Open Collections

Casazza, William

University of British Columbia

The majority of genetic loci that influence complex traits are non-coding. Roughly half of these loci affect gene expression in one or more tissues, despite making up less than 10% of common genetic variants. However, it is increasingly unlikely that genetic effects on gene expression will be sufficient to assign a function to all non-coding risk loci. This thesis splits the problem of assigning new functions to complex trait loci into two steps. First, I aim to increase the number of non-coding loci with a known function without requiring new molecular datasets. In my thesis, I explore context-dependent genetic regulation of molecular traits, where the context that affects this process is either inferred from data or a common phenotypic measure like sex. I then advocate for associating genetic variation with molecular data other than gene expression, primarily on DNA methylation. Second, I outline how to tie these novel molecular functions to genetic risk for complex traits. Importantly, this requires methods and workflows that summarize genetic effects at individual loci across interpretable functional units (genes) and genome-wide genetic risk for complex traits. In my scholarship chapters, I first show that environments inferred from global gene expression can correlate with various phenotypic and environmental variables. These inferred contexts are replicated across samples and can subsequently be used to identify novel context-specific genetic regulation of gene expression. I then show that novel context-specific genetic regulation can be approached in DNA methylation using sex, measured in virtually all genetic and molecular datasets. My later chapters demonstrate how to summarize genetic effects to learn which traits are particularly relevant to these novel regulatory relationships. I start with effects and individual loci and then explore methods to interpret the gene-level influence of genetic effects on DNA methylation. Then I demonstrate how cumulative, genome-wide risk for complex traits can provide new insight into the biological functions underlying complex traits beyond the effects I observe at individual loci. Overall this thesis shows that we require various approaches to discover disease-relevant molecular functions of non-coding genetic loci.

Text; Dataset

2023-08-15

Attribution-NonCommercial-ShareAlike 4.0 International

http://hdl.handle.net/2429/85485

Bioinformatics

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-sa/4.0/