Novel Methods for Multi-ancestry Polygenic Prediction and their Evaluations in 5.1 Million Individuals of Diverse Ancestry

Haoyu Zhang, PhD
Earl Stadtman tenure-track investigator, National Cancer Institute

Polygenic risk scores (PRS) are becoming more predictive of complex traits, but their performance in non-European populations is suboptimal. We introduce CT-SLEB, a method that integrates multiple techniques to calculate PRS from multi-ancestry genome-wide association study (GWAS) summary statistics, improving performance in non-European populations Our evaluation using datasets from 23andMe Inc., the Global Lipids Genetics Consortium, All of Us, and UK Biobank demonstrates CT-SLEBS superior to simpler methods and comparable or superior to a recent computationally intensive method.

Register if attending via Zoom