April 2008

Lihua Julie Zhu

Integrated Analysis Of ChIP-seq/chip using ChIPpeakAnno and GeneNetworkBuilder

Bioconductor Developer Meeting Zurich, Switzerland December 13th -14th 2012

•  Introduction of ChIP-seq and ChIP-chip analysis workflow

•  ChIPpeakAnno

•  GeneNetworkBuilder

•  Analysis of DAF-12 ChIP-chip and Expression Dataset

Outline

HIGH-THROUGHPUT IDENTIFICATION OF DNA BINDING SITES

•  ChIP-seq – ChIP followed by high-throughput

sequencing

•  ChIP-chip – ChIP followed by genome tiling array

analysis

ANALYSIS WORKFLOW

CHIPPEAKANNO

•  Batch annotate enriched peaks – ChIP-seq – ChIP-chip – PAS-seq (Poly(A) Site Sequencing) – Cap Analysis of Gene Expression (CAGE) – Any experiments resulting in a large number of

enriched genomic regions

CHIPPEAKANNO •  Find the nearest genes for each set of peaks and graph the distribution around features. •  Find all genes within a certain distance from the peaks •  Identify enriched Gene Ontology (GO) terms and pathways associated with adjacent genes

of the peaks. •  Label peaks with any annotation of interest

•  a dataset from the literature •  CpG island •  conserved element •  histone modification marks

•  Determine the significance of overlap and drawing Venn diagrams to visualize the extent of the overlap

•  binding sites among replicates •  binding sites among transcription factors within a complex •  binding sites among different experiments such as yours and the ones in literature

•  Retrieve genomic sequences flanking putative binding sites for motif discovery, cloning or PCR amplification

•  Find the peaks with bi-directional promoters with summary statistics •  Summarize motif occurrence in peaks

GENENETWORKBUILDER

DAF-12 EXAMPLE DATASET

•  ChIP-chip peaks were downloaded from GEO at http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE28350 (Hochbaum, Zhang et al. 2011, PLoS Genet 7(7): e1002179)

•  Expression Microarray results were downloaded from (Fisher and Lithgow 2006, Aging Cell 5(2): 127-138).

OVERLAP ANALYSIS AND DISTRIBUTION OF PEAKS AROUND TSS

Replicate1 Replicate2

Replicate3 156

932

12

323

4

148

1

1424

Distance To Nearest TSS

Freq

uenc

y

−10000 −5000 0 5000 10000

05

1015

2025

DISTRIBUTION OF DAF-12-BINDING SITES

downstream

includeFeature

insideoverlapEnd

overlapStart

upstream

Exon% Intron% 5UTR% 3UTR% Proximal Promoter% Immediate Downstream% Enhancer%

Chromosome Region

% b

indi

ng s

ites

010

2030

4050

6070

DAF-12 REGULATORY NETWORK

C02F5.5

mi r -46C33D9.9

C01B10.4

mi r -52

mi r -53

C17A2.4

mi r -228

F01D5.1

srd-16

mb f - 1

mi r -74

mi r -229Y37D8A.5

lsy-2

mi r -75

C50F4.9

mi r -73

hpd -1

F32A5.4

C08E8.3

K10C2.3

Y69E1A.5

mi r -76

htas-1

Y59E9AL.3

dod-24

R10H10.3

sss-1

mi r -240

K09H11.7

pos-1

mi r -43

F55G11.7T01D3.6

cdh-5

T05E12.6

mi r -788

F57G8.7

C33F10.1

F01D5.3

K03H1.12

scl-27

m i r - 2

mi r -242

bre-1

mi r -63nhr -86

hpo-28

Y106G6A.4

F35H8.4

msp-51

clec-66

mi r -72

F44A2.5

ZK1248.17

msp-76

mir-239.1

mi r -87

B0507.10

T11B7.5

pho-1

clec-52

f l i - 1

T05C3.6

nspd-10

Y54G11A.14F21C10.11

Y49G5A.1

F36D1.4

K10D3.6

C27F2.7

nspd-1

ceh-45

meg-2

col-172

z ip -2

tbc -7

mi r -230

wago-9col -19

m i r - 1clec-4

g rd -5

gst -11

dyc-1

nhr -20

T22B7.7

F35E12.5

mi r -243

mi r -44

m ig -5

mi r -80

nhr -42

h lh-30

mi r -40

C24B9.3

i lys-2

mi r -58Y44A6D.1

Y75B8A.23

gst -27

cyp-14A5

col-20

C53A5.2

mi r -34far -3

msp-38

Y67A6A.1

W10C8.5

nhr -10

a t f -6

W01B6.4

ces-2mi r -237

l i n -4

nh r -1

daf -12

grd-10

C12D5.4

F09F7.4

mi r -84

sre-13lnp -1

msp-56

mi r -261

pho-12 T01B11.2

C36C9.1

mi r -42

msp-142le t -7

mi r -795

mi r -37

ug t -6

F11G11.4

ssq-2msp-63

C27D6.3

mi r -51

clec-76

msp-50

fu t - 1

mi r -39

Y40C5A.1

mi r -355

pd i -2

mi r -38

daf -3

mi r -41

l in -13mi r -241

mi r -35

mi r -36

scd-1

C32H11.4

K12H4.7

SUMMARY

•  Analysis of the DAF-12 example dataset shows that enriched GO terms and interaction pathways are consistent with the known functions of DAF-12.

•  Network analysis, using GeneNetworkBuilder with ChIP data and expression data, generated a system-level view of the intertwined connections among the direct and indirect targets of DAF-12, which shows that DAF-12 is a master regulator.

ACKNOWLEDGEMENT •  Mark Robinson for the invitation

and hospitality

•  The Bioconductor package reviewers

–  Nishant Gopalak

–  Marc Carlson

–  Paul Shannon

•  Coauthors –  Jianhong Ou (Developer of

GeneNetworkBuilder), Claude Gazin, Nathan Lawson, Hervé Pagès, Simon Lin, David Lapointe, Michael Green

•  The users of the ChIPpeakAnno

•  The Bioconductor core team, esp.,

•  Patrick Aboyoun

•  Vincent Carey

•  Martin Morgan

•  Ivan Gregoretti

•  Amy Molesworth!•  Khademul Islam!•  Hua Li!•  Noah Dowell

•  Yin Wu

•  Zhiping Weng, Sara Evans , Alan Ritacco, Glenn Maston, Ping Wan, Ellen Kittler