International consortia have generated reference epigenomes for over 100 cell types to better understand epigenetic variation. The NIH Roadmap Epigenomics Project profiled 127 tissues and cell types, building on data from ENCODE. These reference epigenomes are available through public databases like the Epigenome Atlas and WashU Epigenome Browser. Researchers can use the data to explore questions about gene regulation and annotation of genetic variants. The document demonstrates exploring Roadmap data to replicate findings on neural differentiation and epigenomic annotation of multiple sclerosis variants.

1. Using Public Databases to
Inform Research Questions
Day 1

2. Large Reference
Epigenome Projects
Understanding Natural Variation

3. Understanding
Epigenetic
Variation
• International consortia
have pursued the
establishment of reference
epigenomes for a large
number of cell types and
conditions
• Epigenome Roadmap
published in Nature
February 2015
• main findings of the NIH
Roadmap Epigenomics
Program

4. The Beginnings
• ENCODE (the Encyclopedia of DNA
Elements) started in 2003 to identify
functional elements in the non-
coding parts of the genome
• Understand how the genome is
packaged, regulated, and read
• In 2012 Nature, Genome
Research and Genome Biology
published 30 papers all together on
the results of the ENCODE project
• While ENCODE transformed
technology and data analysis, clinical
application was limited
• Most results were from a small
number of cell lines
ENCODE Data types:
• After the human genome had been mapped, it was clear the epigenome
needed to be explored as well
https://www.encodeproject.org/

5. ENCODE to Roadmap
• Funded by NIH, the Roadmap Epigenomic Project was established to generate
epigenomic data from the primary cells and tissues from both healthy individuals
and patients with diseases (e.g. cancer, neurodegenerative and autoimmune
disease)
• To the right are
the tissues and
cell types
profiled
• 127 human
tissues and cell
types
PMID: 25693563

6. Cell Types in Roadmap
• Many of the adult tissues investigated were broken down by cell type or region
• e.g. blood into several types of immune cell, and the brain into regions including the
hippocampus and dorsolateral prefrontal cortex
PMID: 25693562
profiled for histone modification
patterns, DNA accessibility, DNA
methylation and RNA expression

7. Data Sets
Available in
Roadmap
Many more cell/tissue groups…
To see full list:
http://www.nature.com
/nature/journal/v518/n
7539/fig_tab/nature142
48_F2.html
111 reference
epigenomes from
Roadmap with 16
additional epigenomes
from ENCODE

8. International Human Epigenome
Consortium (IHEC)
• Roadmap represents an early component of IHEC
• IHEC plans to determine the epigenomes of every cell type in the human body
— estimated to about a thousand
• IHEC brings together data from several consortiums, including:
• ENCODE
• Roadmap
• BLUEPRINT (http://www.blueprint-epigenome.eu/)- aim to decipher the
epigenomes of more than 100 different types of blood cells
• Canadian Epigenetics, Environment and Health Research Consortium (CEEHRC)

9. Using these Reference Data Sets
• What questions can these data sets answer?
• What questions can they not answer?
• How can we use this data to inform our
study questions?

10. Lets Explore the Data
• There are a couple of ways to view and download this
reference data
• I am going to provide a few examples
• Human Epigenome Atlas provides
interactive Visualization and
Download
• Not going to focus on download
because there are many ways to
view the data online
• http://www.genboree.org/epige
nomeatlas/index.rhtml
click

11. Epigenetic Data
Browsers
Getting an Idea of Content

12. UCSC Genome Browser
• Very user friendly, can search for a region or a gene
• Directions:
• Go to: http://genome.ucsc.edu/
• Click on “Genome Browser”
• Search a favorite gene, click
“submit”
• Things to try:
• zoom
• Adding tracks
• getting track info
click

13. Viewing the Roadmap Data
• From Genboree you can export the Roadmap data
to Genome Browser, but there are two browsers
from WashU that are great for view this data
• WashU Epigenome Browser
• Supports multiple organisms, visualizes chromatin-interaction
data (e.g. Hi-C), performs gene set view, gene plot, and many
other capabilities
• The Roadmap Epigenome Browser
• Powered by the WashU Epigenome Browser, but specific to the
Roadmap data

14. WashU Epigenome Browser
• http://epigenomegateway.wustl.edu/browser/
• Click on “Human h19”
• Then click on “Public Hubs”
• Then click “Reference human epigenomes from
Roadmap Epigenomics Consortium”
click

15. WashU Epigenome Browser
• Then click “Load” Roadmap Data from GEO
• Then click “Loaded” to select the samples to view
click

16. WashU Epigenome Browser
• Can select and view
all epigenetics marks
for one tissue
• Walk through
together
• Can view one type of
epigenetic mark for
multiple tissues
• Prefer to do this in
Roadmap Epigenome
Browser

17. Roadmap Epigenome Browser
• http://epigenomegate
way.wustl.edu/browse
r/roadmap/
• Click on “h19” and
then “Load”
• Can then select
epigenetic mark
• Can select region or
gene to interrogate
• Click submit
Select
Region
Select mark

18. Can see
how the
data
clusters

19. Exploring SNP data in Ensembl
• http://useast.ensembl.org/index.html
• Can also explore in dbSNP but I prefer this interface
• Put rs number in search – example: rs1801133

20. Exploring SNP data in Ensembl
• Can explore
population
genetics from
1000 genomes
project
• Can look at LD
• Must select a
reference
population

21. Exploring Other Public Data in
NCBI Epigenomics
• http://www.ncbi.nlm.nih.gov/epigenomics
• Can browse experiments or samples and then view results

22. Exploring Other Public Data in
NCBI Epigenomics
• Choose “Browse Experiments”
• Choose species, biological source and all features
• “Select all” Experiment IDs
• Click “View on Genome”

23. Exploring Other Public Data in
NCBI Epigenomics
• Can view in
UCSC Genome
Browser
• Click “View at
UCSC”

24. Exploring Other Public Data: GEO
GEO: Gene Expression Omnibus
• A public functional genomics data repository
• Both array and sequencing data stored
• Visit GEO DataSet Site:
http://www.ncbi.nlm.nih.gov/gds/?term=
• Can search a research question of interest
• Many ways to download data into R – can download
processed data or raw data
• Can download directly from GEO
• Can use R library GEOquery
• Both minfi and RnBeads have functions to download GEO data and
format for their specific purpose

25. Task: Replicating some
Roadmap Results
Practicing with these viewers and data sets

26. Brain Epigenomics
• Paper: Dissecting neural differentiation regulatory
networks through epigenetic footprinting
• One of the papers published in the February issue of Nature
• Going to replicate some of the results presented in
Supplementary Figure 1 and Figure 1

27. Brain Epigenomics:
Supp. Figure 1
• Identified 3,396 differentially expressed
genes between undifferentiated ES cells
and the first four neural progenitor cell
stages
• Pluripotency-associated genes such
as OCT4 and NANOG are downregulated
• We’re going to look at H9 cells, H9
derived neuronal progenitor cells, and
H9 derived neuron culture cells
• Data not available yet for all the cell types
presented in paper

28. Brain Epigenomics: Supp. Figure 1
• Go to: http://epigenomegateway.wustl.edu/browser/
• Going to load expression data for these 3 cell types
• Go to Human hg19
• Go to Public Hubs
• Click on “Roadmap Epigenomics Interactive Analysis Hub”
• Load “Complete Consolidated Dataset”
• Load Expression Data (under ES/iPS cells) for:
• H9 cells
• H9 derived neuronal progenitor cells
• H9 derived neuron culture cells
• Explore some of the genes in Supplementary Figure 1 (e.g.
NANOG)

29. Brain Epigenomics: Supp. Figure 1
• Look at some of the genes in
the figure all together by
creating a gene set
• Click “Apps”
• Select Gene and region set
• Upload the list of genes in the
comments of slide
• Can try “Gene set view” to see
all the genes at once
• Under apps look at “Scatter
plot” to compare to plots

30. Brain Epigenomics: Figure 1
Going to replicate part of this figure

31. Brain Epigenomics: Figure 1
• From the previous exercise we can see the
downregulation of NANOG in the derived cells
• Now can add in additional epigenetic marks
• Do they seem to be associated with regulation?
• Can also try to recreate the Figure 1 for SOX2
• Use H9 derived neuronal progenitor cell

32. Exploring Epigenomic Annotation
of Genetic Variants
• Another paper published in February Issue
• Epigenomic annotation of genetic variants using the Roadmap
Epigenome Browser
• Going to look at multiple sclerosis–associated SNPs
identified that are annotated using epigenomic and
expression data from 31 primary human tissues
(orange) and cells (light green).
• The region associated with rs756699 has H3K4me1 mostly
confined to immune-related cell types (solid black box).
• The closest gene, TCF7 (3.8 kb downstream), also shows high
expression in the same group of cell types
• The region surrounding rs307896 has H3K4me1 signal in all
tissues and cell types (dashed black box).
• SNP rs307896 lies in an intron of SAE1, a gene that is also expressed
in all the samples

33. Exploring Epigenomic Annotation
of Genetic Variants

34. Exploring Epigenomic Annotation
of Genetic Variants
• http://epigenomegateway.wustl.edu/browser/roadmap/
• Can also do this in WashU Epigenome Browser, but this has a
nice clustering capability
• Add H3K9me1 and RNA-seq (have to use search)
• Load data
• Clink hg19 to select the samples – select primary cells and
tissues
• In window click “chr” to select the gene TCF7, zoom out
• Click the track header (e.g. “H3K4me1”) to add
“Annotation track”
• Go to population variation
• Add dbSNP