Polygenic prediction across populations is influenced by ancestry, genetic architecture, and methodology

Abstract

Polygenic risk scores built from multi-ancestry genome-wide association studies (GWAS, PRSmulti) have the potential to improve PRS accuracy and generalizability across populations. To provide the best practice to leverage the increasing diversity of genomic studies, we used large-scale simulated and empirical data to investigate how ancestry composition, trait-specific genetic architecture, and PRS methodology affect the performance of PRSmulti as compared to PRS constructed from single-ancestry GWAS (PRSsingle). In both simulations on 6 various scenarios and empirical analyses on 17 anthropometric and blood panel traits, we showed that the accuracy of PRSmulti overall outperformed PRSsingle in the understudied target populations, except for a few comparisons where the understudied population only accounted for a very small proportion of the multi-ancestry GWAS. Further, using substantially fewer samples for traits such as height and mean corpuscular volume from Biobank Japan (BBJ) may achieve comparable accuracies to using 320,000 European (EUR) individuals from UK Biobank (UKBB). Finally, we find that incorporating PRS based on local ancestry-informed GWAS and large-scale EUR-based PRS improved predictive performance than using EUR-based PRS alone in understudied African (AFR) population, especially for less polygenic traits when there are variants with large ancestry-specific effects. Overall, our study provides insights into how ancestry composition and genetic architecture impact polygenic prediction across populations, particularly across imbalanced sample sizes. Our work also highlights the need for increasing diversity in genetic studies to achieve equitable PRS performance across ancestral populations and provides practical guidance on developing PRS from multiple resources.

Polygenic risk score portability for common diseases across genetically diverse populations

“…Specifically, we found that the PRS scores were more predictive when the selection of the appropriate PRS model for each particular individual was ancestry specific, which necessarily requires ancestry estimation prior to PRS estimation. It should be noted that the top selected models for CAD [22][23], were generated with genome-wide summary statistics derived from transethnic studies, which agrees with studies demonstrating the increased PRS predictability in multi-ethnic samples [11]. Although our approach does not solve the transferability problem impacting PRSs, we provide an alternative strategy to identify those conditions and ancestries where PRSs could work efficiently.…”

“…. Conversely, PRS performances relative to Europeans improved when we specifically selected the optimal PRS for a given condition and a given population ancestry, reinforcing the idea that optimal PRS methods are trait and context specific [11].…”

“…Although generalizing a single PRS model useful for all populations supposes the gold standard, it is becoming more established that optimal PRS methods are context specific [11].…”

The Role of Genomic-Informed Risk Assessments in Predicting Dementia Outcomes

“…Our use of mCAIDE was determined by the lack of extensive lifestyle and social determinants of health data, precluding the use of more comprehensive dementia risk scores. Second, the predictive accuracy of PRS models is affected by the sample size ratio between the EUR and minor GWAS, betweenancestry genetic architecture differences (LD, MAF, genetic correlation, heritability, effect size), and LD reference panels 34 , 35 . As such, the choice of PRS model to use for portability across populations will differ between traits.…”

Leveraging haplotype information in heritability estimation and polygenic prediction

“…These methods directly model LD patterns through haplotype information and should therefore be able to tag unobserved causal SNPs, which are poorly captured by an SNP array or imputed genotype data, due to the unobserved SNPs likely being in identity-by-state for long haplotypes being in identity-by-state [41]. Multiple approaches including local ancestry tracts inferred from haplotypes have also been developed to estimate SNP heritability [65] and construct reference-based PRS estimators [57].…”