Introduction Biological Background Method Results Moving Forward
Modeling gene expression using five
histone modifications
Fifth Annual Primes MIT Conference
Lalita Devadas
Mentor: Angela Yen
May 17, 2015
Lalita Devadas — Modeling gene expression using five histone modifications
1/31
Introduction Biological Background Method Results Moving Forward
Outline
1 Biological Background
2 Method
3 Results
4 Moving Forward
Lalita Devadas — Modeling gene expression using five histone modifications
2/31
Introduction Biological Background Method Results Moving Forward
Outline
1 Biological Background
2 Method
3 Results
4 Moving Forward
Lalita Devadas — Modeling gene expression using five histone modifications
3/31
Introduction Biological Background Method Results Moving Forward
Gene Expression
Central dogma of molecular biology
Lalita Devadas — Modeling gene expression using five histone modifications
4/31
Introduction Biological Background Method Results Moving Forward
Gene Expression
Relevance
Important to understanding biological activity
Crucial to advances in medicine
Detection, prevention, and treatment of disease
Lalita Devadas — Modeling gene expression using five histone modifications
5/31
Introduction Biological Background Method Results Moving Forward
Gene Expression
Regulation
Genetic
Sequences of nucleotides (ACTG)
Lalita Devadas — Modeling gene expression using five histone modifications
6/31
Introduction Biological Background Method Results Moving Forward
Gene Expression
Regulation
Epigenetic
Changes to environment surrounding DNA
Lalita Devadas — Modeling gene expression using five histone modifications
7/31
Introduction Biological Background Method Results Moving Forward
Epigenetics
Histone modifications
Chemical changes to histone protein core or protruding
tail
Lalita Devadas — Modeling gene expression using five histone modifications
8/31
Introduction Biological Background Method Results Moving Forward
Epigenomes
Roadmap Project
Lalita Devadas — Modeling gene expression using five histone modifications
9/31
Introduction Biological Background Method Results Moving Forward
Outline
1 Biological Background
2 Method
3 Results
4 Moving Forward
Lalita Devadas — Modeling gene expression using five histone modifications
10/31
Introduction Biological Background Method Results Moving Forward
Data Pipeline
Objective
Density of Histone
Modifications
Prediction
Lalita Devadas — Modeling gene expression using five histone modifications
Level of Gene
Expression
11/31
Introduction Biological Background Method Results Moving Forward
Data Pipeline
Overview
All Histone Data
Best-binning
Relevant Histone Data
Classification
“Off” genes (Unexpressed)
Predict at 0
Classification
“On” genes (Expressed)
Regression
Predicted Expression
Lalita Devadas — Modeling gene expression using five histone modifications
12/31
Introduction Biological Background Method Results Moving Forward
Data Pipeline
Best-bin approach
All Histone Data
Best-binning
Relevant Histone Data
Classification
“Off” genes (Unexpressed)
Predict at 0
Classification
“On” genes (Expressed)
Regression
Predicted Expression
Lalita Devadas — Modeling gene expression using five histone modifications
13/31
Introduction Biological Background Method Results Moving Forward
Best-bin approach
Dividing genes
Lalita Devadas — Modeling gene expression using five histone modifications
14/31
Introduction Biological Background Method Results Moving Forward
Best-bin approach
Choosing best bin
epigenome X, histone mark Y
bin 1
bin 2
bin 3
bin 4
.
.
bin p
.
.
.
bin 81
expression
gene a
gene b
gene c
.
.
.
.
.
p = best bin
Lalita Devadas — Modeling gene expression using five histone modifications
strongest correlation
15/31
Introduction Biological Background Method Results Moving Forward
Data Pipeline
Classification
All Histone Data
Best-binning
Relevant Histone Data
Classification
“Off” genes (Unexpressed)
Predict at 0
Classification
“On” genes (Expressed)
Regression
Predicted Expression
Lalita Devadas — Modeling gene expression using five histone modifications
16/31
Introduction Biological Background Method Results Moving Forward
Types of Models
Random Forest
Random Forest model
Returns majority vote of classification determined by a group
of decision trees
Lalita Devadas — Modeling gene expression using five histone modifications
17/31
Introduction Biological Background Method Results Moving Forward
Types of Models
Random Forest
Random Forest model
Returns majority vote of classification determined by a group
of decision trees
Lalita Devadas — Modeling gene expression using five histone modifications
18/31
Introduction Biological Background Method Results Moving Forward
Data Pipeline
Regression
All Histone Data
Best-binning
Relevant Histone Data
Classification
“Off” genes (Unexpressed)
Predict at 0
Classification
“On” genes (Expressed)
Regression
Predicted Expression
Lalita Devadas — Modeling gene expression using five histone modifications
19/31
Introduction Biological Background Method Results Moving Forward
Types of Models
Linear Model
Linear model
Finds a linear correlation between predictors and response
Lalita Devadas — Modeling gene expression using five histone modifications
20/31
Introduction Biological Background Method Results Moving Forward
Data Pipeline
Overview
All Histone Data
Best-binning
Relevant Histone Data
Classification
“Off” genes (Unexpressed)
Predict at 0
Classification
“On” genes (Expressed)
Regression
Predicted Expression
Lalita Devadas — Modeling gene expression using five histone modifications
21/31
Introduction Biological Background Method Results Moving Forward
Outline
1 Biological Background
2 Method
3 Results
4 Moving Forward
Lalita Devadas — Modeling gene expression using five histone modifications
22/31
Introduction Biological Background Method Results Moving Forward
Epigenomes
Roadmap Project
Lalita Devadas — Modeling gene expression using five histone modifications
23/31
Introduction Biological Background Method Results Moving Forward
Data Pipeline
Objective
Density of Histone
Modifications
Prediction
Level of Gene
Expression
H3K9me3
H3K4me3
H3K4me1
H3K36me3
H3K27me3
Lalita Devadas — Modeling gene expression using five histone modifications
24/31
Introduction Biological Background Method Results Moving Forward
Results of Pipeline
Conclusions
Models created for cultured epigenomes have a much
higher predictive power than those created for tissue
samples
H3K36me3 is the most important histone mark used for
prediction
Lalita Devadas — Modeling gene expression using five histone modifications
25/31
Introduction Biological Background Method Results Moving Forward
Results of Pipeline
Graph
Lalita Devadas — Modeling gene expression using five histone modifications
26/31
Introduction Biological Background Method Results Moving Forward
Specifics of Best Model
Classification Accuracy
Lalita Devadas — Modeling gene expression using five histone modifications
27/31
Introduction Biological Background Method Results Moving Forward
Specifics of Best Model
Regression Accuracy
every data point
represents one gene
r2 value: 0.640
Lalita Devadas — Modeling gene expression using five histone modifications
28/31
Introduction Biological Background Method Results Moving Forward
Outline
1 Biological Background
2 Method
3 Results
4 Moving Forward
Lalita Devadas — Modeling gene expression using five histone modifications
29/31
Introduction Biological Background Method Results Moving Forward
Next Steps
Improve predictive power
Broaden scope of predictors and response
Further analysis of current results
Apply procedure to different data
Release code as a tool for other researchers
Lalita Devadas — Modeling gene expression using five histone modifications
30/31
Introduction Biological Background Method Results Moving Forward
Acknowledgements
I would like to thank:
My mentor, Angela Yen
Prof. Manolis Kellis
Roadmap Project
PRIMES program
My family
Lalita Devadas — Modeling gene expression using five histone modifications
31/31