Approved Research

Validation of a breast cancer risk prediction model incorporating a polygenic score for diverse ancestries

Lay summary

Various types of genetic data can be used to calculate risk of future breast cancer. We have a risk prediction model (riskScore) that combines clinical and genetic information. We intend to test the accuracy of riskScore in a large, diverse, real-world cohort. To assess the accuracy, we will apply our riskScore algorithm to the underlying genetic and clinical data available in UK BioBank, then compare riskScore predictions to actual breast cancer outcomes among UK BioBank participants. Additionally, we want to assess the accuracy of riskScore within diverse ancestries. The overarching public health impact of this project is to improve patient outcomes by validating a breast cancer risk prediction model that may inform targeted screening and earlier diagnosis.