BridgePRS: A powerful trans-ancestry Polygenic Risk Score method

Abstract Polygenic Risk Scores (PRS) are mostly derived using European ancestry genomewide association study (GWAS) data, resulting in substantially lower predictive power in non-European samples. We present BridgePRS, a novel Bayesian PRS method that leverages trans-ancestry GWAS data and compare its performance with PRS-CSx across 19 UK Biobank (UKB) traits in African and South Asian ancestry individuals. Validation in the UKB and the independent Mount Sinai Biome Biobank demonstrates superior performance of BridgePRS.