Analysis of genome-wide genetic variants on liver specific gene expression and chromatin state

Date of Award

2017

Document Type

Restricted Thesis

Terms of Use

© 2017 Julian A. Segert. All rights reserved.

Degree Name

Bachelor of Arts

Department

Biology Department

First Advisor

Bradley Justin Davidson

Abstract

The past decade of genome-wide association studies (GWAS) has produced a wealth of data pertaining to complex diseases, but understanding the causality of identified variants has remained a challenge. One major problem is that the vast majority of candidate variants lie in noncoding regions while downstream analyses have primarily focused on protein coding genes, whose functional consequences upon mutation are often more predictable. Noncoding variants often operate by altering transcription factor binding sites. This can lead to downstream changes in chromatin state and gene expression. To assess global effects on expression, and chromatin state resulting from noncoding variation, we have generated RNA-seq and ChIP-seq data from genotyped, human transplant liver samples. In order to identify genetic variants associated with gene expression variation (eQTLs) on a genomic scale, we have built an analytical framework utilizing RASQUAL¹, which allows considerations of phased haplotypes. The RASQUAL method leverages differences in read depth and allelic composition for exceptional statistical power even with low sample sizes. Studying intermediate phenotypes proximal to well-understood biochemical mechanisms such as histone modification and mRNA expression allows for meaningful inferences into the effects of noncoding variants that can be tested experimentally via reporter assays. Here, we have succeeded in mapping numerous significant eQTLs and caQTLs in liver. Continuations of this work will include functional validation of these QTLs and colocalization with relevant GWAS variants.

This document is currently not available here.

Share

COinS