The effects of adipose tissue gene expression, isoform usage, and cell type composition on cardiometabolic traits

Cardiometabolic traits such as type 2 diabetes and obesity are major health burdens worldwide and understanding the genes and mechanisms that influence these cardiometabolic traits would help develop future treatments for disease. Genome-wide association studies (GWAS) have identified thousands of loci associated with cardiometabolic traits, however determining the gene(s) responsible is challenging. Integrating quantitative trait loci (QTL) in adipose tissue with GWAS loci as well as understanding the relationship between adipose gene expression and measured phenotypes can reveal regulatory mechanisms of cardiometabolic traits. I identified QTL for splice junctions, exons, and genes in 426 adipose tissue samples from the METabolic Syndrome in Men (METSIM) study. I detected isoform-specific associations and I identified sQTL colocalized with GWAS signals to find plausible isoform-specific mechanisms that may contribute to the function of GWAS loci. Additionally, I examined the cell type composition of 859 bulk adipose tissue samples, including the 426 used previously, and investigated the association between 22 cardiometabolic traits and cell type proportion estimates. I identified thousands of associations between cardiometabolic traits and gene expression levels genome-wide that are robust to the variation in cell type proportion across individuals. Finally, I performed an adipose eQTL meta-analysis using five datasets with up to 2,344 samples in the largest eQTL dataset to date in non-blood or brain tissues. I identified conditionally distinct eQTL signals and compared the effect sizes, minor allele frequencies, heritability, and tolerance to loss-of-function variants between primary and non-primary signals. I identified over a thousand candidate genes for cardiometabolic trait GWAS signals, some of which show evidence of causal mediation. While these results require future experimental testing, the work of this dissertation identifies thousands of GWAS candidate genes and potential regulatory mechanisms that may help future drug discoveries and treatments for disease.