Population differentiation of polygenic score predictions under stabilizing selection

Abstract

1Given the many small-effect loci uncovered by genome-wide association studies (GWAS), polygenic scores have become central to the drive for genomic medicine and have spread into various areas including evolutionary studies of adaptation. While promising, these scores are fraught with issues of portability across populations, due to mis-estimated effect sizes and missing causal loci across populations unrepresented in large-scale GWAS. The poor portability of polygenic scores at first seems at odds with the view that much of common genetic variation is shared among populations. Here we investigate one potential cause of this discrepancy, stabilizing selection on complex traits. Somewhat counter-intuitively, while stabilizing selection to the same optimum phenotype leads to lower phenotypic differentiation among populations, it increases genetic differentiation at GWAS loci because it accelerates the turnover of polymorphisms underlying trait variation within populations. We develop theory to show how stabilizing selection impacts the utility of polygenic scores when applied to unrepresented populations. Specifically, we quantify their reduced prediction accuracy and find they can substantially overstate average genetic differences of phenotypes among populations. Our work emphasizes stabilizing selection to the same optimum as a useful null evolutionary model to draw connections between patterns of allele frequency and polygenic score differentiation.