Improving the performance of polygenic scores in diverse populations

• Applicants: Privatdozentin Dr. Amke Caliebe; Professor Michael Nothnagel, Ph.D.
• Subject Area: Epidemiology and Medical Biometry/Statistics ; Human Genetics
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 532723127

Numerous polygenic scores (PGSs) for individual genetic risk assessment of a variety of traits have been proposed in the past decade. These scores have overwhelmingly been developed on sample sets of European ancestry and their transferability to populations of different ancestries has repeatedly been shown to be problematic, thereby potentially aggravating global health disparities. Using extensive simulations mirroring different human population groups, also including admixed ones and isolates, and various disease etiologies, we will systematically investigate 1) which existing single-ancestry methods yield PGSs that are generally well transferable between population groups, 2) under which conditions a transferable PGS can be expected and which guiding principles should be followed to construct a robust and transferable score, 3) under which scenarios proposed trans-ancestry PGS methods perform sufficiently well. Our results will help narrowing worldwide health disparities by making the PGS approach applicable to more human populations.

• DFG Programme: Research Grants
• International Connection: Luxembourg
• Co-Investigator: Professorin Dr. Inke Regina Koenig
• Cooperation Partner: Patrick May, Ph.D.