Using population genetic models, researchers show that negative selection on complex traits limits phenotype prediction accuracy between populations.
Over the past decade, GWASs have uncovered a plethora of trait-associated loci. Experts have attempted to turn these associations into phenotype prediction models that aggregate variants across the genome into a polygenic risk score. However, a major challenge remains in applying these polygenic scores uniformly across populations. Recent analyses have shown that because many of the largest studies have focussed on European populations, polygenic scores may be biased and less informative in non-European populations.
There are several reasons why polygenic scores may not transfer between populations. One is that alleles have different effect sizes in different populations. Another reason is that differences in linkage disequilibrium between populations means researchers may tag the causal variants differently. Finally, another possibility is that polygenic score performance in Europeans may be inflated due to population stratification.
In this paper, published in AJHG, researchers proposed another possibility. They highlighted that each population has its own genetic architecture, owing to the evolutionary processes that give rise to traits. In this scenario, a population’s demographic history influences the number of causal variants and their frequencies, resulting in some phenotypic variance from causal variants that are population specific.
In the study, researchers used population genetic models to show that human demographic history and negative selection on complex traits can result in population-specific genetic architectures. They specifically used simulations under demographic scenarios of recent explosive population growth with varying amounts of negative selection as well as analyses of empirical data to test the role of private variants in complex traits.
The team found that in traits where alleles with the largest effect on the trait are under the strongest selection, approximately half of the heritability can be accounted for by variants in Europe that are absent from Africa. This leads to poor performance in phenotype prediction across these populations.
In their simulation results, they found that when there was no coupling between trait effects and fitness, approximately 30% of the heritability was from private variants. This proportion increased as the coupling increased.
Overall, these findings show that recent population growth and negative selection creates population-specific genetic architectures for phenotype, which directly reduces the accuracy of polygenic scores when applied between populations. This study has further emphasised that genetic association studies need to include more diverse populations to ensure the utility of phenotype prediction in all populations.
Image credit: By kjpargeter – freepik