 1714 1465 1425 22C3 1562 NA 1842 NA 1363 1393

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

22

Component A Component B Component C Date Assay Rep1 Rep2 Rep1 Rep2 Rep1 Rep2

101217 2PR4 1431 1411 1791 1770 1357 1364 POTP 1407 1395 1746 1820 1447 1361 NEIF 1487 1398 1803 1803 1448 1404 R4Q5 1450 1439 1767 1739 1409 1411 D5 1394 1422 1745 1773 1389 1429 ND6 1396 1393 1824 1729 1415 1417 D9 1414 1390 1752 1744 1431 1380 HBB1 1401 1382 1813 1754 1381 1336 ND14 1412 1460 1648 1671 1403 1387 22C3 1435 1410 1818 1741 1435 1372

Estimation of Copy Number Concentration It was decided that all three sets of results (homogeneity stability and certification) would be used to derive the certified values with the λ of the first five ldquoten-assayrdquo plates adjusted by a data-driven estimate of the volumetric bias This bias was incorporated into the statistical model providing better estimates of the λ means while capturing all of the variance components (homogeneity stability between-assay and between data sets) as random effects using Bayesian MCMC programmed in OpenBUGS28 Table 14 lists the estimated values and their uncertainties See Appendix C for details

Table 14 Measured Copy Number Concentrations

Genomic Copies per Nanoliter of Solution Component Mean a u(Mean) b U95(Mean) c CV95

d A 151 07 14 93 B 175 03 06 34 C 145 05 10 69

a Mean the experimentally determined most probable estimate of the true value b u(Mean) the standard uncertainty of the Mean c Mean plusmnU95(Mean) the interval which is believed with about 95 confidence to contain the true value d CV95 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the counting unit one29 2) the validity of the Poisson endpoint transformation for digital PCR endpoint assays when applied to samples providing le 05 copies per droplet 3) the determination that the copies are dispersed as dsDNA and 4) calibrated mean droplet volume measurements made at NIST during sample dilution and mastermix preparation Conversion of Copy Number to Mass Concentration As described in Reference 18 metrologically valid conversion from copies per nanoliter to nanograms per microliter requires quantitative estimates for the number of nucleotide bases per reference human genome and the average nucleotide mass Using the values estimated in that report and asserting that each target copy corresponds to one HHGE the required transformation is

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

23

[DNA] ngmicroL

= [DNA] HHGE

nLtimes

3301 ngHHGE

The combined standard uncertainty is

119906119906 [DNA] ng

microL = 1199061199062 [DNA] HHGE

nL + 0531100

times[DNA] HHGE

nL 2

Both the measured and the conversion factor uncertainties are associated with ldquolargerdquo degrees of freedom Therefor the approximate 95 confidence uncertainty is estimated

11988011988095 [DNA] ng

microL = 2 times 119906119906

[DNA] ngmicroL

Table 15 lists the transformation of the Table 14 copy number values These values agree well with the conventional 260 nm absorbance values averaging (4 plusmn 5) larger

Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution

ngmicroL Component Value a u(Measurement)b u(Conversion)c U95(Value)d CV95

e A 498 23 03 47 93 B 578 10 03 21 36 C 479 17 03 33 70

a Estimated true value b Standard uncertainty component from the ddPCR measurements c Standard uncertainty estimated from the relative uncertainty of the conversion factor d Value plusmnU95(Value) is believed with about 95 confidence to contain the true value e 100U95(Value)Value the 95 coefficient of variation (percent relative uncertainty) These results are metrologically traceable to the International System of Units through 1) the measured estimates of copies per nanoliter and 2) determinations made by us from internationally recognized literature resources of the average number of nucleotide bases per HHGE and the mean molecular weights of the bases18 Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)

While not intended to provide certified values mtDNAnDNA was measured as part of 13 experiments over 5 months Measurements were made using dilution-adjusted λ for the mtND1 mtBatz andor mtKav mitochondrial assays referenced against the dilution-adjusted λ of either the NEIF or HBB1 nDNA assays By chance both DNAs in the component C mixture have a binding site mutation that silences the mtND1 probe Figure 11 graphically summarizes the results by assay and component Table 16 lists the data Table 17 summarizes results from a random effects model that accounts for between-assay variability See Appendix C for details

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

24

Due to the limited number of mtDNA assays used in these measurements these results are metrologically traceable only to the assays used

Figure 11 mtDNAnDNA for Components A B and C

Each dot-and-bar symbol represents a mean mtDNAnDNA and its approximate 95 expanded uncertainty The dotted horizontal lines represent approximate 95 confidence intervals on the ratios

Table 16 mtDNAnDNA Ratios

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

20170711a 1 mtND1 1931 1870 2461 2401 1 mtBatz 1845 1764 2555 2608 3201 3299

20170711b 1 mtND1 1810 1906 2229 2200 1 mtBatz 1785 1837 2108 2192 3347 3287

20170712 1 mtND1 1743 1761 2127 2198 1 mtBatz 1829 1847 2455 2506 3217 3384

20170713 1 mtND1 1716 1800 2211 2199 1 mtBatz 1740 1813 2380 2249 3200 3363

20170714 1 mtND1 1923 1835 2105 1943 1 mtBatz 1935 1851 2136 2119 3294 3013

20170814 1 mtND1 1780 1776 1706 2046 2057 2024 1 mtBatz 1804 1751 1738 1998 2043 1952 2927 2776 2762 1 mtKav 1769 1786 1761 2077 2117 2041 3044 2830 2825

20170824 1 mtND1 1947 1928 1922 2 mtND1 1829 1895 1851 3 mtND1 1674 1617

20170825 1 mtBatz 1667 1813 1708 1598 1928 1916 1915 2832 2840 2788 2725 2 mtBatz 1566 1687 1663 1589 1947 1946 1961 1909 2835 2790 2736 2679 3 mtBatz 1643 1668 1604 1573 1953 1952 1941 1881 2803 2773 2757 2752

20170829 1 mtBatz 1719 1716 1670 1666 1865 1863 1828 1828 2599 2508 2551 2609 2 mtBatz 1581 1573 1601 1612 1944 1964 1947 1895 2617 2483 2626 2602 3 mtBatz 1546 1542 1538 1587 1908 1909 1940 1839 2665 2535 2570 2650

20170831 1 mtND1 1617 1599 1688 1783

ND1

ND1

ND1Ka

v

Kav

Kav

Batz

Batz

Batz

A

B

C

160

210

260

mtD

NAn

DNA

Ratio

Assay

Bind

ing s

ite m

utat

ion

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

25

mtDNA Assay

Component A Component B Component C Dataset Aliquot Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4 Rep1 Rep2 Rep3 Rep4

2 mtND1 1622 1579 1823 1631 3 mtND1 1575 1528 1920 1887

20170907 1 mtBatz 1584 1544 1890 1886 2501 2522 20170913 1 mtND1 1757 1724 1768 1990 1921 1982

1 mtBatz 1743 1671 1766 1990 1924 1958 2805 2825 2829 1 mtKav 1692 1662 1767 1903 1831 1919 2709 2738 2727

20171016 1 mtND1 1924 1944 1893 2304 2341 2332 1 mtBatz 1786 1752 1743 2048 2078 2067 2797 2642 2792 1 mtKav 1530 1542 1523 2399 2479 2493 2804 2789 2842 2 mtBatz 1580 1557 1592 2715 2606 2687 2 mtKav 2103 2100 2049 2799 2628 2559 3 mtND1 1749 1769 1777 2065 2166 2164 3 mtBatz 1965 1889 1947 2314 2280 2258 2639 2535 2657 3 mtKav 1813 1864 1864 1867 1987 1971 2953 2757 2802

Table 17 mtDNAnDNA Summary Results

mtDNAnDNA Component Value a u(Value)b U95(Value)c CV95

d A 1737 16 38 17 B 2058 22 44 21 C 2791 23 47 17

a Estimated true value b Standard uncertainty c Value plusmnU95(Value) is believed with about 95 confidence to contain the true value d 100U95(Mean)Mean the 95 coefficient of variation (percent relative uncertainty) Performance in Commercial qPCR Chemistries

Traditional qPCR assays only quantify relative to a reference standard Thus qPCR assays do not provide absolute [DNA] for the individual materials but they can provide [DNA] of one component relative to another Figure 12 summarizes the BA and CA ratios for the commercial autosomal (Human) targets from the eleven assays evaluated (see Table 8) The Y-chromosome (Male) assays cannot be summarized in this manner as the (all female) component B does not include a Y chromosome Figure 13 summaries the mass [DNA] results for the (all male) component A and (male-female mixture) component C Since this material was prepared as a nominal 13 malefemale mixture the expected malehuman (total) ratio is 14 sjc As displayed in these figures there is intrinsic variability between the SRM 2372a components and between the commercial qPCR chemistries examined This reflects the difference in the targets and amplification chemistries for each of the commercial kits Many of the commercial chemistries on the market to date contain multi-copy targets for both autosomal and Y-chromosomal quantification to increase the sensitivity of the assays

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

26

Figure 12 qPCR Autosomal Quantitation Performance Summary

Open circles denote the BA CA ratios the 17 commercial qPCR assays evaluated each labeled with the code listed in Table 8 The solid black circle represents ratios as estimated by ddPCR The red lines bound the approximate 95 confidence intervals on the ddPCR-estimated ratios

Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary

Open squares denote the ngμL values for SRM 2372a Components A (male) and C (mixture of male and female) for the seven commercial Y-chromosome qPCR assays evaluated each labeled with the code listed in Table 8 The dotted red lines represent malehuman ratios of 13 14 and 15

Appendix D provides a graphical analysis of all the data generated in the performance assessments for each of the targets amplified in the eleven commercial qPCR chemistries and the [DNA] values for all of the assays

dd

IQ_h

IQ_d

IQHY_h

IQHY_d

QFD_h

QFH_h QFHP_h

QFHP_dQFT_h

QFT_d

QPH_h

QPP_hQPP_d

P_h

PQ_h

PQ_d

06

08

10

12

14

07 09 11 13

CA

BA

IQHY_y

QFD_y

QFT_y

QPH_y

QPP_y

P_y

PQ_y

1413

15

10

12

14

16

18

30 40 50 60 70

[DNA

] C n

gμL

[DNA]A nguL

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

27

Qualification of ddPCR as Fit-For-Purpose We have documented the requirements for establishing metrological traceability of ddPCR human genomic assay results expressed as copies per nanoliter of sample in a series of three peer-reviewed papers151617 Through the experimental evidence they are based upon these papers demonstrate that ddPCR can accurately enumerate the number of independently migrating genomic copies in an aqueous extract Our most recent paper establishes the metrological traceability of these ddPCR results when transformed into the nanogram per microliter units used by the forensic and clinical communities18 However our previous studies have not addressed a fundamental requirement for correctly interpreting digital PCR measurements In broad simplification end-point assay systems count the number of independently partitioning entities regardless of the number of copies the entities contain With assays that target just one site per HHGE entities that migrate as ssDNA will yield counts that are twice the value had they migrated as dsDNA30 Should some originally dsDNA be converted over time to ssDNA ddPCR measurements of the number of independently migrating entities would increase although the number of HHGE would be unchanged Measurements of mass [DNA] provided by absorbance spectrophotometry and ddPCR reflect different properties absorbance at 260 nm is sensitive to the spatial configuration of the nucleotides as well as their number while ddPCR counts independently migrating entities The SRM 2372a materials have absorbance spectra shown in Figure 8 consistent with being predominantly dsDNA The SRM 2372 materials now have the absorbance spectra shown in Figure 14 which are consistent with them being entirely ssDNA

Figure 14 Absorbance Spectra of SRM 2372 Components A and B dsDNA is denatured to ssDNA by diluting the source material with an equal volume of 04 molL NaOH hence the NaOH-treated spectra displayed are twice the absorbance of the spectra as acquired The black vertical line marks the 260 nm values used in the conventional conversion of absorbance to mass [DNA]

00

02

04

06

08

10

12

14

16

220 260 300 340

Opt

ical

Den

sity

Wavelength nm

2372-A200720122012-NaOH121417

0

4

6

0

4

6

220 260 300 340

Wavelength nm

2372-B200720122012-NaOH121517

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

28

To ensure the validity of ddPCR-based value assignments it is thus necessary to evaluate the nature of the partitioning entities Should a solution of what is thought to be dsDNA contain an appreciable fraction of ssDNA then the ddPCR results would require adjustment31 To avoid further delay in making SRM 2372a available to the forensic community the following sections detail a series of experiments that have enabled us to conclude with high confidence that the SRM 2372a components have a dsDNA conformation and that for this material ldquoentitiesrdquo can be considered dsDNA ldquocopiesrdquo The remaining stock of the SRM 2372 components also have a dsDNA conformation however the strands in the old materials are much less robustly connected than they are in the new materials Digital Polymerase Chain Reaction Assays

Three nDNA assays were used HBB1 NEIF and ND6 The HBB1 and NEIF assays reproducibly yield very similar results for DNA extracts from many different donors The HBB1 target is at the tip of the short (p) arm of chromosome 11 The NEIF target is in the centromere region of chromosome 2 The ND6 assay was selected for use because it typically yields the lowest results of the ten nDNA assays used to certify the SRM 2372a components (see Figure 10 below) The ND6 target is about halfway down the long (q) arm of chromosome 6 Heat Treatments (Melting of the DNA into Single Strands)

In all experiments the non-heated control and the heat-treated diluted solutions were stored in the same type of PCR reaction tube The starting point for the heat treatment was to determine if there was a difference between snap-cooling on ice and slower cooling at 25 degC Two tubes of the diluted material were placed in a GeneAmp 9700 thermal cycler (Applied Biosystems) and cycled up to 98 degC and held for 15 min After 15 min one tube was removed and placed on ice while the remaining tube was cooled at 1 degCs in the thermal cycler to 25 degC then placed on ice until prepared for ddPCR Additional heat treatments included cycling samples on a GeneAmp 9700 thermal cycler

a) from 98 degC for 5 min then cooling to 25 degC for 2 min During the 25 degC hold a randomly selected tube was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 45 min at 98 degC

b) from 94 degC for 1 min then cooling to 25 degC for 1 min During the 1 min hold at 25 degC a random sample was removed and placed on ice The last tube was removed after 9 cycles of treatment resulting in a total of 9 min at 94 degC

c) from 90 degC for 1 min then cooling to 25 degC for 1 min for a total of 20 cycles Samples were pulled at 25 degC after 1 min 3 min 5 min 10 min and 20 min and placed on ice

After the above experiments heat treatments were performed using a ProFlex thermal cycler This thermal cycler has six temperature zones that can be le 5 degC apart This enables heating

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

29

tubes at the same time but at different temperatures After treatment tubes were cooled at 5 degCs to 4 degC in the thermal cycler While designed for use with PCR plates a MicroAmp 96-well tray for the VeriFlex block enabled its use with single PCR reaction tubes Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a

As a first step in evaluating whether the DNA in the SRM materials disperse as dsDNA or ssDNA a direct comparison of the five materials was made using the HBB1 and NEIF assays Table 18 reports summary statistics for the ddPCR measurements reported as λ copiesdroplet and the estimated mass [DNA] values from the ddPCR results18 from the conventional dsDNA relationship7 and from the conventional ssDNA relationship after denaturing with 04 molL NaOH89 While the mass [DNA] values differ somewhat between the three methods there is no evidence for the SRM materials segregating exclusively or even mostly as ssDNA The extent of the hypochromic interactions among the nucleotides does not accurately reflect the extent to which the two strands of dsDNA have separated

Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a

λ copies per droplet λ copies per droplet ngmicroL Sample Assay Nrep

a Mean CVb ncmbc Mean CVb ddPCRd dsDNAe ssDNAf

2372-A HBB1 4 03645 19 8 03644 20 566g 524 57 2372-A NEIF 4 03644 23 2372-B HBB1 4 03513 23 8 03487 35 580h 536 61 2372-B NEIF 4 03461 47 2372a-A HBB1 4 02603 27 8 02649 30 474 465 454 2372a-A NEIF 4 02694 24 2372a-B HBB1 4 02908 34 8 02891 37 517 556 534 2372a-B NEIF 4 02873 45 2372a-C HBB1 4 02757 28 8 02711 28 485 502 474 2372a-C NEIF 4 02665 16

a Number of technical replicates for the given assay b CV = 100(standard deviation)mean c Combined number of technical replicates ignoring potential differences between the assays d Estimated from the ddPCR λ measurements listed in this Table18 e Estimated from the conventional relationship between the mass [DNA] of dsDNA and absorbance at

260 nm7 using absorbance spectra acquired shortly after solutions were produced f Estimated from the relationship between the mass [DNA] of ssDNA and absorbance at 260 nm after 11

dilution with 04 molL NaOH89 g Values adjusted to measured (132 plusmn19) volume loss of archived units due to evaporation h Values adjusted to measured (71 plusmn25) volume loss of archived due to evaporation

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

30

Quantitative Evaluation of ssDNA Content

Absorbance spectrophotometry is routinely used to quantitatively assess DNA tertiary structure3233 On storage ldquoat physiological pH the intrastrand repulsion between the negatively charged phosphate groups forces the double helix into more rigid rodlike conformationrdquo34 suggesting a mechanism for the observed slow changes in the SRM 2372 materials As demonstrated in Table 18 absorbance at 260 nm does not necessarily correlate with ddPCR behavior Yet ldquofurthermore the repulsion between phosphate groups on opposite strands tends to separate the complementary strandsrdquo34 Therefore a method that is independent of hypochromism is needed to adequately evaluate the proportion (if any) of ldquotruerdquo (completely disjoint) ssDNA in the SRM materials Circular dichroism spectroscopy is sensitive to the conformational structures of DNA in solution35 and different DNA configurations migrate differentially in agarose gel electrophoresis7 Both techniques have been used to qualitatively discriminate ssDNA from dsDNA in ldquopurerdquo single-sequence materials but neither technique appears to be able to quantify small proportions of ssDNA in a complex multi-chromosome genomic material Wilson and Ellison have proposed a real-time chamber digital PCR (cdPCR) technique that can estimate excess proportions of ssDNA copies from the crossing-threshold distribution31 However their method is insufficiently sensitive for low ssDNA proportions It is well known that heating DNA beyond the melting temperature separates mammalian dsDNA into ssDNA36 albeit we have shown using cdPCR that boiling dsDNA for some minutes also renders some target sites non-amplifiable15 In the following section we document a ddPCR-based approach for estimating the proportion if any of ssDNA in a sample that involves less destructive heating While the method is not yet optimized the results from these preliminary studies demonstrate that the SRM 2372 and 2372a components disperse into the ddPCR droplets as dsDNA ddPCR Measurement of ssDNA Proportion Given two otherwise identical aliquots of a given DNA material one subjected to some treatment and the other untreated the ratio between the average number of copies per droplet of treated material and that of its untreated sibling λtλu quantitatively estimates the change in the number of independently dispersing entities Should the treatment reduce the number of accessible amplifiable targets (AAT) λtλu will be less than 1 Reduction in AAT may result from damage to the target itself andor conformational changes to the fragment containing the target that render the target inaccessible or non-amplifiable16 Assuming no denaturation of dsDNA to ssDNA either because of an ineffective treatment or the material having started as ldquopurerdquo ssDNA if the treatment does not itself damage targets then λtλu will equal 1 Should treatment convert at least some dsDNA to ssDNA without damaging targets the number of AAT will increase and λtλu will be greater than 1 Assuming complete conversion of ldquopurerdquo dsDNA to ldquopurerdquo ssDNA without other damage then λtλu will equal 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

31

Values greater than 1 but less than 2 indicate some conversion of dsDNA to ssDNA but also some combination of (a) incomplete denaturation of dsDNA to ssDNA (b) partial renaturation of ssDNA to dsDNA after treatment but before ddPCR analysis (c) damage to targets during the process or ndash of greatest interest ndash (d) the original sample material was an impure mixture of dsDNA and ssDNA The following discussions graphically summarize sets of λtλu followed by tabular listings of the relevant experimental conditions ddPCR measurements and the numerical λtλu values The information in these tables includes

Plate Index Experiment identifier indicates date performed in YYYYMMDD format Material Sample material SRM 2372-A 2372-B 2372a-A 2372a-B or 2372a-C Assay ddPCR assay HBB1 NEIF or ND6 degC Denaturation temperature in degC Min Length of time the treated material was held at denaturation temperature in

minutes Cool Method used to cool treated material from denaturation temperature Snap

Slow R1 or R5 where ldquoSnaprdquo indicates cooling on ice ldquoSlowrdquo cooling at room temperature to 25 degC ldquoR1rdquo cooling in a thermal cycler to 25 degC at 1 degCs and ldquoR5rdquo cooling in a thermal cycler to 4 degC at 5 degCs

F Dilution factor expressed as (volume Material) per (final volume of diluted solution) 12 14 or 18

n Number of technical replicates Mean Mean of technical replicates CV Coefficient of Variation expressed as a percentage 100 Standard Deviation

Mean

λtλu the copies per droplet result of the treated solution λt divided by the copies per droplet result of its untreated sibling λu

u(λtλu) Standard uncertainty of λtλu 119906119906(120582120582t 120582120582ufrasl ) = 120582120582t 120582120582ufrasl CV(120582120582t)100

2

+ CV(120582120582u)100

2

Proof of Principle Figure 15 summarizes two replicate sets of measurements performed to assess whether the SRM 2372 and 2372a materials respond similarly whether the denatured materials renature if allowed to cool slowly and whether results are (reasonably) repeatable Table 19 details the experimental conditions The λtλu values for the four materials are similar all much greater than 1 but not approaching 2 Most of the ratios are on or close to the ldquoslow cooling is the same as snap coolingrdquo line suggesting that renaturation is not a significant concern While the replicate results for the SRM 2372-A and 2372a-A materials do not overlap as closely as those for SRM 2372-B and 2372a-B similar treatment conditions produce similar results

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

32

Figure 15 Comparison of Snap-Cooling to Slow-Cooling

Symbols mark the means of replicate measurements made using the same dilutions on successive days with bars representing plusmn one standard combined uncertainty Results are labeled by material ldquoArdquo and ldquoBrdquo for SRM 2372 components and ldquoaArdquo ldquoaBrdquo and ldquoaCrdquo for SRM 2372a components The diagonal black line represents equality between the snap- and slow-cooling treatments

Table 19 Comparison of Snap-Cooling to Slow-Cooling

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171218 2372-A HBB1 14 4 03830 29

dd20171218 2372-A HBB1 98 15 Slow 14 2 05644 18 1474 0051 dd20171218 2372-A HBB1 98 15 Snap 14 4 05852 15 1528 0051

dd20171218 2372a-A HBB1 14 4 03029 18

dd20171218 2372a-A HBB1 98 15 Slow 14 4 04430 21 1463 0040 dd20171218 2372a-A HBB1 98 15 Snap 14 4 04454 33 1471 0055

dd20171218 2372-B HBB1 14 4 03594 31

dd20171218 2372-B HBB1 98 15 Slow 14 4 05806 18 1615 0058 dd20171218 2372-B HBB1 98 15 Snap 14 4 05825 13 1620 0055

dd20171218 2372a-B HBB1 14 4 03383 15

dd20171218 2372a-B HBB1 98 15 Slow 14 4 05043 38 1491 0061 dd20171218 2372a-B HBB1 98 15 Snap 14 4 05254 20 1553 0039

dd20171219 2372-A HBB1 14 4 03357 23

dd20171219 2372-A HBB1 98 15 Slow 14 4 05340 16 1591 0046 dd20171219 2372-A HBB1 98 15 Snap 14 4 05340 12 1591 0042

A

aA

B

aB

A

aA

B

aB

14

15

16

17

14 15 16 17

λ tλ

u Sn

ap C

oole

d

λtλu Slow Cooled

Rep 1Rep 2Snap = Slow

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

33

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171219 2372a-A HBB1 14 4 02735 24

dd20171219 2372a-A HBB1 98 15 Slow 14 4 04151 03 1518 0037 dd20171219 2372a-A HBB1 98 15 Snap 14 4 03918 21 1433 0045 dd20171219 2372-B HBB1 14 4 03346 22

dd20171219 2372-B HBB1 98 15 Slow 14 4 05476 17 1637 0045 dd20171219 2372-B HBB1 98 15 Snap 14 4 05480 29 1638 0059 dd20171219 2372a-B HBB1 14 4 03158 19

dd20171219 2372a-B HBB1 98 15 Slow 14 4 04798 16 1519 0038 dd20171219 2372a-B HBB1 98 15 Snap 14 4 04856 17 1538 0039

Time at Denaturation Temperature Figure 16 summarizes λtλu for the SRM 2372-B material as a function of denaturation temperature and the cumulative time the solution is held at that temperature Table 20 details the experimental conditions

Figure 16 Change in λtλu with Increasing Time at Several Denaturation Temperatures

The symbols represent the ratio of the mean result of three technical replicates of the untreated and each of the treated solutions the bars represent combined standard uncertainties The lines denote regression fits to the function λtλu = atimese-bt where t is the cumulative time Values and the standard uncertainties of the coefficients are listed in the legend

12

14

16

18

20

0 5 10 15 20

λ t λ

u

Cummulative Treatment Time min

degC a u (a ) b u (b )90 1804 0039 00198 0000494 1788 0029 00099 0002998 1855 0019 00067 00019

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

34

For all three of the denaturation temperatures studied the relationship between λtλu and cumulative time t can be modeled as

λtλu = ae-bt

where a is the extrapolated value of λtλu at t = 0 and b is the rate of change in AAT per minute As expected the rate of loss due to the treatment increases with increasing treatment temperature and cumulative time at that temperature37 Given the much longer treatment time for the 98 degC series the agreement among the three t = 0 parameter (a) values is remarkable This suggests that the first 1 min of treatment efficiently separates dsDNA into ssDNA The pooled relative standard uncertainty of these regression parameter is 17 somewhat smaller than the 25 pooled relative standard uncertainties of the individual λtλu However the plusmnstandard uncertainty interval of the 1-min λtλu results for the 94 degC and 90 degC both overlap the extrapolated t = 0 value suggesting that reliable results can be obtained using a single treatment of short duration

Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20171228 2372-B HBB1 14 3 03650 17 dd20171228 2372-B HBB1 98 5 R1 14 3 06157 12 1687 0035 dd20171228 2372-B HBB1 98 10 R1 14 3 05493 08 1505 0028 dd20171228 2372-B HBB1 98 15 R1 14 3 05019 13 1375 0029 dd20171228 2372-B HBB1 98 20 R1 14 3 04583 21 1256 0034 dd20171228 2372-B HBB1 98 25 R1 14 3 04154 32 1138 0041 dd20171228 2372-B HBB1 98 30 R1 14 3 03784 14 1037 0023 dd20171228 2372-B HBB1 98 35 R1 14 3 03408 10 0934 0019 dd20171228 2372-B HBB1 98 40 R1 14 3 02980 01 0816 0014 dd20171228 2372-B HBB1 98 40 R1 14 3 02815 08 0771 0015

dd20171229 2372-B HBB1 14 3 03708 27

dd20171229 2372-B HBB1 94 1 R1 14 3 06598 26 1779 0056 dd20171229 2372-B HBB1 94 2 R1 14 3 06553 32 1767 0063 dd20171229 2372-B HBB1 94 3 R1 14 3 06301 20 1699 0045 dd20171229 2372-B HBB1 94 4 R1 14 3 06252 14 1686 0037 dd20171229 2372-B HBB1 94 5 R1 14 3 06577 14 1773 0039 dd20171229 2372-B HBB1 94 6 R1 14 3 06156 07 1660 0030 dd20171229 2372-B HBB1 94 7 R1 14 3 06118 10 1650 0033 dd20171229 2372-B HBB1 94 8 R1 14 3 06256 12 1687 0035 dd20171229 2372-B HBB1 94 9 R1 14 3 05999 10 1618 0032

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

35

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180104 2372-B HBB1 14 4 03713 25 dd20180104 2372-B HBB1 90 1 R1 14 3 06916 22 1863 0052 dd20180104 2372-B HBB1 90 3 R1 14 3 06465 06 1741 0032 dd20180104 2372-B HBB1 90 5 R1 14 3 06431 20 1732 0045 dd20180104 2372-B HBB1 90 10 R1 14 3 06014 25 1620 0049 dd20180104 2372-B HBB1 90 15 R1 14 3 06024 10 1623 0032 dd20180104 2372-B HBB1 90 20 R1 14 3 06003 14 1617 0035

Temperature Figure 17 summarizes the effect of denaturation temperature using a single treatment of 05 min on the 2372-B and 2372a-B materials Table 21 details the experimental conditions

Figure 17 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

The symbols represent the ratio of the mean result of four technical replicates of the untreated and three replicates of the treated solutions the bars represent combined standard uncertainties In all cases the duration of the treatment was 05 min The lines denote regression fits to the linear function λtλu = a +bT where Tis the denaturation temperature

The SRM 2372 and 2372a materials respond very differently to changes in denaturation temperature The older spectrophotometrically ssDNA material suffers somewhat greater loss in λtλu as temperature increases The new spectrophotometrically dsDNA material suffers much greater loss in λtλu as temperature decreases

14

15

16

17

18

19

20

82 86 90 94 98

λ tλ

u

Temperature degC

2372-B

2372a-B

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

36

Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180109 2372-B HBB1 14 4 03577 25 dd20180109 2372-B HBB1 100 05 R5 14 3 06314 16 1765 0051

dd20180104a 2372-B HBB1 14 4 03577 05

dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032 dd20180104a 2372-B HBB1 94 05 R5 14 3 06459 25 1806 0046 dd20180104a 2372-B HBB1 90 05 R5 14 2 06591 08 1842 0017 dd20180104a 2372-B HBB1 88 05 R5 14 3 06589 05 1842 0013 dd20180104a 2372-B HBB1 84 05 R5 14 3 06729 10 1881 0022 dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

We hypothesize that in the older SRM 2372 material the between-strand bond is relatively fragile and can be completely severed at relatively low temperature while avoiding much temperature-dependent AAT loss In contrast the between-strand bond in the SRM 2372a material requires higher temperature to completely disassociate the strands This suggests that there is not a sample-independent ldquooptimumrdquo combination of denaturation temperature and treatment duration However 98 degC for 05 min appears to provide comparable results for these two materials DNA Concentration in the ddPCR Solution Figure 18 summarizes the effect of [DNA] in the ddPCR solution with SRM 2372-B at a denaturation temperature of 98 degC and treatment duration of 05 min Table 22 details the experimental conditions

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

37

Figure 18 Change in λtλu with DNA Concentration in ddPCR Solution

The solid black symbols represent the ratio of the mean result of four technical replicates of the untreated and treated solutions with the λtλu values shown on the left-hand vertical axis the bars represent combined standard uncertainties The solid blue circles represent λu and the open red circles λt with values shown on the right-hand vertical axis at this graphical scale the standard deviations of the technical replicates are covered by the symbol The diagonal blue and red lines represent the log-linear relationships between just the 18 and 14 λ values

The primary effect of [DNA] appears to be the loss of linearity in λt with copiesdroplet values greater than about 1 copiesdroplet There may be some advantage to use somewhat greater dilutions than the routine 14 however the small increase in λtλu may be offset by the increased variability with small λu This suggests the use of dilutions that give λu values of (01 to 04) copiesdroplet

Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180117 2372-B HBB1 12 4 07127 11 dd20180117 2372-B HBB1 98 05 R5 12 4 12331 14 1730 0031

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 3 06151 15 1843 0040

dd20180117 2372-B HBB1 18 4 01549 18 dd20180117 2372-B HBB1 98 05 R5 18 4 02897 20 1870 0051

12 14 18

170

175

180

185

190

02

04

06

08

101214

λ T

arge

t D

ropl

et

λ tλ

u

Dilution Factor

λ₊λᵤλᵤλ₊

λtλu

λu

λt

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

38

Equivalence of Assays Figure 19 summarizes the agreement between the HBB1 NEIF and ND6 assays at a denaturation temperature of 98 degC and treatment duration of 05 min Table 23 details the experimental conditions

Figure 19 Comparison of Assays

The symbols compared the λtλu results from the HBB1 and either the NEIF or ND6 assays Each λtλu value is the ratio of the mean result of four technical replicates of an untreated and three or four replicates of the treated solutions The bars represent combined standard uncertainties The diagonal line represents equality between the assays

There is little or no difference between the λtλu results from the HBB1 NEIF and ND6 assays suggesting that the chromosomal location of the assay target has little influence Since the NEIFHBB1 comparison includes results for both the SRM 2372 and 2372a materials this equivalence is not material-specific

14

16

18

20

14 16 18 20

λ tλ

u NE

IF o

r ND6

λtλu HBB1

NEIFHBB1

ND6HBB1

2nd Assay = HBB1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

39

Table 23 Comparison of Assays

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu)

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034 dd20180105 2372a-B HBB1 94 05 R5 14 3 05510 22 1741 0047 dd20180105 2372a-B HBB1 90 05 R5 14 3 05270 23 1665 0047 dd20180105 2372a-B HBB1 88 05 R5 14 3 05032 09 1590 0029 dd20180105 2372a-B HBB1 86 05 R5 14 3 04880 18 1542 0036 dd20180105 2372a-B HBB1 84 05 R5 14 3 04760 04 1504 0025

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045 dd20180105 2372a-B ND6 94 05 R5 14 3 05526 36 1711 0074 dd20180105 2372a-B ND6 90 05 R5 14 3 05326 18 1649 0051 dd20180105 2372a-B ND6 88 05 R5 14 3 05077 22 1572 0052 dd20180105 2372a-B ND6 86 05 R5 14 3 05038 23 1560 0053 dd20180105 2372a-B ND6 84 05 R5 14 3 04906 10 1519 0041

dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

40

ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials

Figure 20 summarizes the currently available λtλu results obtained by denaturing at 98 degC for 05 min rapidly cooling in the thermal cycler with dilutions that provide asymp 03 copiesdroplet The figure also displays the consensus result and its 95 confidence region estimated using the Hierarchical Bayes method provided in the NIST Consensus Builder (NICOB) system3839 Table 24 details the experimental conditions

Figure 20 Evaluation of Limiting Ratio

The symbols represent the ratio of the mean result of three or four technical replicates of the untreated and treated solutions the bars represent combined standard uncertainties The thick horizontal line represents a consensus value covering all displayed results the dotted horizontal lines bracket the 95 confidence interval about the consensus value

The Hierarchical Bayes consensus result for these values is 183 with a 95 confidence interval of plusmn003 All but three of the 16 λtλu values are within one standard uncertainty of this consensus value with most of the λtλu means lying within the 95 confidence interval This suggests that there are no statistically significant differences between the SRM 2372 and 2372a materials with regard to ssDNA proportion

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

17

18

19

20

λ tλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

41

Table 24 Evaluation of Limiting Ratio

Treatment λ Ratio Plate Index Material Assay degC Min Cool F n Mean CV λtλu u(λtλu) dd20180110 2372-A HBB1 14 3 03949 33 dd20180110 2372-A HBB1 98 05 R5 14 4 06989 16 1770 0065

dd20180111 2372-A NEIF 14 4 03971 11 dd20180111 2372-A NEIF 98 05 R5 14 4 07080 37 1783 0069

dd20180104a 2372-B HBB1 14 4 03577 05 dd20180104a 2372-B HBB1 98 05 R5 14 3 06592 17 1843 0032

dd20180108 2372-B HBB1 14 3 03463 09 dd20180108 2372-B HBB1 98 05 R5 14 3 06562 14 1895 0031

dd20180109 2372-B HBB1 14 3 03577 25 dd20180109 2372-B HBB1 98 05 R5 14 3 06495 21 1816 0059

dd20180110 2372-B HBB1 14 4 02269 44 dd20180110 2372-B HBB1 98 05 R5 14 4 04156 12 1832 0084

dd20180111 2372-B NEIF 14 4 02324 11 dd20180111 2372-B NEIF 98 05 R5 14 4 04249 46 1828 0087

dd20180117 2372-B HBB1 14 4 03338 15 dd20180117 2372-B HBB1 98 05 R5 14 4 06151 15 1843 0040

dd20180110 2372a-A HBB1 14 4 02596 13 dd20180110 2372a-A HBB1 98 05 R5 14 4 04721 15 1819 0036

dd20180111 2372a-A NEIF 14 4 02698 14 dd20180111 2372a-A NEIF 98 05 R5 14 4 04755 15 1762 0035

dd20180105 2372a-B HBB1 14 4 03165 16 dd20180105 2372a-B HBB1 98 05 R5 14 3 05901 09 1864 0034

dd20180110 2372a-B HBB1 14 4 02845 27 dd20180110 2372a-B HBB1 98 05 R5 14 3 05283 29 1857 0074

dd20180111 2372a-B NEIF 14 4 02943 25 dd20180111 2372a-B NEIF 98 05 R5 14 3 05448 19 1851 0058

dd20180105 2372a-B ND6 14 4 03229 25 dd20180105 2372a-B ND6 98 05 R5 14 3 05790 01 1793 0045

dd20180110 2372a-C HBB1 14 4 02443 27 dd20180110 2372a-C HBB1 98 05 R5 14 4 04614 27 1889 0072

dd20180111 2372a-C NEIF 14 4 02568 23 dd20180111 2372a-C NEIF 98 05 R5 14 4 04747 24 1849 0062

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

42

The agreement among the measured λtλu is more directly visualized using estimated unilateral ldquoDegrees of Equivalencerdquo d Each d is the absolute difference between the measured and consensus values

119889119889 = |120582120582119905119905 120582120582119906119906frasl minus consensus|

and is associated with a 95 confidence interval U95(d) that combines the uncertainty estimates of the measured and consensus values The U95(d) calculation details are specific to the consensus estimation method39 Values of d plusmn U95(d) that include 0 indicate that the λtλu result does not differ from the majority at about a 95 level of confidence Figure 21 displays the d for the λtλu results for the Hierarchical Bayes consensus value39 By this figure of merit all the λtλu can be considered as members of a single population

Figure 21 Degrees of Equivalence

The symbols represent the Degree of Equivalence d for the λtλu results relative to the Hierarchical Bayes consensus value The bars represent 95 level of confidence uncertainty intervals d plusmn U95(d) The horizontal line marks the d = 0 ldquono differencerdquo value Results with d plusmn U95(d) that include this line are not significantly different from the majority

2372

-A

2372

-B

2372

a-A

2372

a-B

2372

a-C

-024

-018

-012

-006

000

006

012

018

024

Degr

ees o

f Equ

ival

ence

λtλ

u

HBB1NEIFND6

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

43

Conclusions and Recommendations With high confidence all tubes of the SRM 2372a solution have the same nDNA copy number content within measurement repeatability Additionally within measurement repeatability all tubes of the SRM 2372a have the same mtDNAnDNA With high confidence the nDNA copy number content of the SRM 2372a solution is thermally stable from 4 degC to 37 degC over an extended period The solution should not be shipped or stored below 4 degC With high confidence the ddPCR measurements used to establish the entity and mass concentrations and the mtDNAnDNA ratios are adequately precise and unbiased With high confidence the DNA in the SRM 2372a solutions is now dsDNA and can be expected to remain dsDNA for at least ten years when stored at 4 degC Recommended Certification Values

Since the component materials are believed to be homogeneous are of very similar composition and were prepared and processed similarly the observed differences in the measured entity concentration uncertainties are likely artifacts of the measurement process To limit over-interpretation of these differences we recommend that the copy per nanoliter values of the three components be assigned same 95 relative uncertainty interval (CV95) based on the largest of the measured CV95 values (93 ) rounded up to 10 Since the ngμL transformation does not appreciably inflate this relative uncertainty we recommend that the mass concentration values also be assigned CV95 of 10 Similarly we recommend that the mtDNAnDNA ratios be rounded to integer values all three components be assigned the same CV95 and that this uncertainty be based on the largest (21 ) of the measured CV95 rounded up to 25 Table 25 lists the values that we recommend be used in the Certificate of Analysis

Table 25 Recommended Values and 95 Uncertainties for Certification

Component Units Value U95(Value) a A Copies per nanoliter 151 15 B Copies per nanoliter 175 18 C Copies per nanoliter 145 15 A ngμL 498 50

B ngμL 578 58 C ngμL 479 48 A mtDNAnDNA 174 4

B mtDNAnDNA 206 5 C mtDNAnDNA 279 7

a The interval Value plusmn U95(Value) is believed with about 95 confidence to contain the true value of the measurand

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

44

Use of SRM 2372a for Value-Assigning In-House Reference Materials

The within-component and between-assay variability observed for the commercial qPCR assays is not unexpected since different qPCR methods exploit qualitatively different molecular targets Given the intrinsic variability of the human genome the prospect of developing a human DNA source material that responds equally to all current and future qPCR assays is improbable All three components should be used to elucidate potential bias when using these materials to value assign in-house DNA reference and control materials for qPCR assays

Certificate of Analysis In accordance with ISO Guide 31 2000 a NIST SRM certificate is a document containing the name description and intended purpose of the material the logo of the US Department of Commerce the name of NIST as a certifying body instructions for proper use and storage of the material certified property value(s) with associated uncertainty(ies) method(s) used to obtain property values the period of validity if appropriate and any other technical information deemed necessary for its proper use A Certificate is issued for an SRM certified for one or more specific physical or engineering performance properties and may contain NIST reference information or both values in addition to certified values A Certificate of Analysis is issued for an SRM certified for one or more specific chemical properties Note ISO Guide 31 is updated periodically check with ISO for the latest version [httpswwwnistgovsrmsrm-definitions] For the most current version of the Certificate of Analysis for NIST SRM 2372a Human DNA Quantification Standard please visit httpswww-snistgovsrmorsview_detailcfmsrm=2372a

References

1 National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research Department of Health Education and Welfare (1978) Belmont Report Ethical Principles and Guidelines for the Protection of Human Subjects of Research DHEW Publication No (OS) 78-0012 httpvideocastnihgovpdfohrp_belmont_reportpdf

2 CDCNIH Biosafety in Microbiological and Biomedical Laboratories 5th ed HHS publication No (CDC) 21-1112 Chosewood LC Wilson DE Eds US Government Printing Office Washington DC (2009) httpwwwcdcgovbiosafetypublicationsbmbl5indexhtm

3 Duewer DL Kline MC Redman JW Newall PJ Reeder DJ NIST Mixed Stain Studies 1 and 2 Interlaboratory Comparison of DNA Quantification Practice and Short Tandem Repeat Multiplex Performance with Multiple-Source Samples J Forensic Sci 200146(5)1199-1210 PMID 11569565

4 Kline MC Duewer DL Redman JW Butler JM NIST Mixed Stain Study 3 DNA quantitation accuracy and its influence on short tandem repeat multiplex signal intensity Anal Chem 200375(10)2463-2469 httpsdoiorg101021ac026410i

5 Kline MC Duewer DL Redman JW Butler JM Results from the NIST 2004 DNA Quantitation Study J Forensic Sci 200550(3)571-578 PMID 15932088

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

45

6 Kline MC Duewer DL Travis JC Smith MV Redman JW Vallone PM Decker AE Butler JM Production and Certification of Standard Reference Material 2372 Human DNA Quantitation Standard Anal Bioanal Chem 2009394(4)1183-1192 httpsdoiorg101007s00216-009-2782-0

7 Sambrook J and Russell DW (2001) Molecular Cloning a Laboratory Manual Cold Spring Harbor Laboratory Press Cold Spring Harbor New York

8 EU-JRC Protocol NK603 ndash Community Reference Laboratory Event-specific method for the quantification of maize line NK603 using real-time PCR Section 33 Spectrophotometric Measurement of DNA Concentration httpgmo-crljrceceuropaeusummariesNK603-WEB-ProtocolValidationpdf

9 ISO 215712005(E) Foodstuffs ndash Methods of analysis for the detection of genetically modified organisms and derived products ndash Nucleic acid extraction Annex B Methods for quantitation of extracted DNA pp 34-36 International Organization for Standardization Geneva (2005)

10 Certificate of Analysis Standard Reference Material 2372 Human DNA Quantitation Standard (2013) Gaithersburg MD USA Standard Reference Materials Program National Institute of Standards and Technology httpswww-snistgovsrmorsview_certcfmsrm=2372

11 Vogelstein B Kinzler KW Digital PCR Proc Natl Acad Sci USA 1999969236-9241 12 Hindson BJ Ness KD Masquelier DA Belgrader P Heredia NJ Makarewicz AJ Bright IJ Lucero

MY Hiddessen AL Legler TC Kitano TK Hodel MR Petersen JF Wyatt PW Steenblock ER Shah PH Bousse LJ Troup CB Mellen JC Wittmann DK Erndt NG Cauley TH Koehler RT So AP Dube S Rose KA Montesclaros L Wang S Stumbo DP Hodges SP Romine S Milanovich FP White HE Regan JF Karlin-Neumann GA Hindson CM Saxonov S Colston BW High-Throughput Droplet Digital PCR System for Absolute Quantitation of DNA Copy Number Anal Chem 2011838604-8610 httpsdoiorg101021ac202028g

13 Pinheiro LB Coleman VA Hindson CM Herrmann J Hindson BJ Bhat S Emslie KR Evaluation of a Droplet Digital Polymerase Chain Reaction Format for DNA Copy Number Quantification AnalChem 2012841003-1011 httpsdoiorg101021ac202578x

14 Corbisier P Pinheiro L Mazoua S Kortekaas AM Chung PY Gerganova T Roebben G Emons H Emslie K DNA copy number concentration measured by digital and droplet digital quantitative PCR using certified reference materials Anal Bioanal Chem 2015407(7)1831-1840 httpsdoiorg101007s00216-015-8458-z

15 Duewer DL Kline MC Romsos EL Real-time cdPCR opens a window into events occurring in the first few PCR amplification cycles Anal Bioanal Chem 2015407(30)9061-9069 httpsdoiorg101007s00216-015-9073-8

16 Kline MC Romsos EL Duewer DL Evaluating Digital PCR for the Quantification of Human Genomic DNA Accessible Amplifiable Targets Anal Chem 201688(4)2132-2139 httpsdoiorg101021acsanalchem5b03692

17 Kline MC Duewer DL Evaluating Droplet Digital Polymerase Chain Reaction for the Quantification of Human Genomic DNA Lifting the Traceability Fog Anal Chem 201789(8)4648-4654 httpsdoiorg101021acsanalchem7b00240

18 Duewer DL Kline MC Romsos EL Toman B Evaluating Droplet Digital PCR for the quantification of human genomic DNA Conversion from Copies per Nanoliter to Nanogram Genomic DNA per Microliter Anal Bioanal Chem 2018in press httpsdoiorg101007s00216-018-0982-1

19 Dagata JA Farkas N Kramer JA Method for measuring the volume of nominally 100 μm diameter spherical water in oil emulsion droplets NIST Special Publication 260-184 2016 httpsdoiorg106028NISTSP260-184

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

20 ISO 170342016 General requirements for the competence of reference material producers International Organization for Standardization Geneva (2016)

21 NCBI Standard Nucleotide BLAST httpsblastncbinlmnihgovBlastcgiPAGE_TYPE=BlastSearch

22 Timken MD Swango KL Orrego C Buoncristiani MR A Duplex Real-Time qPCR Assay for the Quantification of Human Nuclear and Mitochondrial DNA in Forensic Samples Implications for Quantifying DNA in Degraded Samples J Forensic Sci 200550(5)1044-60 PMID 16225209

23 Walker JA Hedges DJ Perodeau BP Landry KE Stoilova N Laborde ME Shewale J Sinha SK Batzer MA Multiplex polymerase chain reaction for simultaneous quantitation of human nuclear mitochondrial and male Y-chromosome DNA application in human identification Anal Biochem 200533789ndash97 httpsdoiorg101016jab200409036

24 Kavlick MF Lawrence HS Merritt RT Fisher C Isenberg A Robertson JM Budowle B Quantification of Human Mitochondrial DNA Using Synthesized DNA Standards J Forensic Sci 201156(6) 1457ndash1463 httpsdoiorg101111j1556-4029201101871x

25 Stulnig TM Amberger A Exposing contaminating phenol in nucleic acid preparations BioTechniques 199416(3) 403-404

26 Majumdar N Wessel T Marks J Digital PCR Modeling for Maximal Sensitivity Dynamic Range and Measurement Precision PLoS ONE 201510(3)e0118833 httpsdoi101371journalpone0118833

27 Tellinghuisen J Partition Volume Variability in Digital Polymerase Chain Reaction Methods Polydispersity Causes Bias but Can Improve Precision Anal Chem 2016881283-12187 httpsdoiorg101021acsanalchem6b03139

28 Lunn DJ Spiegelhalter D Thomas A Best N (2009) Statistics in Medicine 283049ndash3082 29 De Biegravevre P Dybkaer R Fajgelj A Hibbert DB Metrological traceability of measurement results in

chemistry Concepts and implementation Pure Appl Chem 201183(10)1873-1935 httpsdoiorg101007s00769-011-0805-y

30 Huggett JF Foy CA Benes V Emslie K Garson JA Haynes R Hellemans J Kubista M Mueller RD Nolan T Pfaffl MW Shipley GL Vandesompele J Wittwer CT Bustin SA The Digital MIQE Guidelines Minimum Information for Publication of Quantitative Digital PCR Experiments Clin Chem 201359(6)892-902 httpsdoiorg101373clinchem2013206375

31 Wilson PJ Ellison SLR Extending digital PCR analysis by modelling quantification cycle data BMC Bioinfo 201617421 httpsdoiorg101186s12859-016-1275-3

32 Wang X Son A Effects of pretreatment on the denaturation and fragmentation of genomic DNA for DNA hybridization Environ Sci Processes Impacts 2013152204 httpsdoiorg101039c3em00457k

33 Nwokeoji AO Kilby PM Portwood DE Dickman MJ Accurate Quantification of Nucleic Acids Using Hypochromicity Measurements in Conjunction with UV Spectrophotometry Anal Chem 201789(24)13567-13574 https101021acsanalchem7b04000

34 Farkas DH Kaul KL Wiedbrauk DL Kiechle FL Specimen collection and storage for diagnostic molecular pathology investigation Arch Pathol Lab Med 1996120(6)591-6

35 Kypr J Kejnovskaacute I Renčiuk D Vorličkovaacute M Circular dichroism and conformational polymorphism of DNA Nucleic Acids Res 200937(6)1713-1725 httpsdoiorg101093nargkp026

36 Strachan T Read AP (1999) Human Molecular Genetics 2nd Ed John Wiley amp Sons NY

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

37 Bhat S McLaughlin JL Emslie KR Effect of sustained elevated temperature prior to amplification on template copy number estimation using digital polymerase chain reaction Analyst 2011136724ndash32 httpsdoiorg101039c0an00484g

38 NIST Consensus Builder (NICOB) httpsconsensusnistgov 39 Koepke A Lafarge T Toman B Possolo A NIST Consensus Builder Userrsquos Manual

httpsconsensusnistgovNISTConsensusBuilder-UserManualpdf

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

46

Appendix A DNA Extraction Method The following method has been adapted from Miller et al (1988)A1 Procedure

1 Transfer approximately 5 mL of buffy coat into a 50 mL sterile polypropylene centrifuge tube and add 30 mL Red Blood Cell Lysis Buffer

2 Mix by gentle inversion and let stand at room temperature for 15 min

3 Centrifuge at approximately 209 rads (2000 rpm) for 20 min

4 Discard supernatant

5 Add 9 mL of Nucleic Lysis Buffer and re-suspend the pelleted cells by agitating with a plastic transfer pipette

6 Add 300 microL of 20 sodium dodecyl sulfate (SDS) and 18 mL of Protease K Solution (freshly prepared 2 mg to 5 mg Protease K per 1 mL Protease K Buffer)

7 Mix by inversion followed by incubation overnight or longer at 37 ordmC to 40 ordmC in a shaking water bath The length of incubation is determined by the time required to solubilize the cells present in the tube

8 Add 4 mL saturated ammonium acetate solution (in deionized water stored at 2 ordmC to 8 ordmC) Shake vigorously for 15 s

9 Centrifuge at approximately 471 rads (4500 rpm) for 10 min

10 Pipette supernatant into a 50 mL sterile polypropylene centrifuge tube and add two volumes of room temperature absolute ethanol Mix by gentle inversion until the DNA aggregates and floats to the top of the tube

11 Spool the aggregated DNA from the current tube into a fresh tube of 80 ethanol using a sterile inoculation loop At this step multiple tubes of precipitated DNA may be pooled to make a large lot of the purified material Mix by gentle inversion

12 Transfer the aggregated DNA to a fresh dry 50 mL polypropylene centrifuge tube Remove as much ethanol as possible by squeezing the DNA against the wall of the tube with a sterile inoculation loop Air-dry the material in a laminar flow hood

13 Resolubilize DNA in TE-4 buffer and store material at 4 ordmC Reagents

Red Blood Cell Lysis Buffer 144 mmolL ammonium chloride 1 mmolL sodium bicarbonate in deionized (DI) water Store at 2 ordmC to 8 ordmC

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

47

Nucleic Lysis Buffer 10 mmolL Tris-HCl 400 mmolL sodium chloride 20 mmolL disodium ethylenediaminetetraacetic acid in DI Water adjust pH to 82 Store at 2 ordmC to 8 ordmC

Protease K Buffer

20 mmolL disodium ethylenediaminetetraacetic acid 10 (mass per volume) Sodium Dodecyl Sulfate (SDS) in DI water Store at 2 ordmC to 8 ordmC

Protease K

Sigma-Aldrich cat P6556 Tritirachium album lyophilized powder Store at ndash20 ordmC

Ammonium Acetate

192 molL ammonium acetate (saturated) in DI water Store at 2 ordmC to 8 ordmC

Absolute Ethanol (200 proof)

Store at room temperature Reference

A1 Miller SA Dykes DD Polesky HF A simple salting out procedure for extracting DNA from human nucleated cells Nucleic Acids Res 198816(3)1215 httpsdoiorg101093nar1631215

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

48

Appendix B Spectrophotometric Measurement Protocol The following is for use with a scanning recording spectrophotometer The procedures include the spectrophotometer calibration measurement of the ldquonativerdquo dsDNA and measurement of the NaOH denatured ssDNA The procedure for evaluating solutions after treatment with NaOH is adapted from references89 Spectrophotometer Calibration

Establish Baseline Set the spectrophotometer to acquire data from 220 nm to 345 nm in steps of 01 nm Zero the instrument with nothing (air) in the sample and reference cells Acquire an air-air baseline of the instrument with nothing (air) in the sample and reference cells If the trace has significant structure or excessive noise take corrective action If the trace is acceptable proceed with data collection Enable Verification of Wavelength Calibration SRM 2034 is a solution of holmium oxide in 10 perchloric acid contained in a sealed quartz cuvette delivering an intrinsic UVVis wavelength standard for absorption spectrophotometryB1 Evaluation of the location of peak maxima in the spectra of this solution enables verification of the spectrophotometers wavelength calibration Place SRM 2034 in the sample cell Create and name a file to hold the verification spectra Perform three replicate SRM 2034-to-air scans saving each scan to the file Remove SRM 2034 from the spectrophotometer and store appropriately Enable Verification of Absorbance Calibration SRM 2031 consists of three metal-on-fused silica filters held in cuvette-sized filter holders plus an empty filter holder The three filters are certified for transmittance and transmittance density (-log10(transmittance) effectively absorbance) at several wavelengths with nominal wavelength-independent transmittances of 10 30 and 90 Evaluation of the absorbance of each of these three filters at the certified 250 nm 280 nm and 340 nm wavelengths enables verification of the spectrophotometers absorbance calibration Place the empty SRM 2031 filter holder in the reference cell Remove the protective covers from the 10 transmittance filter and place in the sample cell Perform three replicate scans saving each scan to file Remove the filter from the spectrophotometer replace the protective covers and store appropriately Repeat for the 30 and 90 filters Close the file containing the verification spectra For ldquonativerdquo dsDNA Samples

Verify Optical Balance of Cuvettes The measurements of the DNA-containing samples are made using a 100 microL quartz cuvette in the sample cell compared to an optically matched cuvette in the reference cell The optical equivalence of these two cuvettes can be verified by filling both cuvettes with appropriate volumes of a sample solution matrix (TE-4 buffer) that does not contain DNA Fill two dry

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

49

clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes followed by a methanol rinse in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquonativerdquo dsDNA remove and clean both the reference cuvette and the sample cuvette For NaOH denatured ssDNA Samples

Prepare a 04 molL NaOH Stock Solution To a 15 mL disposable centrifuge tube add 016 g NaOH pellets add deionized (DI) water to the 10 mL line and mix until all the NaOH pellets have dissolved Centrifuge at 524 rads (5000 rpm) for 5 min For every 9 samples to be analyzed aliquot 1 mL of this 04 molL NaOH stock solution into a 15 mL screw top tube Use these tubes to prepare the NaOH blanks and DNA samples In a 15 mL screw cap vial mix 200 microL TE-4 buffer with 200 microL 04 molL sodium hydroxide solution to achieve a final NaOH concentration of 02 molL Vortex 5 s then centrifuge at 1257 rads (12 000 rpm) for 1 min Fill two dry clean quartz cuvettes with this solution Visually inspect the cell windows to ensure that no bubbles or stray material are present Zero the instrument with buffer in both the sample and reference cells Acquire a blank-to-blank baseline scan If the trace has significant structure or excessive noise take corrective action Once the trace is acceptable create and suitably name a file to hold the spectra for this series Perform three replicate scans saving each scan to file Prepare the Sample Solutions For each unit of material to be measured (Tube or vial)

a) Vortex the unit for 5 s b) Centrifuge at 1257 rads (12 000 rpm) for 1 min c) Remove a 100 microL aliquot and place in a new labeled vial d) Add 100 microL of the freshly prepared 04 molL NaOH solution (with the same pipet) e) Vortex 5 s f) Centrifuge at 1257 rads (12 000 rpm) for 1 min

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

50

Measure Sample Solutions Leave the filled reference cuvette in place Remove the sample cuvette clean and dry with a series of DI water washes rinse with methanol in a cuvette washer dry the cuvette with a vacuum assist To the dry cuvette add 100 microL of the NaOH denatured DNA solution to be measured Place the loaded cuvette in the spectrophotometer sample cell Perform three replicate scans saving each scan to file Remove the sample cuvette and repeat the cleaning and re-filling procedure for each sample measured After the last measurement of the ldquodenaturedrdquo ssDNA close the file holding the spectra for this series Remove and clean both the reference cuvette and the sample cuvette After All Sample Measurements are Complete

Enable Verification of Spectrophotometer Stability The stability of the spectrophotometer over the course of the data collection can be evaluated by comparing spectra for SRM 2034 and 2031 taken at the start of data collection with spectra taken at the end of data collection Repeat the wavelength and absorbance verification measurements described above collecting the spectra in a suitably named file Clean Up Transfer all relevant data files from the spectrophotometers storage to a transfer device Clean the spectrophotometer work area as necessary and restore the SRMs and cuvettes to their appropriate storage locations Evaluation

Export the files to a spreadsheet Average the replicate spectra Subtract the average absorbance at A320 from the absorbance at A260 resulting in the corrected absorbance at A260 The conventional spectrophotometric estimate of the mass concentration of the dsDNA in the sample is achieved by multiplying the corrected absorbance by 50 The conventional estimate for the NaOH-treated ssDNA is achieved by multiplying the corrected absorbance by 37times2 where the ldquotimes2rdquo adjusts for the 1rarr2 volumetric dilution of the sample Note reliable spectrophotometric measurements require corrected A260 to be greater than 005 Reference

B1 Travis JC Acosta JC Andor G Bastie J Blattner P Chunnilall CJ Crosson SC Duewer DL Early EA Hengstberger F Kim C-S Liedquist L Manoocheri F Mercader F Metzdorf J Mito A Monardh LAG Nevask S Nilsson M Noeumll M Rodriguez AC Ruiacutez A Schirmacher A Smith MV Valencia G van Tonder N Zwinkels J Intrinsic Wavelength Standard Absorption Bands in Holmium Oxide Solution for UVvisible Molecular Absorption Spectrophotometry J Phys Chem Ref Data 200534(1)41-56

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

51

Appendix C SRM 2372a Statisticianrsquos report Extracted from Blaza Toman

November 30 2017 1 Introduction Candidate material for SRM 2372a is composed of three components labeled A B and C All components are human genomic DNA and prepared at NIST from buffy coat white blood cells from single-source anonymous donors Component A and B are single source (malefemale respectively) and Component C is a 1 male to 3 female mixture of two single-source donors Droplet digital PCR (ddPCR) assays were used to obtain measurements of the concentration of copies per microliter of sample Ten NIST-developed human nuclear DNA and three literature-derived human mitochondrial DNA PCR assays were used in this study

11 OpenBUGS Analysis All parameters in the following sections were estimated using Markov Chain Monte Carlo (MCMC) implemented in the OpenBUGS freewareC1 OpenBUGS is a software application for the Bayesian analysis of complex statistical models

12 OpenBUGS Exemplar Output All following sections list the OpenBUGs model (ldquocoderdquo) initialization values (ldquoinitsrdquo) if they are needed and the data used to evaluate the various data sets discussed in the body of this report homogeneity stability nDNA certification and mtDNAnDNA value assignment The provided information in each of the following sections is sufficient to produce estimates but the estimates may not exactly reproduce the exemplar results provided The MCMC paradigm does not produce ldquoanalyticalrdquo values but approaches them asymptotically as the number o model updates increases All exemplar results presented here are based upon the following bull 10-fold ldquothinningrdquo to minimize autocorrelation there were 10 MCMC estimates for every

model update used in parameter estimation bull a ldquoburn inrdquo of 5 000 updates before beginning parameter estimation and bull 10 000 model updates

The following statistics are provided in the tables listing the exemplar results

mean arithmetic mean of the parameterrsquos model update values sd standard deviation of the parameterrsquos model update values MC_error Monte Carlo standard deviation of the mean meanradic(number of model updates) val25pc 25th percentile of the model update values median 50th percentile of the model update values val975pc 975th percentile of the model update values startnumber of ldquoburn inrdquo updates sample number of model update values

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

52

2 Homogeneity assessment 21 Data

Samples of the material of all three components were drawn from 22 different boxes Two assays (NEIF ND6) were used to obtain 3 replicates of the measurement in copies per nanoliter for each sample and component For the complete homogeneity data see

Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component

22 Model The data were analyzed using a three-level Gaussian hierarchical ANOVA model

119910119910119894119894119894119894119894119894~119873119873120572120572119894119894119894119894119894119894 120590120590119908119908119894119894119905119905ℎ119894119894119894119894 119887119887119887119887119887119887 1198941198941198941198942

where ~N() denotes ldquodistributed as a normal distribution of some mean μ and variancerdquo 119894119894 denotes component (1 for A 2 for B 3 for C) 119895119895 denotes assay (1 for NEIF 2 for ND6) 119896119896 denotes box (1hellip22) and 120572120572119894119894119894119894119894119894~119873119873micro119894119894119894119894120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119887119887119887119887119887119887119894119894

2

All variance components were given a Half Cauchy prior as described in Reference C2 Note The Half Cauchy prior is specified by ldquo~dnorm(0 00016)I(0001)rdquo The ldquoIrdquo notation is for censoring It cuts off any random draws from the Normal distribution that are lower than 0001 An informative prior is needed when there are less than four data per distribution and the half Cauchy is a good general-purpose prior

23 OpenBUGS code and data Homogeneity code ModelBegin for(i in 13)xiNsb[i] ~ dnorm(0 00016)I(0001) chSqNsb[i] ~ dgamma(0505) sigb[i] lt- xiNsb[i]sqrt(chSqNsb[i]) for(i in 13)for(j in 12)mu[ij]~dnorm(010E-5) for(i in 13) muA[i]lt-mean(mu[i])sigbP[i]lt-100(muA[i]sqrt(sigb[i])) for(i in 13)for(j in 12)xiNs[ij] ~ dnorm(0 00016)I(0001) chSqNs[ij]~dgamma(0505) sigt[ij]lt-xiNs[ij]sqrt(chSqNs[ij]) sigtP[ij]lt-100(mu[ij]sqrt(sigt[ij])) for(i in 13)for(j in 12)for(k in 122)alpha[ijk]~dnorm(mu[ij]sigb[i]) for(idx in 1393)lamb[idx]~dnorm(alpha[cmp[idx]asy[idx]box[idx]]sigt[cmp[idx]asy[idx]]) ModelEnd Inits None required

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

53

Homogeneity Data lamb[] asy[] box[] cmp[] rep[] 1632 1 1 1 1 1529 1 1 1 2 1511 1 1 1 3 1616 2 1 1 1 1640 2 1 1 2 1582 2 1 1 3 1531 1 2 1 1 1525 1 2 1 2 1549 1 2 1 3 1615 2 2 1 1 1589 2 2 1 2 1601 2 2 1 3 1524 1 3 1 1 1540 1 3 1 2 1533 1 3 1 3 1582 2 3 1 1 1152 2 3 1 2 1555 2 3 1 3 1454 1 4 1 1 1458 1 4 1 2 1427 1 4 1 3 1627 2 4 1 1 1605 2 4 1 2 1607 2 4 1 3 1465 1 5 1 1 1540 1 5 1 2 1504 1 5 1 3 1607 2 5 1 1 1633 2 5 1 2 1586 2 5 1 3 1512 1 6 1 1 1491 1 6 1 2 1491 1 6 1 3 1511 2 6 1 1 1483 2 6 1 2 1524 2 6 1 3 1418 1 7 1 1 1480 1 7 1 2 1455 1 7 1 3 1479 2 7 1 1 1557 2 7 1 2 1656 2 7 1 3 1496 1 8 1 1 1560 1 8 1 2 1501 1 8 1 3 1514 2 8 1 1 1615 2 8 1 2 1540 2 8 1 3 1588 1 9 1 1 1581 1 9 1 2 1637 1 9 1 3 1474 2 9 1 1 1438 2 9 1 2 1454 2 9 1 3 1574 1 10 1 1 1513 1 10 1 2 1570 1 10 1 3 1594 2 10 1 1 1713 2 10 1 2 1669 2 10 1 3

1557 1 11 1 1 1508 1 11 1 2 1633 1 11 1 3 1571 2 11 1 1 1611 2 11 1 2 1631 2 11 1 3 1500 1 12 1 1 1476 1 12 1 2 1515 1 12 1 3 1626 2 12 1 1 1586 2 12 1 2 1608 2 12 1 3 1672 1 13 1 1 1634 1 13 1 2 1804 1 13 1 3 1490 2 13 1 1 1516 2 13 1 2 1497 2 13 1 3 1551 1 14 1 1 1568 1 14 1 2 1538 1 14 1 3 1617 2 14 1 1 1614 2 14 1 2 1639 2 14 1 3 1456 1 15 1 1 1448 1 15 1 2 1523 1 15 1 3 1525 2 15 1 1 1402 2 15 1 2 1436 2 15 1 3 1573 1 16 1 1 1520 1 16 1 2 1617 1 16 1 3 1634 2 16 1 1 1627 2 16 1 2 1575 2 16 1 3 1516 1 17 1 1 1516 1 17 1 2 1628 1 17 1 3 1613 2 17 1 1 1567 2 17 1 2 1567 2 17 1 3 1612 1 18 1 1 1535 1 18 1 2 1672 1 18 1 3 1529 2 18 1 1 1495 2 18 1 2 1528 2 18 1 3 1598 1 19 1 1 1514 1 19 1 2 1596 1 19 1 3 1571 2 19 1 1 1535 2 19 1 2 1511 2 19 1 3 1575 1 20 1 1 1568 1 20 1 2 1586 1 20 1 3 1652 2 20 1 1 1545 2 20 1 2 1627 2 20 1 3 1658 1 21 1 1

1611 1 21 1 2 1578 1 21 1 3 1647 2 21 1 1 1607 2 21 1 2 1612 2 21 1 3 1547 1 22 1 1 1528 1 22 1 2 1532 1 22 1 3 1537 2 22 1 1 1513 2 22 1 2 1536 2 22 1 3 1790 1 1 2 1 1909 1 1 2 2 1885 1 1 2 3 1870 2 1 2 2 1860 2 1 2 3 1712 1 2 2 1 1719 1 2 2 2 1704 1 2 2 3 1775 2 2 2 1 1718 2 2 2 2 1756 2 2 2 3 1693 1 3 2 1 1755 1 3 2 2 1749 1 3 2 3 1824 2 3 2 1 1483 2 3 2 2 1896 2 3 2 3 1754 1 4 2 1 1745 1 4 2 2 1739 1 4 2 3 1670 2 4 2 1 1632 2 4 2 2 1814 2 4 2 3 1757 1 5 2 1 1687 1 5 2 2 1735 1 5 2 3 1683 2 5 2 1 1759 2 5 2 2 1703 2 5 2 3 1697 1 6 2 1 1709 1 6 2 2 1714 1 6 2 3 1693 2 6 2 1 1723 2 6 2 2 1722 2 6 2 3 1722 1 7 2 1 1625 1 7 2 2 1703 1 7 2 3 1690 2 7 2 1 1684 2 7 2 2 1688 2 7 2 3 1884 1 8 2 1 1802 1 8 2 2 1772 1 8 2 3 1864 2 8 2 1 1887 2 8 2 2 1821 2 8 2 3 1713 1 9 2 1 1734 1 9 2 2 1727 1 9 2 3

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

54

1709 2 9 2 1 1756 2 9 2 2 1758 2 9 2 3 1759 1 10 2 1 1809 1 10 2 2 1738 1 10 2 3 1764 2 10 2 1 1816 2 10 2 2 1787 2 10 2 3 1749 1 11 2 1 1782 1 11 2 2 1740 1 11 2 3 1741 2 11 2 1 1792 2 11 2 2 1784 2 11 2 3 1748 1 12 2 1 1712 1 12 2 2 1718 1 12 2 3 1797 2 12 2 1 1614 2 12 2 2 1790 2 12 2 3 1632 1 13 2 1 1668 1 13 2 2 1701 1 13 2 3 1626 2 13 2 1 1566 2 13 2 2 1681 2 13 2 3 1752 1 14 2 1 1535 1 14 2 2 1673 1 14 2 3 1724 2 14 2 1 1741 2 14 2 2 1685 2 14 2 3 1871 1 15 2 1 1835 1 15 2 2 1887 1 15 2 3 1770 2 15 2 1 1752 2 15 2 2 1798 2 15 2 3 1688 1 16 2 1 1725 1 16 2 2 1743 1 16 2 3 1744 2 16 2 1 1697 2 16 2 2 1710 2 16 2 3 1789 1 17 2 1 1769 1 17 2 2 1825 1 17 2 3 1716 2 17 2 1 1762 2 17 2 2 1638 2 17 2 3 1826 1 18 2 1 1780 1 18 2 2 1849 1 18 2 3 1805 2 18 2 1 1801 2 18 2 2 1931 2 18 2 3 1751 1 19 2 1 1759 1 19 2 2 1880 1 19 2 3 1827 2 19 2 1 1766 2 19 2 2

1786 2 19 2 3 1738 1 20 2 1 1678 1 20 2 2 1768 1 20 2 3 1752 2 20 2 1 1747 2 20 2 2 1768 2 20 2 3 1742 1 21 2 1 1836 1 21 2 2 1713 1 21 2 3 1746 2 21 2 1 1744 2 21 2 2 1764 2 21 2 3 1761 1 22 2 1 1743 1 22 2 2 1660 1 22 2 3 1666 2 22 2 1 1692 2 22 2 2 1819 2 22 2 3 1493 1 1 3 1 1385 1 1 3 2 1510 1 1 3 3 1565 2 1 3 1 1562 2 1 3 2 1466 2 1 3 3 1516 1 2 3 1 1513 1 2 3 2 1577 1 2 3 3 1490 2 2 3 1 1547 2 2 3 2 1570 2 2 3 3 1451 1 3 3 1 1528 1 3 3 2 1506 1 3 3 3 1485 2 3 3 1 1505 2 3 3 2 1471 2 3 3 3 1440 1 4 3 1 1486 1 4 3 2 1520 1 4 3 3 1529 2 4 3 1 1508 2 4 3 2 1506 2 4 3 3 1523 1 5 3 1 1472 1 5 3 2 1483 1 5 3 3 1513 2 5 3 1 1520 2 5 3 2 1491 2 5 3 3 1437 1 6 3 1 1493 1 6 3 2 1522 1 6 3 3 1492 2 6 3 1 1481 2 6 3 2 1461 2 6 3 3 1485 1 7 3 1 1477 1 7 3 2 1463 1 7 3 3 1475 2 7 3 1 1547 2 7 3 2 1485 2 7 3 3 1433 1 8 3 1

1405 1 8 3 2 1478 1 8 3 3 1497 2 8 3 1 1501 2 8 3 2 1495 2 8 3 3 1463 1 9 3 1 1499 1 9 3 2 1505 1 9 3 3 1503 2 9 3 1 1491 2 9 3 2 1516 2 9 3 3 1534 1 10 3 1 1458 1 10 3 2 1454 1 10 3 3 1448 2 10 3 1 1465 2 10 3 2 1474 2 10 3 3 1533 1 11 3 1 1514 1 11 3 2 1555 1 11 3 3 1488 2 11 3 1 1470 2 11 3 2 1483 2 11 3 3 1450 1 12 3 1 1525 1 12 3 2 1497 1 12 3 3 1496 2 12 3 1 1483 2 12 3 2 1519 2 12 3 3 1537 1 13 3 1 1527 1 13 3 2 1515 1 13 3 3 1499 2 13 3 1 1479 2 13 3 2 1475 2 13 3 3 1515 1 14 3 1 1490 1 14 3 2 1494 1 14 3 3 1469 2 14 3 1 1480 2 14 3 2 1466 2 14 3 3 1511 1 15 3 1 1568 1 15 3 2 1594 1 15 3 3 1515 2 15 3 1 1543 2 15 3 2 1565 2 15 3 3 1449 1 16 3 1 1452 1 16 3 2 1471 1 16 3 3 1543 2 16 3 1 1448 2 16 3 2 1491 2 16 3 3 1388 1 17 3 1 1421 1 17 3 2 1394 1 17 3 3 1452 2 17 3 1 1435 2 17 3 3 1475 1 18 3 1 1455 1 18 3 2 1445 1 18 3 3 1458 2 18 3 1

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

55

1531 2 18 3 2 1459 2 18 3 3 1468 1 19 3 1 1525 1 19 3 2 1446 1 19 3 3 1435 2 19 3 1 1478 2 19 3 2 1421 2 19 3 3 1549 1 20 3 1

1531 1 20 3 2 1499 1 20 3 3 1499 2 20 3 1 1465 2 20 3 2 1468 2 20 3 3 1530 1 21 3 1 1440 1 21 3 2 1484 1 21 3 3 1499 2 21 3 1

1454 2 21 3 2 1442 1 22 3 1 1487 1 22 3 2 1436 1 22 3 3 1589 2 22 3 1 1591 2 22 3 2 1556 2 22 3 3 END

Table C-1 Example OpenBUGS Homogeneity Summary

mean sd MC_error val25pc median val975pc start sample mu[11] 15460 0122 000123 1522 1546 1570 5001 10000 mu[12] 15640 0137 000141 1537 1564 1591 5001 10000 mu[21] 17490 0113 000118 1727 1749 1772 5001 10000 mu[22] 17480 0129 000125 1722 1748 1774 5001 10000 mu[31] 14870 0073 000073 1473 1487 1501 5001 10000 mu[32] 14960 0069 000063 1482 1496 1509 5001 10000 muA[1] 15550 0092 000092 1536 1555 1573 5001 10000 muA[2] 17490 0085 000086 1732 1749 1765 5001 10000 muA[3] 14910 0050 000046 1481 1491 1501 5001 10000 sigbP[1] 3269 04985 000514 2376 3244 4324 5001 10000 sigbP[2] 2548 04385 000461 1733 2533 3472 5001 10000 sigbP[3] 1850 0292 000260 1329 1832 2477 5001 10000

sigtP[11] 2774 0298 000260 2268 2750 3442 5001 10000 sigtP[12] 4036 0429 000438 3290 4005 4986 5001 10000 sigtP[21] 2664 0301 000321 2155 2640 3313 5001 10000 sigtP[22] 3807 0395 000377 3116 3772 4662 5001 10000 sigtP[31] 2313 0250 000209 1880 2291 2873 5001 10000 sigtP[32] 1812 0195 000170 1473 1797 2243 5001 10000

24 Results The material appears to be homogeneous as can be seen in Figure C-1 Based on these results the between-tube variability is of the same order as within-tube variability

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

56

Component A Assay NEIF Component A Assay ND6 box plot gamma[11]

[111][112][113]

[114]

[115][116]

[117]

[118]

[119]

[1110][1111]

[1112]

[1113]

[1114]

[1115]

[1116][1117]

[1118]

[1119][1120]

[1121]

[1122]

114

box plot gamma[12]

[121][122]

[123]

[124][125]

[126]

[127][128]

[129]

[1210]

[1211][1212]

[1213]

[1214]

[1215]

[1216]

[1217]

[1218][1219]

[1220][1221]

[1222]

115

Component B Assay NEIF Component B Assay ND6

box plot gamma[21]

[211]

[212]

[213][214]

[215][216]

[217]

[218]

[219]

[2110][2111]

[2112]

[2113][2114]

[2115]

[2116]

[2117]

[2118]

[2119]

[2120]

[2121]

[2122]

129

box plot gamma[22]

[221]

[222][223]

[224][225][226][227]

[228]

[229]

[2210][2211]

[2212]

[2213]

[2214]

[2215]

[2216][2217]

[2218]

[2219]

[2220][2221][2222]

128

Component C Assay NEIF Component C Assay ND6

box plot gamma[31]

[311]

[312]

[313][314]

[315][316]

[317]

[318]

[319][3110]

[3111]

[3112]

[3113]

[3114]

[3115]

[3116]

[3117]

[3118]

[3119]

[3120]

[3121]

[3122]109

box plot gamma[32]

[321][322]

[323]

[324][325]

[326]

[327][328][329]

[3210]

[3211]

[3212]

[3213][3214]

[3215]

[3216]

[3217]

[3218]

[3219]

[3220][3221]

[3222]

11

Figure C-1 Homogeneity Assessment

Note These results were obtained using the 1-4 gravimetric dilution-adjusted λ but uncorrected for droplet volume

or the 1-10 volumetric dilution That is the vertical axis in each panel should be multiplied by 07349times10 = 7349 This constant difference in scale does not impact the homogeneity analysis Error bars span approximate 95 confidence intervals

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

57

3 Certification measurements 31 Data

Originally the experiment consisted of 5 plates containing the 3 components (A B C) and 12 assays Ten of the assays were for genomic DNA identification (for the certified value) and two of the assays were for mitochondrial detection which is used for informational values (and are not included in this data set) Each of the plates contained samples from a new vial from different boxes (noted in the ldquovial usedrdquo portion the number references the box) The assay positions were alternated across the five plates as well For the plate design and which tubes were used see

Figure 6 Basic Plate Layout Design for Certification Measurements It was later determined that the homogeneity and stability data sets were obtained using a different set of pipettes than the certification data set and since all data was to be combined for certification another plate was obtained using the original pipette For the complete certification data see

Table 13 Certification Data as λacute Copy Number per Nanoliter of Component Stability data was measured at different temperatures over several weeks using assays NEIF and ND6 Only the 4degC data was also used for certification For the complete stability data see

Table 12 Stability Data as λacute Copy Number per Nanoliter of Component

32 Model It was decided that all three sets of data should be used to derive the certified values In this way the uncertainty of the certified values would better reflect any potential heterogeneity in the samples as well as any additional unknown factors that may have effected the results Because two different pipettes were used to obtain different sets of data there was a significant bias in some of the measurements Specifically the ldquocertificationrdquo data from the first five plates was obtained using pipette set ldquo1rdquo and the remaining measurements were obtained using pipette set ldquo2rdquo This bias was introduced into the statistical model as part of the mean Possible effects of different assays and type of data (heterogeneity certification and stability) were accounted for in the statistical model in terms of a random effect Specifically the response for each component (A B C) was represented as

119910119910119894119894119894119894119894119894~1198731198731205741205741198941198941198941198941198941198941205901205901198941198942 where ~N() denotes ldquodistributed as a normal distribution of some mean and variancerdquo i = 123 denotes type of measurement heterogeneity certification and stability j = 1hellip10 denotes assay 2PR4 POTP NEIF R4Q5 D5 ND6 D9 HBB1 ND14 22C3 k = 1 2 denotes pipette set 120574120574119894119894119894119894119894119894 = 120572120572119894119894119894119894 + 120573120573119894119894 120572120572119894119894119894119894~119873119873(micro1198941198941205901205901205721205722) 120573120573119894119894 = 120573120573 for data type 1 and 120573120573119894119894 = 0 for data types 2 and 3 For a given 2372a component the variance 1205901205901205721205722 measures the between assay variability and 1205901205901198941198942 measures the repeatability variability for each data type The data was analyzed using Bayesian MCMC programmed in OpenBUGSC2 with non-informative Gaussian priors for the means and Gamma priors for the variance components

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

58

33 OpenBUGS code inits and data Certification code ModelBegin for(i in 13)mu[i]~dnorm(010E-5) sig[i]~dgamma(10E-510E-5) sigalph~dgamma(10E-510E-5) for(i in 110)alpha[1i]~dnorm(mu[1] sigalph) for(i in 12)alpha[2i]~dnorm(mu[2] sigalph) alpha[3i]~dnorm(mu[3]sigalph) b~dnorm(010E-5) beta[1]lt-b beta[2]lt-00 for(i in 1Ntot)mean[i]lt-alpha[typ[i]asy[i]]+beta[pip[i]] for(i in 1Ntot)lamb[i]~dnorm(mean[i]sig[typ[i]]) ModelEnd Certification Inits list(sig=c(111) sigalph=1) Certification Data for Component A list(Ntot=275) lamb[] asy[] cmp[] typ[] pip[] 1667 1 1 1 1 1655 2 1 1 1 1647 3 1 1 1 1594 4 1 1 1 1655 5 1 1 1 1629 6 1 1 1 1655 7 1 1 1 1644 8 1 1 1 1645 9 1 1 1 1545 10 1 1 1 1698 1 1 1 1 1685 2 1 1 1 1665 3 1 1 1 1658 4 1 1 1 1683 5 1 1 1 1647 6 1 1 1 1652 7 1 1 1 1611 8 1 1 1 1707 9 1 1 1 1603 10 1 1 1 1668 1 1 1 1 1587 2 1 1 1 1678 3 1 1 1 1628 4 1 1 1 1738 5 1 1 1 1667 6 1 1 1 1695 7 1 1 1 1757 8 1 1 1 1771 9 1 1 1 1762 10 1 1 1 1739 1 1 1 1 1702 2 1 1 1 1614 3 1 1 1 1762 4 1 1 1 1649 5 1 1 1 1678 6 1 1 1 1653 7 1 1 1

175 8 1 1 1 1764 9 1 1 1 1759 10 1 1 1 1638 1 1 1 1 1676 2 1 1 1 1616 3 1 1 1 1559 4 1 1 1 1608 5 1 1 1 1611 6 1 1 1 1517 7 1 1 1 1557 8 1 1 1 1604 9 1 1 1 1675 10 1 1 1 1667 1 1 1 1 1638 2 1 1 1 1624 3 1 1 1 1587 4 1 1 1 1652 5 1 1 1 1589 6 1 1 1 1607 7 1 1 1 1582 8 1 1 1 1625 9 1 1 1 1616 10 1 1 1 1888 1 1 1 1 1868 2 1 1 1 1953 3 1 1 1 1779 4 1 1 1 1924 5 1 1 1 1861 6 1 1 1 1824 7 1 1 1 1809 8 1 1 1 1864 9 1 1 1 193 10 1 1 1 1939 1 1 1 1 1815 2 1 1 1 1902 3 1 1 1 1871 4 1 1 1 2124 5 1 1 1

1787 6 1 1 1 1844 7 1 1 1 1799 8 1 1 1 1857 9 1 1 1 1942 10 1 1 1 1597 1 1 1 1 1533 2 1 1 1 1462 3 1 1 1 1493 4 1 1 1 1538 5 1 1 1 1492 6 1 1 1 1528 7 1 1 1 1548 8 1 1 1 1513 9 1 1 1 1562 10 1 1 1 159 1 1 1 1 1404 2 1 1 1 1533 3 1 1 1 1519 4 1 1 1 1586 5 1 1 1 1516 6 1 1 1 1538 7 1 1 1 1581 8 1 1 1 1514 9 1 1 1 1431 1 1 1 2 1407 2 1 1 2 1487 3 1 1 2 145 4 1 1 2 1394 5 1 1 2 1396 6 1 1 2 1414 7 1 1 2 1401 8 1 1 2 1412 9 1 1 2 1435 10 1 1 2 1411 1 1 1 2 1395 2 1 1 2 1398 3 1 1 2 1439 4 1 1 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

59

1422 5 1 1 2 1393 6 1 1 2 139 7 1 1 2 1382 8 1 1 2 146 9 1 1 2 141 10 1 1 2 1632 1 1 2 2 1531 1 1 2 2 1524 1 1 2 2 1454 1 1 2 2 1465 1 1 2 2 1512 1 1 2 2 1418 1 1 2 2 1496 1 1 2 2 1588 1 1 2 2 1574 1 1 2 2 1557 1 1 2 2 15 1 1 2 2 1672 1 1 2 2 1551 1 1 2 2 1573 1 1 2 2 1516 1 1 2 2 1612 1 1 2 2 1598 1 1 2 2 1575 1 1 2 2 1658 1 1 2 2 1547 1 1 2 2 1529 1 1 2 2 1525 1 1 2 2 154 1 1 2 2 1458 1 1 2 2 154 1 1 2 2 1491 1 1 2 2 148 1 1 2 2 156 1 1 2 2 1581 1 1 2 2 1513 1 1 2 2 1508 1 1 2 2 1476 1 1 2 2 1634 1 1 2 2 1568 1 1 2 2 152 1 1 2 2 1516 1 1 2 2 1535 1 1 2 2 1514 1 1 2 2 1568 1 1 2 2 1611 1 1 2 2 1528 1 1 2 2 1511 1 1 2 2 1549 1 1 2 2 1533 1 1 2 2 1427 1 1 2 2 1504 1 1 2 2 1491 1 1 2 2 1455 1 1 2 2

1501 1 1 2 2 1637 1 1 2 2 157 1 1 2 2 1633 1 1 2 2 1515 1 1 2 2 1804 1 1 2 2 1538 1 1 2 2 1617 1 1 2 2 1628 1 1 2 2 1672 1 1 2 2 1596 1 1 2 2 1586 1 1 2 2 1578 1 1 2 2 1532 1 1 2 2 1616 2 1 2 2 1615 2 1 2 2 1582 2 1 2 2 1627 2 1 2 2 1607 2 1 2 2 1511 2 1 2 2 1479 2 1 2 2 1514 2 1 2 2 1474 2 1 2 2 1594 2 1 2 2 1571 2 1 2 2 1626 2 1 2 2 149 2 1 2 2 1617 2 1 2 2 1634 2 1 2 2 1613 2 1 2 2 1529 2 1 2 2 1571 2 1 2 2 1652 2 1 2 2 1647 2 1 2 2 1537 2 1 2 2 164 2 1 2 2 1589 2 1 2 2 1152 2 1 2 2 1605 2 1 2 2 1633 2 1 2 2 1483 2 1 2 2 1557 2 1 2 2 1615 2 1 2 2 1438 2 1 2 2 1713 2 1 2 2 1611 2 1 2 2 1586 2 1 2 2 1516 2 1 2 2 1614 2 1 2 2 1627 2 1 2 2 1567 2 1 2 2 1495 2 1 2 2 1535 2 1 2 2 1545 2 1 2 2 1607 2 1 2 2

1513 2 1 2 2 1582 2 1 2 2 1601 2 1 2 2 1555 2 1 2 2 1607 2 1 2 2 1586 2 1 2 2 1524 2 1 2 2 1656 2 1 2 2 154 2 1 2 2 1454 2 1 2 2 1669 2 1 2 2 1631 2 1 2 2 1608 2 1 2 2 1497 2 1 2 2 1639 2 1 2 2 1575 2 1 2 2 1567 2 1 2 2 1528 2 1 2 2 1511 2 1 2 2 1627 2 1 2 2 1612 2 1 2 2 1536 2 1 2 2 1568 1 1 3 2 152 1 1 3 2 1486 1 1 3 2 1441 1 1 3 2 1706 1 1 3 2 144 1 1 3 2 15 1 1 3 2 1537 1 1 3 2 1387 1 1 3 2 1463 1 1 3 2 157 1 1 3 2 1429 1 1 3 2 1667 1 1 3 2 1632 1 1 3 2 1589 1 1 3 2 1474 2 1 3 2 1543 2 1 3 2 1555 2 1 3 2 1599 2 1 3 2 1494 2 1 3 2 1624 2 1 3 2 1577 2 1 3 2 1593 2 1 3 2 1479 2 1 3 2 1517 2 1 3 2 1538 2 1 3 2 159 2 1 3 2 1479 2 1 3 2 1479 2 1 3 2 1522 2 1 3 2 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

60

Table C-2 Example OpenBUGS Certification Summary for Component A

mean sd MC_error val25pc median val975pc start sample b 2584 02948 0005376 2001 2588 3158 100001 100000 mu[1] 1417 02706 0005366 1364 1416 147 100001 100000 mu[2] 1559 008368 4776E-4 1543 1559 1576 100001 100000 mu[3] 1533 01504 0001419 1504 1533 1563 100001 100000

Certification Data for Component B list(Ntot=278) lamb[] asy[] cmp[] typ[] pip[] 1869 1 2 1 1 185 2 2 1 1 1709 3 2 1 1 1856 4 2 1 1 1848 5 2 1 1 1828 6 2 1 1 1861 7 2 1 1 1825 8 2 1 1 1871 9 2 1 1 1776 10 2 1 1 196 1 2 1 1 1781 2 2 1 1 1707 3 2 1 1 1787 4 2 1 1 1829 5 2 1 1 1826 6 2 1 1 1858 7 2 1 1 1869 8 2 1 1 1773 9 2 1 1 1902 10 2 1 1 2174 1 2 1 1 2202 2 2 1 1 2172 3 2 1 1 223 5 2 1 1 2094 6 2 1 1 1949 7 2 1 1 1971 8 2 1 1 2123 9 2 1 1 2107 10 2 1 1 2231 1 2 1 1 2292 2 2 1 1 2216 3 2 1 1 2078 4 2 1 1 2192 5 2 1 1 2034 6 2 1 1 211 7 2 1 1 2123 8 2 1 1 2066 9 2 1 1 2046 10 2 1 1 181 1 2 1 1 1865 2 2 1 1 1806 3 2 1 1 1878 4 2 1 1 1824 5 2 1 1 1768 6 2 1 1 1811 7 2 1 1 1763 8 2 1 1 1865 9 2 1 1 1815 10 2 1 1

1806 1 2 1 1 1801 2 2 1 1 1786 3 2 1 1 1768 4 2 1 1 179 5 2 1 1 167 6 2 1 1 1749 7 2 1 1 1741 8 2 1 1 1781 9 2 1 1 1866 10 2 1 1 1784 1 2 1 1 1735 2 2 1 1 1778 3 2 1 1 1835 4 2 1 1 1794 5 2 1 1 17 6 2 1 1 1799 7 2 1 1 1765 8 2 1 1 1754 9 2 1 1 1792 10 2 1 1 1821 1 2 1 1 1813 2 2 1 1 1813 3 2 1 1 1749 4 2 1 1 1848 5 2 1 1 1711 6 2 1 1 1779 7 2 1 1 1769 8 2 1 1 1725 9 2 1 1 1814 10 2 1 1 1718 1 2 1 1 1753 2 2 1 1 1773 3 2 1 1 1777 4 2 1 1 1803 5 2 1 1 1718 6 2 1 1 1802 7 2 1 1 1834 8 2 1 1 1731 9 2 1 1 1842 10 2 1 1 1768 1 2 1 1 1698 2 2 1 1 1792 3 2 1 1 174 4 2 1 1 1795 5 2 1 1 1693 6 2 1 1 1743 7 2 1 1 1745 8 2 1 1 1714 9 2 1 1 1791 1 2 1 2

1746 2 2 1 2 1803 3 2 1 2 1767 4 2 1 2 1745 5 2 1 2 1824 6 2 1 2 1752 7 2 1 2 1813 8 2 1 2 1648 9 2 1 2 1818 10 2 1 2 177 1 2 1 2 182 2 2 1 2 1803 3 2 1 2 1739 4 2 1 2 1773 5 2 1 2 1729 6 2 1 2 1744 7 2 1 2 1754 8 2 1 2 1671 9 2 1 2 1741 10 2 1 2 179 1 2 2 2 1712 1 2 2 2 1693 1 2 2 2 1754 1 2 2 2 1757 1 2 2 2 1697 1 2 2 2 1722 1 2 2 2 1884 1 2 2 2 1713 1 2 2 2 1759 1 2 2 2 1749 1 2 2 2 1748 1 2 2 2 1632 1 2 2 2 1752 1 2 2 2 1871 1 2 2 2 1688 1 2 2 2 1789 1 2 2 2 1826 1 2 2 2 1751 1 2 2 2 1738 1 2 2 2 1742 1 2 2 2 1761 1 2 2 2 1909 1 2 2 2 1719 1 2 2 2 1755 1 2 2 2 1745 1 2 2 2 1687 1 2 2 2 1709 1 2 2 2 1625 1 2 2 2 1802 1 2 2 2 1734 1 2 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

61

1809 1 2 2 2 1782 1 2 2 2 1712 1 2 2 2 1668 1 2 2 2 1535 1 2 2 2 1835 1 2 2 2 1725 1 2 2 2 1769 1 2 2 2 178 1 2 2 2 1759 1 2 2 2 1678 1 2 2 2 1836 1 2 2 2 1743 1 2 2 2 1885 1 2 2 2 1704 1 2 2 2 1749 1 2 2 2 1739 1 2 2 2 1735 1 2 2 2 1714 1 2 2 2 1703 1 2 2 2 1772 1 2 2 2 1727 1 2 2 2 1738 1 2 2 2 174 1 2 2 2 1718 1 2 2 2 1701 1 2 2 2 1673 1 2 2 2 1887 1 2 2 2 1743 1 2 2 2 1825 1 2 2 2 1849 1 2 2 2 188 1 2 2 2 1768 1 2 2 2 1713 1 2 2 2 166 1 2 2 2 1775 2 2 2 2 1824 2 2 2 2 167 2 2 2 2 1683 2 2 2 2 1693 2 2 2 2 169 2 2 2 2 1864 2 2 2 2 1709 2 2 2 2 1764 2 2 2 2

1741 2 2 2 2 1797 2 2 2 2 1626 2 2 2 2 1724 2 2 2 2 177 2 2 2 2 1744 2 2 2 2 1716 2 2 2 2 1805 2 2 2 2 1827 2 2 2 2 1752 2 2 2 2 1746 2 2 2 2 1666 2 2 2 2 187 2 2 2 2 1718 2 2 2 2 1483 2 2 2 2 1632 2 2 2 2 1759 2 2 2 2 1723 2 2 2 2 1684 2 2 2 2 1887 2 2 2 2 1756 2 2 2 2 1816 2 2 2 2 1792 2 2 2 2 1614 2 2 2 2 1566 2 2 2 2 1741 2 2 2 2 1752 2 2 2 2 1697 2 2 2 2 1762 2 2 2 2 1801 2 2 2 2 1766 2 2 2 2 1747 2 2 2 2 1744 2 2 2 2 1692 2 2 2 2 186 2 2 2 2 1756 2 2 2 2 1896 2 2 2 2 1814 2 2 2 2 1703 2 2 2 2 1722 2 2 2 2 1688 2 2 2 2 1821 2 2 2 2 1758 2 2 2 2 1787 2 2 2 2

1784 2 2 2 2 179 2 2 2 2 1681 2 2 2 2 1685 2 2 2 2 1798 2 2 2 2 171 2 2 2 2 1638 2 2 2 2 1931 2 2 2 2 1786 2 2 2 2 1768 2 2 2 2 1764 2 2 2 2 1819 2 2 2 2 1666 1 2 3 2 1626 1 2 3 2 188 1 2 3 2 1653 1 2 3 2 1733 1 2 3 2 1551 1 2 3 2 1884 1 2 3 2 1718 1 2 3 2 1549 1 2 3 2 1723 1 2 3 2 194 1 2 3 2 164 1 2 3 2 1565 1 2 3 2 1787 1 2 3 2 1655 2 2 3 2 167 2 2 3 2 1673 2 2 3 2 1705 2 2 3 2 1755 2 2 3 2 1625 2 2 3 2 1708 2 2 3 2 1705 2 2 3 2 1862 2 2 3 2 1686 2 2 3 2 1723 2 2 3 2 1909 2 2 3 2 1859 2 2 3 2 1897 2 2 3 2 1972 2 2 3 2 END

Table C-3 Example OpenBUGS Certification Summary for Component B

mean sd MC_error val25pc median val975pc start sample b 09337 03387 0007504 02695 09371 1592 100001 100000 mu[1] 1764 03101 0007483 1704 1763 1825 100001 100000 mu[2] 1748 008014 4719E-4 1733 1748 1764 100001 100000 mu[3] 1735 02338 0003093 1688 1736 1781 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

62

Certification data for Component C list(Ntot=280) lamb[] asy[] cmp[] typ[] pip[] 163 1 3 1 1 1596 2 3 1 1 1574 3 3 1 1 159 4 3 1 1 161 5 3 1 1 1589 6 3 1 1 1569 7 3 1 1 1565 8 3 1 1 1564 9 3 1 1 1548 10 3 1 1 1706 1 3 1 1 1576 2 3 1 1 1538 3 3 1 1 1555 4 3 1 1 1607 5 3 1 1 1548 6 3 1 1 1572 7 3 1 1 1681 8 3 1 1 1492 9 3 1 1 1655 10 3 1 1 1499 1 3 1 1 1576 2 3 1 1 1545 3 3 1 1 1584 4 3 1 1 1577 5 3 1 1 1599 6 3 1 1 1587 7 3 1 1 152 8 3 1 1 1479 9 3 1 1 1512 10 3 1 1 1497 1 3 1 1 1611 2 3 1 1 156 3 3 1 1 1546 4 3 1 1 1529 5 3 1 1 1478 6 3 1 1 1504 7 3 1 1 1551 8 3 1 1 1438 9 3 1 1 1494 10 3 1 1 1452 1 3 1 1 1347 2 3 1 1 1395 3 3 1 1 1442 4 3 1 1 1448 5 3 1 1 1463 6 3 1 1 1402 7 3 1 1 1474 8 3 1 1 1466 9 3 1 1 1439 10 3 1 1 1424 1 3 1 1 1397 2 3 1 1 135 3 3 1 1 1459 4 3 1 1 1602 5 3 1 1 1435 6 3 1 1 1433 7 3 1 1 1481 8 3 1 1 1482 9 3 1 1

1568 10 3 1 1 147 1 3 1 1 1411 2 3 1 1 1512 3 3 1 1 1452 4 3 1 1 1567 5 3 1 1 148 6 3 1 1 1442 7 3 1 1 143 8 3 1 1 1413 9 3 1 1 1486 10 3 1 1 1444 1 3 1 1 1427 2 3 1 1 1439 3 3 1 1 1468 4 3 1 1 1524 5 3 1 1 1424 6 3 1 1 1459 7 3 1 1 1467 8 3 1 1 1411 9 3 1 1 148 10 3 1 1 1485 1 3 1 1 1422 2 3 1 1 1438 3 3 1 1 1439 4 3 1 1 1515 5 3 1 1 1409 6 3 1 1 1469 7 3 1 1 1453 8 3 1 1 1465 9 3 1 1 1363 10 3 1 1 1458 1 3 1 1 1459 2 3 1 1 1465 3 3 1 1 1468 4 3 1 1 1457 5 3 1 1 1412 6 3 1 1 1474 7 3 1 1 1428 8 3 1 1 1425 9 3 1 1 1393 10 3 1 1 1357 1 3 1 2 1447 2 3 1 2 1448 3 3 1 2 1409 4 3 1 2 1389 5 3 1 2 1415 6 3 1 2 1431 7 3 1 2 1381 8 3 1 2 1403 9 3 1 2 1435 10 3 1 2 1364 1 3 1 2 1361 2 3 1 2 1404 3 3 1 2 1411 4 3 1 2 1429 5 3 1 2 1417 6 3 1 2 138 7 3 1 2 1336 8 3 1 2 1387 9 3 1 2

1372 10 3 1 2 1493 1 3 2 2 1516 1 3 2 2 1451 1 3 2 2 144 1 3 2 2 1523 1 3 2 2 1437 1 3 2 2 1485 1 3 2 2 1433 1 3 2 2 1463 1 3 2 2 1534 1 3 2 2 1533 1 3 2 2 145 1 3 2 2 1537 1 3 2 2 1515 1 3 2 2 1511 1 3 2 2 1449 1 3 2 2 1388 1 3 2 2 1475 1 3 2 2 1468 1 3 2 2 1549 1 3 2 2 153 1 3 2 2 1442 1 3 2 2 1385 1 3 2 2 1513 1 3 2 2 1528 1 3 2 2 1486 1 3 2 2 1472 1 3 2 2 1493 1 3 2 2 1477 1 3 2 2 1405 1 3 2 2 1499 1 3 2 2 1458 1 3 2 2 1514 1 3 2 2 1525 1 3 2 2 1527 1 3 2 2 149 1 3 2 2 1568 1 3 2 2 1452 1 3 2 2 1421 1 3 2 2 1455 1 3 2 2 1525 1 3 2 2 1531 1 3 2 2 144 1 3 2 2 1487 1 3 2 2 151 1 3 2 2 1577 1 3 2 2 1506 1 3 2 2 152 1 3 2 2 1483 1 3 2 2 1522 1 3 2 2 1463 1 3 2 2 1478 1 3 2 2 1505 1 3 2 2 1454 1 3 2 2 1555 1 3 2 2 1497 1 3 2 2 1515 1 3 2 2 1494 1 3 2 2 1594 1 3 2 2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

63

1471 1 3 2 2 1394 1 3 2 2 1445 1 3 2 2 1446 1 3 2 2 1499 1 3 2 2 1484 1 3 2 2 1436 1 3 2 2 1565 2 3 2 2 149 2 3 2 2 1485 2 3 2 2 1529 2 3 2 2 1513 2 3 2 2 1492 2 3 2 2 1475 2 3 2 2 1497 2 3 2 2 1503 2 3 2 2 1448 2 3 2 2 1488 2 3 2 2 1496 2 3 2 2 1499 2 3 2 2 1469 2 3 2 2 1515 2 3 2 2 1543 2 3 2 2 1452 2 3 2 2 1458 2 3 2 2 1435 2 3 2 2 1499 2 3 2 2 1499 2 3 2 2 1589 2 3 2 2 1562 2 3 2 2 1547 2 3 2 2 1505 2 3 2 2 1508 2 3 2 2 152 2 3 2 2

1481 2 3 2 2 1547 2 3 2 2 1501 2 3 2 2 1491 2 3 2 2 1465 2 3 2 2 147 2 3 2 2 1483 2 3 2 2 1479 2 3 2 2 148 2 3 2 2 1543 2 3 2 2 1448 2 3 2 2 1531 2 3 2 2 1478 2 3 2 2 1465 2 3 2 2 1454 2 3 2 2 1591 2 3 2 2 1466 2 3 2 2 157 2 3 2 2 1471 2 3 2 2 1506 2 3 2 2 1491 2 3 2 2 1461 2 3 2 2 1485 2 3 2 2 1495 2 3 2 2 1516 2 3 2 2 1474 2 3 2 2 1483 2 3 2 2 1519 2 3 2 2 1475 2 3 2 2 1466 2 3 2 2 1565 2 3 2 2 1491 2 3 2 2 1435 2 3 2 2 1459 2 3 2 2

1421 2 3 2 2 1468 2 3 2 2 1556 2 3 2 2 1424 1 3 3 2 1427 1 3 3 2 1381 1 3 3 2 1394 1 3 3 2 1473 1 3 3 2 133 1 3 3 2 1446 1 3 3 2 1424 1 3 3 2 133 1 3 3 2 1452 1 3 3 2 1512 1 3 3 2 1314 1 3 3 2 1525 1 3 3 2 1475 1 3 3 2 1542 1 3 3 2 1461 2 3 3 2 1443 2 3 3 2 1547 2 3 3 2 1394 2 3 3 2 1364 2 3 3 2 1395 2 3 3 2 1522 2 3 3 2 1483 2 3 3 2 1452 2 3 3 2 1482 2 3 3 2 1439 2 3 3 2 1458 2 3 3 2 155 2 3 3 2 1502 2 3 3 2 152 2 3 3 2 END

Table C-4 Example OpenBUGS Certification Summary for Component C

mean sd MC_error val25pc median val975pc start sample b 09626 01701 0002428 06285 09633 1293 100001 100000 mu[1] 1399 01566 0002413 1369 1399 143 100001 100000 mu[2] 1491 005701 2386E-4 148 1491 1503 100001 100000 mu[3] 1449 0132 0001181 1423 1449 1475 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

64

34 Results Table C-5 lists the results in CopiesnL for the three components

Table C-5 Parameter Values for Components A B and C

Posterior Distribution Summary Statistics Component Parameter Mean SD 25 50 975

A

β 260 029 203 260 318 μ1 1415 027 1362 1415 1467 μ2 1559 010 1539 1559 158 μ3 1533 016 1501 1533 1565

B

β 095 034 029 095 162 μ1 1762 031 1701 1762 1822 μ2 1748 010 1728 1748 1769 μ3 1735 024 1687 1735 1783

C

β 097 017 064 097 130 μ1 1399 016 1368 1399 1429 μ2 1491 008 1476 1491 1507 μ3 1449 014 1421 1449 1477

Since in each case the three data set types represent independent measurements the three estimates can be combined to obtain a consensus value In this way a between data set type variability is included in the final standard uncertainty Table C-6 lists the results in Copies per Nanoliter obtained using the NIST Consensus Builder (NICOB)C3

Table C-6 Certified Values for Components A B and C

Component Mean u(Mean) Uk=2(Mean) A 151 07 14 B 175 03 06 C 145 05 10

4 Mito-to-nuclear DNA ratios (mtDNAnDNA)

41 Data Measurements were also obtained on the same samples using three mitochondrial detection assays mtND1 mtBatz and mtKav and a single nDNA assay (genomic) to compute the ratios mtDNAnDNA For the complete mtDNAnDNA data see

Table 16 mtDNAnDNA Ratios

42 Model For each component the data was analyzed using the following random effects model to account for between assay variability

119903119903119894119894~119873119873120572120572119894119894 1205901205901198861198861198861198861198861198861198861198861198861198862

where 119894119894 denotes assay (for A and B 1 = mtND1 2 = mtBatz 3 = mtKav for C 1 = mtBatz 2 = mtKav) and 120572120572119894119894~119873119873micro120590120590119887119887119887119887119905119905119908119908119887119887119887119887119894119894 119886119886119886119886119886119886119886119886119886119886

2

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

65

43 OpenBUGS code inits and data Two versions of the same model are required one for the results from the three mtDNA assays available for components A and B mtND1 mtBatz and mtKav and the other for the results from the two mtDNA assays available for component C mtBatz and mtKav mtDNAnDNA code for components A and B ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 13)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for components A and B list(sigt=c(111)) Data for Component A list(Ndata=102) set[] rep[] asy[] rat[] 1 1 1 1931 1 2 1 187 1 1 2 1845 1 2 2 1764 2 1 1 181 2 2 1 1906 2 1 2 1785 2 2 2 1837 3 1 1 1743 3 2 1 1761 3 1 2 1829 3 2 2 1847 4 1 1 1716 4 2 1 180 4 1 2 174 4 2 2 1813 5 1 1 1923 5 2 1 1835 5 1 2 1935 5 2 2 1851 6 1 1 1757 6 2 1 1724 6 3 1 1768 6 1 2 1743 6 2 2 1671 6 3 2 1766 6 1 3 1692 6 2 3 1662 6 3 3 1767 7 1 1 1924 7 2 1 1944 7 3 1 1893 7 1 2 1786

7 2 2 1752 7 3 2 1743 7 1 3 153 7 2 3 1542 7 3 3 1523 7 1 2 158 7 2 2 1557 7 3 2 1592 7 1 3 2103 7 2 3 210 7 3 3 2049 7 1 1 1749 7 2 1 1769 7 3 1 1777 7 1 2 1965 7 2 2 1889 7 3 2 1947 7 1 3 1813 7 2 3 1864 7 3 3 1864 8 1 1 178 8 2 1 1776 8 3 1 1706 8 1 2 1804 8 2 2 1751 8 3 2 1738 8 1 3 1769 8 2 3 1786 8 3 3 1761 9 1 2 1584 9 2 2 1544 10 1 1 1947 10 2 1 1928 10 3 1 1922 10 1 1 1829

10 2 1 1895 10 3 1 1851 10 1 1 1674 10 2 1 1617 11 1 2 1667 11 2 2 1813 11 3 2 1708 11 4 2 1598 11 1 2 1566 11 2 2 1687 11 3 2 1663 11 4 2 1589 11 1 2 1643 11 2 2 1668 11 3 2 1604 11 4 2 1573 12 1 2 1719 12 2 2 1716 12 3 2 167 12 4 2 1666 12 1 2 1581 12 2 2 1573 12 3 2 1601 12 4 2 1612 12 1 2 1546 12 2 2 1542 12 3 2 1538 12 4 2 1587 13 1 1 1617 13 2 1 1599 13 1 1 1622 13 2 1 1579 13 1 1 1575 13 2 1 1528 END

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

66

Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A

mean sd MC_error val25pc median val975pc start sample mu 1730 1443 004419 1703 1730 1759 100001 100000

data for Component B list(Ndata=87) set[] rep[] asy[] rat[] 1 1 1 2461 1 2 1 2401 1 1 2 2555 1 2 2 2608 2 1 1 2229 2 2 1 2200 2 1 2 2108 2 2 2 2192 3 1 1 2127 3 2 1 2198 3 1 2 2455 3 2 2 2506 4 1 1 2211 4 2 1 2199 4 1 2 2380 4 2 2 2249 5 1 1 2105 5 2 1 1943 5 1 2 2136 5 2 2 2119 6 1 1 1990 6 2 1 1921 6 3 1 1982 6 1 2 1990 6 2 2 1924 6 3 2 1958 6 1 3 1903 6 2 3 1831

6 3 3 1919 7 1 1 2304 7 2 1 2341 7 3 1 2332 7 1 2 2048 7 2 2 2078 7 3 2 2067 7 1 3 2399 7 2 3 2479 7 3 3 2493 7 1 1 2065 7 2 1 2166 7 3 1 2164 7 1 2 2314 7 2 2 2280 7 3 2 2258 7 1 3 1867 7 2 3 1987 7 3 3 1971 8 1 1 2046 8 2 1 2057 8 3 1 2024 8 1 2 1998 8 2 2 2043 8 3 2 1952 8 1 3 2077 8 2 3 2117 8 3 3 2041 9 1 2 1890 9 2 2 1886

10 1 2 1928 10 2 2 1916 10 3 2 1915 10 1 2 1947 10 2 2 1946 10 3 2 1961 10 4 2 1909 10 1 2 1953 10 2 2 1952 10 3 2 1941 10 4 2 1881 11 1 2 1865 11 2 2 1863 11 3 2 1828 11 4 2 1828 11 1 2 1944 11 2 2 1964 11 3 2 1947 11 4 2 1895 11 1 2 1908 11 2 2 1909 11 3 2 1940 11 4 2 1839 12 1 1 1688 12 2 1 1783 12 1 1 1823 12 2 1 1631 12 1 1 1920 12 2 1 1887 END

Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B

mean sd MC_error val25pc median val975pc start sample mu 2045 2246 0111 1999 2045 2088 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

67

mtDNAnDNA code for component C ModelBegin mu ~ dnorm(0 00016)I(0001) chSqNs ~ dgamma(0505) sigr lt- musqrt(chSqNs) for(i in 12)alpha[i]~dnorm(musigr) sigt[i]~dgamma(10E-510E-5) for(i in 1Ndata)rat[i]~dnorm(alpha[asy[i]]sigt[asy[i]]) ModelEnd mtDNAnDNA inits for component C list(sigt=c(11)) mtDNAnDNA data for component C list(Ndata=66) set[] rep[] asy[] rat[] 1 1 1 3201 1 2 1 3299 2 1 1 3347 2 2 1 3287 3 1 1 3217 3 2 1 3384 4 1 1 3200 4 2 1 3363 5 1 1 3294 5 2 1 3013 6 1 1 2805 6 2 1 2825 6 3 1 2829 6 1 2 2709 6 2 2 2738 6 3 2 2727 7 1 1 2797 7 2 1 2642 7 3 1 2792 7 1 2 2804 7 2 2 2789

7 3 2 2842 7 1 1 2715 7 2 1 2606 7 3 1 2687 7 1 2 2799 7 2 2 2628 7 3 2 2559 7 1 1 2639 7 2 1 2535 7 3 1 2657 7 1 2 2953 7 2 2 2757 7 3 2 2802 8 1 1 2927 8 2 1 2776 8 3 1 2762 8 1 2 3044 8 2 2 2830 8 3 2 2825 9 1 1 2501 9 2 1 2522 10 1 1 2832 10 2 1 2840

10 3 1 2788 10 4 1 2725 10 1 1 2835 10 2 1 2790 10 3 1 2736 10 4 1 2679 10 1 1 2803 10 2 1 2773 10 3 1 2757 10 4 1 2752 11 1 1 2599 11 2 1 2508 11 3 1 2551 11 4 1 2609 11 1 1 2617 11 2 1 2483 11 3 1 2626 11 4 1 2602 11 1 1 2665 11 2 1 2535 11 3 1 2570 11 4 1 2650 END

Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C

mean sd MC_error val25pc median val975pc start sample mu 2774 2367 01176 2729 2774 2821 100001 100000

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

68

Table C-10 Results for mtDNAnDNA Posterior Distribution Summary Statistics

Component Mean Std Dev 25 50 975 A 1730 14 1703 1730 1758 B 2045 23 1994 2047 2087 C 2774 23 2729 2774 2821

5 ReferencesC1 Lunn DJ Spiegelhalter D Thomas A Best N Statistics in Medicine 2009283049ndash3082C2 Polson NG Scott JG Bayesian Analysis 20127887ndash902C3 NIST Consensus Builder Software and userrsquos guide freely available at

httpsconsensusnistgov

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

69

Appendix D Performance in Commercial qPCR Chemistries The performance of the SRM 2372a materials in commercial qPCR chemistries was assessed using manufacturerrsquos recommended protocols Each panel in Figures D-1 to D-4 displays the average cycle threshold (Ct) as a function of DNA concentration for the 5-point dilutions of SRM 2372a components In all cases the external calibrant was SRM 2372 component A Error bars represent approximate 95 confidence intervals Table D-1 lists the estimated [DNA] for the three SRM 2372a components for all commercial assays evaluated This Table also lists the Component BComponent A (BA) and Component CComponent A (CA) ratios calculated from the [DNA] estimates

Figure D-1 Results for the Innogenomics Commercial qPCR assays

20

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantLong(IQ_h)

ABCStd

[A] = (384 plusmn 26) ngμL[B] = (441 plusmn 24) ngμL[C] = (516 plusmn 24) ngμL

Rsup2 = 0999 se(Ct) = 01016

17

18

19

20

21

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantShort(IQ_d)

ABCStd

[A] = (435 plusmn 28) ngμL[B] = (543 plusmn 12) ngμL[C] = (510 plusmn 28) ngμL

Rsup2 = 0999 se(Ct) = 007

24

25

26

27

28

29

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYLong(IQHY_h)

ABCStd

[A] = (387 plusmn 24) ngμL[B] = (448 plusmn 21) ngμL[C] = (559 plusmn 31) ngμL

Rsup2 = 0998 se(Ct) = 01314

16

18

20

22

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYShort(IQHY_d)

ABCStd

[A] = (353 plusmn 23) ngμL[B] = (267 plusmn 15) ngμL[C] = (390 plusmn 18) ngμL

Rsup2 = 1000 se(Ct) = 00420

22

24

26

28

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

InnoQuantHYY(IQHY_y)

ABCStd

[A] = (572 plusmn 13) ngμL[B] = 0 ngμL[C] = (137 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

70

Figure D-2 Results for the Thermo Fisher Commercial qPCR assays

21

22

23

24

25

26

27

28

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HumanHuman(QFH_h)

ABCStd

[A] = (464 plusmn 13) ngμL[B] = (529 plusmn 19) ngμL[C] = (472 plusmn 22) ngμL

Rsup2 = 1000 se(Ct) = 004

22

23

24

25

26

27

28

29

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoHuman(QFD_h)

ABCStd

[A] = (701 plusmn 53) ngμL[B] = (561 plusmn 11) ngμL[C] = (472 plusmn 11) ngμL

Rsup2 = 0994 se(Ct) = 01922

24

26

28

30

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler DuoMale(QFD_y)

ABCStd

[A] = (706 plusmn 49) ngμL[B] = 0 ngμL[C] = (122 plusmn 02) ngμL

Rsup2 = 0999 se(Ct) = 007

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPLarge(QFHP_h)

ABCStd

[A] = (340 plusmn 18) ngμL[B] = (432 plusmn 05) ngμL[C] = (342 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 00720

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler HPSmall(QFHP_d)

ABCStd

[A] = (434 plusmn 15) ngμL[B] = (514 plusmn 10) ngμL[C] = (466 plusmn 07) ngμL

Rsup2 = 1000 se(Ct) = 005

18

19

20

21

22

23

24

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioLarge(QFT_h)

ABCStd

[A] = (362 plusmn 23) ngμL[B] = (430 plusmn 06) ngμL[C] = (378 plusmn 07) ngμL

Rsup2 = 0999 se(Ct) = 00920

21

22

23

24

25

26

27

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioSmall(QFT_d)

ABCStd

[A] = (468 plusmn 08) ngμL[B] = (505 plusmn 08) ngμL[C] = (523 plusmn 14) ngμL

Rsup2 = 1000 se(Ct) = 00318

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantifiler TrioY(QFT_y)

ABCStd

[A] = (372 plusmn 11) ngμL[B] = 0 ngμL[C] = (110 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 003

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

71

Figure D-3 Results for the Qiagen Commercial qPCR assays

Figure D-4 Results for the Promega Commercial qPCR assays

19

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresAutosomal(QPH_h)

ABCStd

[A] = (554 plusmn 21) ngμL[B] = (479 plusmn 05) ngμL[C] = (381 plusmn 09) ngμL

Rsup2 = 1000 se(Ct) = 00418

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex HyresY(QPH_y)

ABCStd

[A] = (358 plusmn 10) ngμL[B] = 0 ngμL[C] = (118 plusmn 03) ngμL

Rsup2 = 0999 se(Ct) = 007

16

17

18

19

20

21

22

23

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProHuman(QPP_h)

ABCStd

[A] = (525 plusmn 49) ngμL[B] = (480 plusmn 42) ngμL[C] = (467 plusmn 45) ngμL

Rsup2 = 0999 se(Ct) = 01019

20

21

22

23

24

25

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProDegradation(QPP_d)

ABCStd

[A] = (567 plusmn 18) ngμL[B] = (558 plusmn 08) ngμL[C] = (484 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 00410

15

20

25

30

35

-50 -40 -30 -20 -10 00 10

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

Quantiplex ProY(QPP_y)

ABCStd

[A] = (538 plusmn 27) ngμL[B] = (00 plusmn 00) ngμL[C] = (162 plusmn 11) ngμL

Rsup2 = 1000 se(Ct) = 003

16

18

20

22

24

-05 00 05 10 15 20

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorAutosomal(P_h)

ABCStd

[A] = (759 plusmn 64) ngμL[B] = (775 plusmn 24) ngμL[C] = (1056 plusmn 17) ngμL

Rsup2 = 0995 se(Ct) = 01818

20

22

24

26

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PlexorY(P_y)

ABCStd

[A] = (596 plusmn 51) ngμL[B] = 0 ngμL[C] = (117 plusmn 01) ngμL

Rsup2 = 0994 se(Ct) = 020

18

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantAutosomal(PQ_h)

ABCStd

[A] = (436 plusmn 13) ngμL[B] = (606 plusmn 20) ngμL[C] = (504 plusmn 10) ngμL

Rsup2 = 1000 se(Ct) = 00218

20

22

24

26

-05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantDegradation(PQ_d)

ABCStd

[A] = (420 plusmn 18) ngμL[B] = (529 plusmn 16) ngμL[C] = (530 plusmn 06) ngμL

Rsup2 = 1000 se(Ct) = 00516

18

20

22

-10 -05 00 05 10 15

Cros

sing

Thr

esho

ld C

ycle

s

Log₁₀([DNA])

PowerQuantY(PQ_y)

ABCStd

[A] = (410 plusmn 20) ngμL[B] = 0 ngμL[C] = (139 plusmn 05) ngμL

Rsup2 = 0999 se(Ct) = 006

_____________________________________________________________________________________________________ This publication is available free of charge from

httpsdoiorg106028NIS

TSP

260-189

72

Table D-1 Summary [DNA] Results for the Commercial qPCR Assays Estimated [DNA] ngμL Ratios

A B C BA CA Assaya xb u(x)c xb u(x)c xb u(x)c yd u(y)e yd u(y)e

IQd 435 28 543 12 510 28 125 008 117 010 IQh 384 26 441 24 516 24 115 010 134 011

IQHYd 353 23 267 15 390 18 076 006 111 009 IQHYh 387 24 448 21 559 31 116 009 144 012

Ph 759 64 775 24 1056 16 102 009 139 012 PQd 420 18 529 16 530 06 126 007 126 006 PQh 436 13 606 20 504 10 139 006 116 004

QFDh 701 53 561 11 472 11 080 006 067 005 QFHh 464 13 529 19 472 22 114 005 102 006

QFHPd 434 15 514 10 466 07 118 005 107 004 QFHPh 340 18 432 05 342 03 127 007 101 005

QFTd 468 08 505 08 523 14 108 003 112 004 QFTh 362 23 430 06 378 07 118 008 104 007 QPHh 554 21 479 05 380 09 086 003 069 003 QPPd 567 18 558 08 484 11 098 003 085 003 QPPh 525 49 480 42 467 45 092 012 089 012

IQHYy 572 13 NDf 137 06

NRg

024 001 Py 596 51 NDf 117 01 020 002

PQy 410 20 NDf 139 05 034 002 QFDy 706 49 NDf 122 02 017 001 QFTy 372 11 NDf 110 06 030 002 QPHy 358 10 NDf 118 03 033 001 QPPy 538 27 0002 0001 162 11 030 003

a Assay codes are listed in Table 8 b Mean of dilution-adjusted results from five dilutions of the SRM 2372a component c Standard uncertainty of the mean estimated from the stdanrd deviation of the five dilutions d Ratio of the estimated [DNA] of component B or C relative to that of component A e Standard uncertainty of the ratio estimated from the standard uncertainties of the individual components f Not detected Component B was prepared using tissue from only female donors it is not expected to

respond to male-specific assays g Not relevant

  • Abstract
  • Keywords
  • Technical Information Contact for this SRM
  • Table of Contents
  • Table of Tables
  • Table of Figures
  • Purpose and Description
  • Warning SRM 2372a is a Human Source Material
  • Storage and Use
  • History and Background
    • Figure 1 Sales History of SRM 2372
      • Experimental Methods
        • Polymerase Chain Reaction Assays
          • Table 1 NIST-Developed Human Nuclear DNA Assays
          • Figure 2 Exemplar Droplet Patterns for the Ten nDNA Assays
          • Table 2 NIST-Optimized Human Mitochondrial DNA Assays
          • Figure 3 Exemplar Droplet Patterns for the Three mtDNA Assays
            • Droplet Digital Polymerase Chain Reaction Measurements
              • Nuclear DNA (nDNA)
                • Table 3 Mastermix Setup for Human Nuclear DNA ddPCR Assays
                  • Mitochondrial DNA (mtDNA)
                    • Table 4 Mastermix Setup for Human Mitochondrial DNA ddPCR Duplex Assays
                    • Table 5 Mastermix Setup for Human Mitochondrial DNA ddPCR Monoplex Assays
                        • Sample Preparation
                        • Absorbance Measurements
                        • SRM Unit Production
                        • Component Homogeneity
                          • Figure 4 Plate Layout Design for Homogeneity Measurements
                            • Component Stability
                              • Figure 5 Plate Layout Design for Stability Measurements
                                • Component Certification Measurements
                                  • Table 6 Tubes used for Certification Measurements
                                  • Figure 6 Basic Plate Layout Design for Certification Measurements
                                    • Mitochondrial DNA Measurements
                                      • Table 7 Run Parameters for Mitochondrial DNA
                                        • Performance in Commercial qPCR Chemistries
                                          • Table 8 Commercial qPCR Chemistries Tested with SRM 2372a
                                          • Figure 7 Plate Layout Design for Performance Assessments
                                              • Measurement Results
                                                • Absorbance Spectrophotometry
                                                  • Table 9 Absorbance at Selected Wavelengths
                                                  • Figure 8 Absorbance Spectra of SRM 2372a Components A B and C
                                                    • Droplet Volume
                                                    • Certification Measurements
                                                      • Homogeneity Results
                                                        • Table 10 Homogeneity Data as λacute Copy Number per Nanoliter of Component
                                                        • Table 11 Within- and Between-Tube Relative Standard Deviations
                                                          • Stability Results
                                                            • Figure 9 ddPCR Stability Measurements as Function of Time
                                                            • Table 12 Stability Data as λacute Copy Number per Nanoliter of Component
                                                              • Ten-Assay Certification Results
                                                                • Figure 10 Comparison of Assay Performance
                                                                • Table 13 Certification Data as λacute Copy Number per Nanoliter of Component
                                                                  • Estimation of Copy Number Concentration
                                                                    • Table 14 Measured Copy Number Concentrations
                                                                      • Conversion of Copy Number to Mass Concentration
                                                                        • Table 15 Estimated Mass Concentrations Nanograms DNA per Microliters Solution
                                                                            • Mitochondrial to Nuclear DNA Ratio (mtDNAnDNA)
                                                                              • Figure 11 mtDNAnDNA for Components A B and C
                                                                              • Table 16 mtDNAnDNA Ratios
                                                                              • Table 17 mtDNAnDNA Summary Results
                                                                                • Performance in Commercial qPCR Chemistries
                                                                                  • Figure 12 qPCR Autosomal Quantitation Performance Summary
                                                                                  • Figure 13 qPCR Y-ChromosomeAutosomal Ratio Performance Summary
                                                                                      • Qualification of ddPCR as Fit-For-Purpose
                                                                                        • Figure 14 Absorbance Spectra of SRM 2372 Components A and B
                                                                                        • Digital Polymerase Chain Reaction Assays
                                                                                        • Heat Treatments (Melting of the DNA into Single Strands)
                                                                                        • Comparison of Mass Concentration Estimates for SRM 2372 and SRM 2372a
                                                                                          • Table 18 Comparison of Mass Concentration Estimates for SRM 2372 and 2372a
                                                                                            • Quantitative Evaluation of ssDNA Content
                                                                                              • ddPCR Measurement of ssDNA Proportion
                                                                                              • Proof of Principle
                                                                                                • Table 19 Comparison of Snap-Cooling to Slow-Cooling
                                                                                                  • Time at Denaturation Temperature
                                                                                                    • Table 20 Change in λtλu with Increasing Time at Several Denaturation Temperatures
                                                                                                      • Temperature
                                                                                                        • Table 21 Change in λtλu with Denaturation Temperature for SRM 2372 and 2372a
                                                                                                          • DNA Concentration in the ddPCR Solution
                                                                                                            • Table 22 Change in λtλu with DNA Mass Concentration in Treated Solution
                                                                                                              • Equivalence of Assays
                                                                                                                • Table 23 Comparison of Assays
                                                                                                                    • ssDNA Proportion Assessment of the SRM 2372 and 2372a Materials
                                                                                                                      • Table 24 Evaluation of Limiting Ratio
                                                                                                                          • Conclusions and Recommendations
                                                                                                                            • Recommended Certification Values
                                                                                                                              • Table 25 Recommended Values and 95 Uncertainties for Certification
                                                                                                                                • Use of SRM 2372a for Value-Assigning In-House Reference Materials
                                                                                                                                  • Certificate of Analysis
                                                                                                                                  • References
                                                                                                                                  • Appendix A DNA Extraction Method
                                                                                                                                    • Procedure
                                                                                                                                    • Reagents
                                                                                                                                    • Reference
                                                                                                                                      • Appendix B Spectrophotometric Measurement Protocol
                                                                                                                                        • Spectrophotometer Calibration
                                                                                                                                          • Establish Baseline
                                                                                                                                          • Enable Verification of Wavelength Calibration
                                                                                                                                          • Enable Verification of Absorbance Calibration
                                                                                                                                            • For ldquonativerdquo dsDNA Samples
                                                                                                                                              • Verify Optical Balance of Cuvettes
                                                                                                                                                • For NaOH denatured ssDNA Samples
                                                                                                                                                  • Prepare a 04 molL NaOH Stock Solution
                                                                                                                                                  • Prepare the Sample Solutions
                                                                                                                                                  • Measure Sample Solutions
                                                                                                                                                    • After All Sample Measurements are Complete
                                                                                                                                                      • Enable Verification of Spectrophotometer Stability
                                                                                                                                                      • Clean Up
                                                                                                                                                        • Evaluation
                                                                                                                                                        • Reference
                                                                                                                                                          • Appendix C SRM 2372a Statisticianrsquos report
                                                                                                                                                            • 1 Introduction
                                                                                                                                                            • 11 OpenBUGS Analysis
                                                                                                                                                            • 12 OpenBUGS Exemplar Output
                                                                                                                                                            • 2 Homogeneity assessment
                                                                                                                                                            • 21 Data
                                                                                                                                                            • 22 Model
                                                                                                                                                            • 23 OpenBUGS code and data
                                                                                                                                                              • Table C-1 Example OpenBUGS Homogeneity Summary
                                                                                                                                                                • 24 Results
                                                                                                                                                                  • Figure C-1 Homogeneity Assessment
                                                                                                                                                                    • 3 Certification measurements
                                                                                                                                                                    • 31 Data
                                                                                                                                                                    • 32 Model
                                                                                                                                                                    • 33 OpenBUGS code inits and data
                                                                                                                                                                      • Table C-2 Example OpenBUGS Certification Summary for Component A
                                                                                                                                                                      • Table C-3 Example OpenBUGS Certification Summary for Component B
                                                                                                                                                                      • Table C-4 Example OpenBUGS Certification Summary for Component C
                                                                                                                                                                        • 34 Results
                                                                                                                                                                          • Table C-5 Parameter Values for Components A B and C
                                                                                                                                                                          • Table C-6 Certified Values for Components A B and C
                                                                                                                                                                            • 4 Mito-to-nuclear DNA ratios (mtDNAnDNA)
                                                                                                                                                                            • 41 Data
                                                                                                                                                                            • 42 Model
                                                                                                                                                                            • 43 OpenBUGS code inits and data
                                                                                                                                                                              • Table C-7 Example OpenBUGS mtDNAnDNA Summary for Component A
                                                                                                                                                                              • Table C-8 Example OpenBUGS mtDNAnDNA Summary for Component B
                                                                                                                                                                              • Table C-9 Example OpenBUGS mtDNAnDNA Summary for Component C
                                                                                                                                                                              • Table C-10 Results for mtDNAnDNA
                                                                                                                                                                                • 5 References
                                                                                                                                                                                  • Appendix D Performance in Commercial qPCR Chemistries
                                                                                                                                                                                    • Table D-1 Summary [DNA] Results for the Commercial qPCR Assays
Page 12: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 13: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 14: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 15: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 16: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 17: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 18: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 19: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 20: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 21: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 22: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 23: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 24: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 25: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 26: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 27: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 28: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 29: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 30: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 31: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 32: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 33: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 34: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 35: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 36: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 37: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 38: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 39: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 40: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 41: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 42: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 43: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 44: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 45: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 46: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 47: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 48: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 49: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 50: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 51: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 52: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 53: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 54: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 55: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 56: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 57: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 58: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 59: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 60: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 61: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 62: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 63: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 64: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 65: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 66: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 67: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 68: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 69: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 70: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 71: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 72: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 73: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 74: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 75: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 76: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 77: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 78: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 79: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 80: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 81: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .
Page 82: Certification of Standard Reference Material® 2372a Human DNA … · 2018. 3. 26. · NIST Special Publication 260-189 . Certification of . Standard Reference Material ® 2372a .