Abstract 4956: Transcriptome-wide association study identifies new prostate cancer susceptibility genes in the OncoArray data

Abstract

Abstract Genome-wide association studies (GWAS) have identified over 150 genomic regions harboring risk variants for prostate cancer which explain one third of all familial risk. However, with some notable exceptions, the causal variants and target susceptibility genes at these risk loci have yet to be identified. Recent work has shown a strong overlap between loci associated with gene expression levels (eQTLs) in prostate tissue and GWAS loci, which suggests that the causal mechanism at a significant proportion of risk loci includes causal alleles that regulate expression levels of nearby susceptibility genes. While overlapping eQTLs with GWAS is a powerful method to prioritize susceptibility genes, it is often the case that multiple eQTLs co-localize at the GWAS risk region (due to linkage disequilibrium (LD) and correlations across transcript levels). This prohibits the identification of the true susceptibility gene as opposed to spurious co-localization at the same locus. We recently leveraged gene expression imputation to perform transcriptome-wide association studies (TWAS) as a principled approach to measure the strength of association between gene expression and disease status. Here, we use imputed expression to identify new susceptibility genes for prostate cancer in the OncoArray GWAS data. We integrate gene expression data from more than 44 tissues across ~4,000 individuals with GWAS of prostate cancer from the OncoArray in ~140,000 individuals. Our approach identified 118 susceptibility genes for prostate cancer that reside in 90 independent loci across the genome. Of these, we report 7 genes located more than 0.5 Megabases away from any previously reported GWAS loci for prostate cancer, thus providing new risk loci. Second, we use TWAS to investigate genes previously reported as susceptibility genes for prostate cancer through overlaps of eQTL and GWAS. We find 36 (out of 86 previously reported genes) to be significant in TWAS. Overall, our findings highlight the power of integrating gene expression data with GWAS and provide testable hypotheses for future functional validation of prostate cancer risk. Citation Format: Nicholas Mancuso, Wei Zheng, Kathryn Penney, The PRACTICAL Consortium, Zsofia Kote-Jarai, Christopher Haiman, Simon Gayther, Matthew Freedman, Bogdan Pasaniuc. Transcriptome-wide association study identifies new prostate cancer susceptibility genes in the OncoArray data [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 4956. doi:10.1158/1538-7445.AM2017-4956