Low generalizability of polygenic scores in African populations due to genetic and environmental diversity

Abstract

African populations are vastly underrepresented in genetic studies but have the most genetic variation and face wide-ranging environmental exposures globally. Because systematic evaluations of genetic prediction had not yet been conducted in ancestries that span African diversity, we calculated polygenic risk scores (PRS) in simulations across Africa and in empirical data from South Africa, Uganda, and the UK to better understand the generalizability of genetic studies. PRS accuracy improves with ancestry-matched discovery cohorts more than from ancestry-mismatched studies. Within ancestrally and ethnically diverse South Africans, we find that PRS accuracy is low for all traits but varies across groups. Differences in African ancestries contribute more to variability in PRS accuracy than other large cohort differences considered between individuals in the UK versus Uganda. We computed PRS in African ancestry populations using existing European-only versus ancestrally diverse genetic studies; the increased diversity produced the largest accuracy gains for hemoglobin concentration and white blood cell count, reflecting large-effect ancestry-enriched variants in genes known to influence sickle cell anemia and the allergic response, respectively. Differences in PRS accuracy across African ancestries originating from diverse regions are as large as across out-of-Africa continental ancestries, requiring commensurate nuance.

Transferability of genetic loci and polygenic scores for cardiometabolic traits in British Pakistanis and Bangladeshis

“…The copyright holder for this preprint this version posted June 24, 2021. ; https://doi.org/10.1101/2021.06.22.21259323 doi: medRxiv preprint 6 We used the Michigan Imputation Server 16 to perform imputation with the GenomeAsia pilot reference panel 17 , imputing from 336,133 autosomal, biallelic SNPs with matched alleles. Eagle v2.4 and Minimac v4 were used for phasing and imputation, respectively.…”

“…These have important implications to translational applications of genetic data such as causal inference with MR which could prioritise different prevention strategies or drug targets between ancestries, and clinical risk prediction. Whilst the predictive performance of PGSs derived from EUR populations in non-EUR individuals decreases with genetic distance [3][4][5][6] , the extent to which this attenuation is due to genetic drift (differences in linkage disequilibrium and allele frequency ) versus heterogeneity of causal genetic effects remains unclear. Furthermore, the potential clinical utility of a CAD PGS in a real-world healthcare system is largely unknown, since previous studies have mostly examined research cohorts composed of volunteers who are healthier and wealthier than average (e.g.…”

Stability of Polygenic Scores Across Discovery Genome-Wide Association Studies

“…5; 20-23 Given that the GWAS required for computing PGS have been disproportionately run for European-ancestry populations, [24][25][26][27][28] a fundamental challenge will be ensuring that diverse populations have equitable access to medically beneficial PGS, 29 as it has been demonstrated that that PGS are less predictive when the target and discovery populations have differing genetic ancestry or varying degrees of admixture. [30][31][32][33][34] Previous studies have evaluated PGS performance in terms of how well they predict phenotypes at the population level. However, if PGS are going to be adopted in the precisionmedicine setting, it is also necessary to examine how well PGS perform at predicting the risk for individual patients.…”

Quantitative Human Paleogenetics: What can Ancient DNA Tell us About Complex Trait Evolution?

“…In aggregate, these issues combine to substantially diminish the portability of polygenic scores between populations. Indeed, in present-day populations, the predictive accuracy of PRS degrades approximately linearly with increasing genetic distance from the cohort used to ascertain the GWAS (Scutari et al, 2016;Martin et al, 2017aMartin et al, , 2019Kim et al, 2018;Bitarello and Mathieson, 2020;Mostafavi et al, 2020;Majara et al, 2021). Even within a single ancestry group, the correlation between PRS calculated from different discovery GWAS shows considerable variance (Schultz et al, 2021).…”