Deciphering genetic associations using genome-wide epigenomics approaches

Abstract

Genetic mapping of the drivers of complex human phenotypes and disease through the genome-wide association study (GWAS) has identified thousands of causal genetic loci in the human population. However, genetic mapping approaches can often only reveal a particular causal locus, not the molecular mechanism through which it acts. Biological interpretation of these genetic results is thus a bottleneck for turning results from GWAS into meaningful biological insights for human biology. Genetic mapping of complex human traits has revealed that most common variants influencing human phenotypes have weak effect sizes and reside outside protein-coding regions, complicating biological interpretation of their function. In this thesis we use computational and experimental approaches to study the non-coding genome. In particular, we focus on using epigenomic signatures to characterize non-coding transcriptional regulatory elements and predict regulatory segments of DNA disrupted by genetic variants. In Chapter 2, we describe how genome-wide maps of epigenomic modifications can be used to characterize and discover new GWAS loci. In Chapter 3, we outline an experimental method for the high-throughput assessment of putative transcriptional regulatory elements. In summary, our research highlights the value of interpreting human genetics information through an epigenomic lens, and provides a glimpse into the possible biological insights that manifest from the intersection of these two areas of research.