Improving GWAS-based Discovery for Early-onset Breast Cancer

In genome-wide association studies (GWAS), hundreds of thousands of genetic features are compared between individuals with a disease and without. To date, GWAS of breast cancer have identified dozens of susceptibility genes, but additional susceptibility genes likely exist. This may be especially true for early onset breast cancer which is believed to have stronger genetic determinants than later-onset disease. Most GWAS use very simple statistical methods to identify important genes, but we believe that these methods can be improved and enabling the identification of genes not detected in initial GWAS analyses. In this proposal, we describe methods that leverage knowledge of human biology and newly-developed statistical techniques to improve GWAS-based discovery. In pilot work, these methods have shown great promise for identifying new genes, and we propose to apply these methods to a GWAS of early-onset breast cancer.