RESEARCH ARTICLE In Silico Functional Pathway Annotation of 86 Established Prostate Cancer Risk Variants

Heritability is one of the strongest risk factors of prostate cancer, emphasizing the importance of the genetic contribution towards prostate cancer risk. To date, 86 established prostate can-cer risk variants have been identified by genome-wide association studies (GWAS). To deter-mine if these risk variants are located near genes that interact together in biological networks or pathways contributing to prostate cancer initiation or progression, we generated gene sets based on proximity to the 86 prostate cancer risk variants. We took two approaches to gener-ate gene lists. The first strategy included all immediate flanking genes, up- and downstream of the risk variant, regardless of distance from the index variant, and the second strategy included genes closest to the index GWASmarker and to variants in high LD (r20.8 in Euro-peans) with the index variant, within a 100 kb window up- and downstream. Pathway mapping of the two gene sets supported the importance of the androgen receptor-mediated signaling in prostate cancer biology. In addition, the hedgehog andWnt/β-catenin signaling pathways were identified in pathway mapping for the flanking gene set. We also used the HaploReg re-source to examine the 86 risk loci and variants high LD (r20.8) for functional elements. We found that there was a 12.8 fold (p = 2.9 x 10-4) enrichment for enhancer motifs in a stem cell line and a 4.4 fold (p = 1.1 x 10-3) enrichment of DNase hypersensitivity in a prostate adeno-carcinoma cell line, indicating that the risk and correlated variants are enriched for transcrip-tional regulatory motifs. Our pathway-based functional annotation of the prostate cancer risk variants highlights the potential regulatory function that GWAS risk markers, and their highly correlated variants, exert on genes. Our study also shows that these genes may function cooperatively in key signaling pathways in prostate cancer biology