Cancer predisposition variants in All of Us

Adam Bates will discuss results from leveraging the All of Us Research Program to study previously identified pathogenic germline variants and their relationship to cancer occurrence. This work highlights the importance of using diverse study populations to make cancer research more broadly applicable. 

Overview of Presentation

• We took over 500 pathogenic germline variants that had been discovered in TCGA and identified individuals in the All of Us Research Program that had these variants. 

• Presence of these variants was generally associated with increased cancer occurrence in some groups of genetic similarity, but not others. 

• Variants in specific genes showed strong associations with cancer phenotypes, especially in EUR-like individuals. For example, unsurprisingly, we saw strong associations between variants in BRCA2 and breast/ovarian cancer. 

• Importantly, pathogenic germline variants were not distributed evenly across the genetic similarity groups, showing significant enrichment in EUR-like individuals. 

• While our statistical power was somewhat limited in the All of Us version 6 dataset, we are excited about the improved power that will come in future All of Us releases.