Study Highlights Limitations of Polygenic Risk Scores Based on Single-Ancestry Data

Polygenic risk scores (PRS) developed using single-ancestry genome-wide association study (GWAS) data perform poorly in people of other ethnicities, according to a paper appearing this week in Genome Biology, highlighting the need for genetic research in different populations. To date, GWASs have largely been conducted using populations of European descent. As a result, PRS have been effectively used to classify whether individuals of European descent are more likely to develop complex diseases but their accuracy in other ethnicities remains unclear. To investigate, a team led by scientists from the Georgia Institute of Technology evaluated the performance of existing PRS for prostate cancer — a disease that particularly impacts men of African descent — using data from the UK Biobank and a novel dataset from nearly 3,000 prostate cancer cases and controls from Ghana, Nigeria, Senegal, and South Africa. They find that polygenic predictions were more effective for European individuals than African individuals and, notably, that the inclusion of genetic data from African Americans only modestly improved PRS performance in sub-Saharan African individuals. "Unless well-powered GWAS[s] are undertaken in diverse populations, the accuracy and utility of PRS will be sub-optimal, exacerbating disparities in risk prediction and subsequent disease management," the study's authors write.