Rare coding variants in GWAS identified loci with breast cancer risk

Abstract

To date, common genetic variants in ~ 109 loci have been identified for breast cancer risk via genome-wide association studies (GWAS). Most common variants found in GWAS are located in non-coding region, thus impeding the direct interpretation of their functional effects. They may be involved in regulation of gene expression, and rare functional variants in the coding region of these genes may change gene structure and function. These rare coding variants may contribute to breast cancer risk. Candidate genes were identified through systematic analysis of expression quantitative trait loci (eQTL) using three major sources; the Cancer Genome Atlas (TCGA), the Genotype-Tissue Expression (GTEx), and Molecular Taxonomy of Breast Cancer International Consortium (METABRIC). We prioritized functional variants into two groups, nonsynonymous and loss-of-function (LOF), using functional prediction algorithm. A total of 9,004 cases and 11,996 controls from three ethnic groups including Chinese, European Americans, and African Americans were investigated to examine associations of rare coding variants with breast cancer risk. Our comprehensive association analysis including additive and recessive models revealed potential candidate genes and rare coding variants associated with breast cancer risk.