BridgePRS : A powerful trans-ancestry Polygenic Risk Score method.

Area of Science:

• Genetics and Genomics
• Bioinformatics
• Precision Medicine

Background:

• Polygenic Risk Scores (PRS) are crucial for precision medicine but are often inaccurate in non-European populations due to reliance on European-ancestry Genome-Wide Association Studies (GWAS) data.
• This ancestry bias limits the equitable application of PRS in global health initiatives.
• Developing ancestry-aware PRS methods is essential for reducing health disparities.

Approach:

• Introducing BridgePRS, a novel Bayesian method that enhances PRS accuracy in under-represented populations by utilizing shared genetic effects across ancestries.
• Evaluating BridgePRS performance using simulated data and real-world UK Biobank and Biobank Japan GWAS summary statistics across 19 traits in African, South Asian, and East Asian individuals.
• Comparing BridgePRS against PRS-CSx and single-ancestry PRS methods, with out-of-cohort validation in the Mount Sinai BioMe Biobank.

Key Points:

• Simulations show BridgePRS performance improves relative to PRS-CSx under conditions of increased uncertainty, such as lower heritability and higher genetic diversity.
• Real-data analyses confirm BridgePRS superior predictive accuracy in African ancestry samples, demonstrating a 60% boost in mean R² for out-of-cohort prediction compared to PRS-CSx.
• BridgePRS offers a computationally efficient, full PRS analysis pipeline suitable for diverse populations.

Conclusions:

• BridgePRS significantly enhances the accuracy and utility of Polygenic Risk Scores in non-European populations.
• This method represents a substantial advancement for precision medicine, promoting more equitable genetic risk prediction.
• BridgePRS provides a powerful tool for researchers and clinicians aiming to serve diverse global populations.