Advisory Committee Chair
Greg Cooper
Advisory Committee Members
Greg Barsh
Rick Myers
Nita Limdi
Devin Absher
Dan Bullard
Document Type
Thesis
Date of Award
2018
Degree Name by School
Master of Science (MS) College of Arts and Sciences
Abstract
From its foundations, modern genetic science has focused on understanding the causal connections between genotypes and phenotypes. The emergence of tools and techniques for microarray-based genomic science has enabled hundreds of genome-wide association studies (GWAS) to rapidly uncover thousands of previously unknown links between genetic variants and complex phenotypes, including human diseases (Li et al., 2011b). The output of genomic association research has yielded significant insights into basic biology and disease, including the identification of disease DNA variants that are possibly targetable by pharmaceuticals and gene therapeutics. Despite evident successes of the GWAS paradigm, opinion is divided on the overall value of GWAS research. Few complex phenotypes have been satisfactorily resolved through GWAS evidence alone, and the translation of association research into tangible clinical benefits has been limited. In this proposal we highlight two major problems preventing traditional GWAS studies from reaching their full promise. First, trait-associated variants often fall outside of coding gene regions, suggesting that their effects are mediated through gene regulation events rather than through causally simpler protein structural disruptions. Secondly, it is difficult to identify truly causal variants from pools of candidate variants. We believe that these two important problems can be addressed by the development of association studies that target gene expression information and that are conditioned by knowledge of functional noncoding DNA. Maps generated from gene expression association studies identify expression quantitative trait loci (eQTLs), DNA loci correlated with gene expression. We further believe, based on our own research and from outside studies, that genome-wide functional assays can be combined with eQTL maps to predict expression-causal variants and to simultaneously provide hints to the mechanisms by which these variants exhibit their effects on gene expression. Knowledge of eQTL causation will provide basic genomic knowledge and invaluable assistance to GWAS interpretation by providing plausible regulatory mechanisms for observed trait effects. We intend to substantially improve upon our current ability to predict causal expression-associated variants through the use of diverse functional annotation data and through the appropriate application of computational and statistical methods.
Recommended Citation
Weaver, Benjamin Todd, "The Computational Discovery Of Genetic Determinants Of Gene Regulatory Variation In Human Liver" (2018). All ETDs from UAB. 3291.
https://digitalcommons.library.uab.edu/etd-collection/3291