bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Авг. 21, 2024

A major obstacle hindering the broad adoption of polygenic scores (PGS) is their lack "portability" to people that differ-in genetic ancestry or other characteristics-from GWAS samples in which effects were estimated. Here, we use UK Biobank measure change PGS prediction accuracy as a continuous function individuals' genome-wide dissimilarity sample ("genetic distance"). Our results highlight three gaps our understanding portability. First, extremely noisy at individual level and not well predicted by distance. In fact, variance explained comparably socioeconomic measures. Second, trends portability vary across traits. For several immunity-related traits, drops near zero quickly even intermediate levels This quick drop may reflect associations being more ancestry-specific traits than Third, show qualitative can depend on used. instance, for white blood cell count, (reduction mean squared error) increases with Together, cannot be understood through global groupings alone. There are other, understudied factors influencing portability, such specifics evolution trait its architecture, social context, construction score. Addressing these aid development application inform equitable genomic research.

Язык: Английский