Use Case 13 - Genboree

Introduction to the Fourth Epigenome
Informatics Workshop
NIH Roadmap
Epigenomics Data Analysis and Coordination Center
May 17-18, 2012
Houston, Texas

Workshop Objective:
Catalyze Conversion of Epigenomic Profiling Data
into Biological Insights through Integrative Analysis
• Introduction to Epigenomics, Epigenome Informatics
• Methods (Assays, Data Processing)
• Standards (Metadata, Interoperability)
• Data Resources (Human Epigenome Atlas)
• Epigenomic Tools (Genboree Workbench Use Cases)
• Collaborative Opportunities / Networking / Exchange of Experience
2

Thursday, May 17th 2012
Session 1
9:00 – 9:45 am Introduction to the Workshop, Epigenome Analysis & Genboree –
Matt Roth
Session 2
9:45 – 10:45 am Use case preparation – Setting up projects, databases, groups,
accessing files, user privaleges, navigating Genboree, toolsets, submitting jobs,
etc. (BRL staff)
10:45 – 11:00 am Break
11:00 – 1:00 pm Hands-on Case Studies: Epigenomic variation between tissues
(Part 1 of 3) (BRL staff)
1:00 – 2:00 pm Lunch
3

Thursday, May 17 th , 2012
Session 3
2:00 – 3:30 pm Hands-on Case Studies: Epigenomic variation between
individuals (Part 2 of 3)
3:30 – 5:00 pm Hand-on Case Studies: Epigenomic variation in cancer
(Part 3 of 3)
6:00 pm Depart for dinner together (place TBD) or dinner on your own
Friday, May 18 th , 2012
8:30 – 9:00 am Continental breakfast (outside auditorium)
Session 4
9:00 – 9:15 am Review of Day 1 and preview of Day 2 – Matt Roth
4

Friday, May 18 th , 2012
9:15 – 10:00 am Quantitative profiling of histone modifications, peak calling and
segmentation of epigenomic signals, Chip-Seq, RNA-Seq –
Cristi Coarfa (EDACC)
10:00 – 12:00 pm Hands-on Case Study: Chip-Seq analysis (BRL staff)
12:00 – 1:00 pm Boxed lunch (outside auditorium)
Session 5
1:00 – 2:30 pm Hands-on Case Study: RNA-Seq analysis (BRL staff)
2:30 – 3:00 pm Hands-on Case Study: Virtual data integration using Genboree
(BRL staff)
3:00 – 4:30 pm Hands-on: semi-structured time for completing data
analysis/case studies, informal discussions, and feedback from
users on practical changes that can improve tools (all)
4:30 – 5:00 pm Open discussion and wrap-up
5

Workshop Participants:
Brief introductions to facilitate networking
Your goals in attending workshop
6

Bioinformatics Research Laboratory (BRL)
Govind Kunde, MS
Rob Waterland Lab
Viren Amin
Graduate Student
7

NIH Roadmap Epigenomics:
Reference Epigenomes
and the Human Epigenome Atlas
NIH Roadmap
Epigenomics Data Analysis and Coordination Center (EDACC)
May 17-18, 2012
Houston, Texas

NIH Roadmap Epigenomics Project
Hypothesis:
Origins of health and susceptibility to disease are, in part,
result of epigenetic regulation
Goal:
Transform biomedical research by
• Developing comprehensive reference epigenome maps
• Developing new technologies for epigenomic analyses
Including cyberinfrastructure for epigenomic research 9

NIH Roadmap Epigenomics
Project: Data Flow
Quarterly cumulative releases of the Human Epigenome Atlas
(Human Epigenome Atlas Release 6 completed)

Epigenomics Portal at NCBI

Human Epigenome Browser at Wash U

Epigenomic MetaData Standards
www.ihec-epigenomes.org

Epigenomic Metadata

(2010)
Alan Harris

DNA Methylation
• C5 position of cytosines primarily in CpGs, but also in non-
CpGs primarily in embryonic stem cells
• CpG Islands
• Regulation of cellular processes
– Transcription
– Defense against endogenous retroviruses
– Embryonic development
– X chromosome inactivation
– Imprinting

DNA Methylation Methods
• Bisulfite-based Methods
– MethylC-seq: whole genome shotgun bisulfite sequencing
– RRBS: Reduced Representation Bisulfite Sequencing
• MspI digestion
• Enrichment-based Methods
– MeDIP-seq: Methylated DNA Immunoprecipitation
• 5-methylcytosine antibody
– MBD-seq: Methyl-Binding Domain
• MBD2 protein methyl-CpG binding domain
– MRE-seq: Methylation-sensitive Restriction Enzyme
• Parallel HpaII, Hin6I, and AciI digestion

Chromatin ImmunopreciPitation
Sequencing – ChIP-Seq

Methylation Data Analysis Software
Software Features
BISMARK Supports both single end and pair-end reads.
Uses bowtie aligner.
PASH 3.0 Methylation & SNP’s.
Uses low memory & High speed alignment
BSMAP Maps both single/pair-end reads.
Uses SOAP aligner.
Methylcoder Maps both single/pair-end reads.
Handles also color space reads (SOLiD).
BS-Seq Uses Gaussian Mixture model (GMM) to identify the
probability of A vs G vs C vs T.
GMM available only to Arabidopsis genome
BRAT Maps both single/pair-end reads.
Trims low quality bases.
Improves unique mapping for pair-end reads.
Kismeth Web-based tool.
Designed for plant methylation data.

Genboree Workbench (cont’d)

Various Data Types
(tracks, files, ROIs, etc)
Tells the tool to use
this data/file
Tells the tool
where to
deposit results

Specific
information on
files/samples
selected in the
“Data Selector”

Epigenome Atlas Release 5
over 1500 experiments
www.epigenomeatlas.org
27

Epigenome Atlas Release 5
www.epigenomeatlas.org
28

Biology Across Key Genetic Elements (promoters, exons, UTR, etc):
Many ROI (i.e. annotation) Tracks Available
ROI tracks available
Promoter 5’ UTR Exon 1 Exon 2 3’ UTR
Introns

ROI track
Data Tracks from Epigenomic Experiments Projected
On To ROI (i.e. annotation) Tracks
Data track
LCP 1 LCP 2 LCP 3
Promoters: LCP = Low-CpG promoters (as defined in Weber et al., Nature Genetics (2007)

Compute Pearson Correlation Coefficient Between Experiments:
Similarity Matrix is Output as Heatmap
Cell line 1 Cell line 2
LCP 1 LCP 2 LCP 3 LCP 1 LCP 2 LCP 3
(2005) RDCT: R: A language and environment for
statistical computing. In. Vienna, Austria: R Foundation
for Statistical Computing; 2005.
Cell line 3 (not shown)
Cell line 2
Cell line 1
Cell line 1
Cell line 2
Cell line 3 (not shown)

Viewing selections
32

Including More Genes in the Same
Pathway
33

The data for several of the workshop Use
Cases was kindly provided by Dr. Jonathan
Mill (King’s College London, UK), and is
under review for this publication:
"Tissue-specific epigenetic variation across brain and blood:
functional annotation of the human brain methylome". Matthew
Davies 1 , Manuela Volta 1 , Abhishek Dixit 1 , Simon Lovestone 1 , Cristian
Coarfa 2 , R. Alan Harris 2 , Aleksandar Milosavljevic 2 , Claire Troakes 1 ,
Safa Al-Sarraj 1 , Richard Dobson 1 , Leonard C. Schalkwyk 1 , Jonathan
Mill 1 *
1 Institute of Psychiatry, King’s College London. UK. 2 Baylor College of Medicine,
Houston, Texas. USA. *Corresponding Author: Dr. Jonathan Mill, Address: Institute of
Psychiatry, SGDP Centre, De Crespigny Park, Denmark Hill, London.
Since the paper is under review (Davies et al), it can not be shared with anyone outside
of the workshop at this time, and we ask that you consider the data confidential. We will
notify you when the data can be shared. Thank you for your understanding.

Use Case 1: Genomewide Patterns of Methylation can Distinguish Between Blood, Cerebellum, and Cortex
Epigenomic Tracks:
-Blood
-Cerebellum
-Cortex
Use Case 2: Breast Cell Types Cluster Based on Their MeDIP-seq Profiles (Epigenome Atlas and UCSF REMC data)
Epigenomic Tracks:
-Breast Luminal Epithelium
-Breast Myoepithelial
-Breast Stem Cell
ROIs
ROIs
Data from Davies et al (in review)
36

Use Case 5: Methylation of some features discriminate tissue type better than others
Similar to Use Case 1 & 2 but uses different ROIs to illustrate how different features produce different
similarity matrices (heatmaps).
Epigenomic Tracks
-Blood
-Cerebellum
-Cortex
ROIs
Data from Davies et al (in review)
37

Use Case 9: Coordinated Changes of Epigenomic Marks Across Tissue Types
Epigenomic Tracks:
-H1 cell line
-IMR90 cell line
Collates score tracks into one data matrix, export to Excel
Column headers = experiments
Rows = ROIs
Data from Davies et al (in review)
Scatter plots
38

Use Case 12: Assess breast cancer cell type of origin
“Your” Epigenomic Tracks:
-Breast Luminal Epithelium
-Breast Myoepithelial
-Breast Stem Cell
Public Epigenomic Tracks:
-Breast Luminal Epithelium
-Breast Myoepithelial
-Breast Stem Cell
Data from Davies et al (in review)
LIMMA: Smyth, G. K. Statistical Applications in Genetics and Molecular Biology (2005)
39

Use Case 13: Analysis of epigenomic variation in breast tumors
Use Case 13a: Cluster all 16 breast tissue samples
16 450 K Samples (Dedeurwaerder, S.et al., 2011)
-8 normal breast samples
-8 cancerous breast samples
ROIs (HCP vs LCP)
Data from Davies et al (in review)
LCP
HCP
Normal
Tumor
Normal
Tumor
40

Use Case 13: Analysis of epigenomic variation in breast tumors
Use Case 13b: Compare 450K profiles (8 tumor, 8 normal) against
reference epigenomes from the Epigenome Atlas
16 450 K Samples (Dedeurwaerder, S.et al.,2011)
-8 normal breast samples
-8 cancerous breast samples
ROIs
Data from Davies et al (in review)
LCP
HCP
41

Use Case 13: Analysis of epigenomic variation in breast tumors
Use Case 13c: Since most breast tumor samples appear to contain excess of blood & immune
cells, comparison of normal and tumor tissue may reveal differentially methylated
genes (and corresponding pathways). Identify differentially methylated probes,
genes, and pathways using LIMMA & online resources
16 450 K Samples (Dedeurwaerder)
-8 normal breast samples
-8 cancerous breast samples
ROIs
Gene List
Data from Davies et al (in review)
42

BED
files
Use Case 14: Chip-Seq and RNA-Seq Data Analysis
MACS results (file in Genboree) Visualize in Genboree
H3K4me3, H3K4me1
Data from Davies et al (in review)
Zhang et al, Genome Biology (2008)
43

FASTQ,
BAM
files
Use Case 14: Chip-Seq and RNA-Seq Data Analysis
Gene expression diffs (file in Genboree)
Data from Davies et al (in review)
Trapnell et al, Bioinformatics (2009)
Trapnell et al, Nature Biotech (2010)
Trapnell et al, Nature Biotech (2010)
Visualization/pathway analysis
44

Workshop Evaluation (link)
Workshop
?
Epigenomic Profiling Biological Insights
Integrative Analysis
• Introduction to Workshop, Epigenome Informatics, Genboree
• Methods (Assays, Data Processing)
• Standards (Metadata, Interoperability)
• Data Resources (Human Epigenome Atlas)
• Tools (Epigenomic Toolset, Genboree Workbench)
• Use Cases / Case Studies
• Collaborative Opportunities / Networking / Exchange of Experience
45

Acknowledgments
• BRL
– Aleksandar Milosavljevic
– Cristian Coarfa, Alan R Harris
– Elke Norwig-Eastaugh
– Viren Amin
• BRL core
– Matt Roth, Kevin Riehle
• Genboree programming team
– Andrew Jackson, Sameer Paithankar, Sriram Raghuram
• Former BRL members
– Chia-Chin Wu, Arpit Tandon
• Govind Kunde (Waterland lab)
• Peak Calling
– Anshul Kundaje
– Bob Thurman (UW)
– Noam Shoresh (BI)
– Martin Hirst (BCGSC)
– Lee Daniels (NIH)
– Wei Li (BCM)
• RNA-Seq
– Adrian Lee, Joe Gray
– Laising Yen, Kalpana Kannan
– Sean McGuire, Christine Pichot