The contribution of coding variants in cancer GWAS susceptibility regions to cancer risk

Abstract

Deepening our understanding of the genetic architecture of cancer could provide insights into cancer biology to help mitigate the health effects of increasing global cancer incidence. Elucidating the shared genetic etiology of cancer could reveal novel loci involved in cancer susceptibility. We investigated rare coding variation in genes near cancer susceptibility regions to identify genes with pleiotropic cancer associations. We used a nested case-control design to sample 195,507 participants with whole exome sequencing data from the UK Biobank cohort. We then conducted rare coding variant analyses in a generalized linear mixed-effects model framework to robustly control for population stratification and family relatedness. Our analyses identified a significant association between predicted deleterious rare coding variation in BRCA2 and cancer diagnosis. We also found a significant rare variant association in MC1R and suggestive associations in several other genes. These results highlight the potential for investigating genetic contributions to cross-cancer risk using large population-based samples that include individuals representing diverse genetic ancestries.