Leveraging fine-mapping and multipopulation training data to improve cross-population polygenic risk scores

Abstract

Polygenic risk scores (PRS) based on European training data suffer reduced accuracy in non-European target populations, exacerbating health disparities. This loss of accuracy predominantly stems from LD differences, MAF differences (including population-specific SNPs), and/or causal effect size differences. PRS based on training data from the non-European target population do not suffer from these limitations, but are currently limited by much smaller training sample sizes. Here, we propose PolyPred, a method that improves cross-population polygenic prediction by combining two complementary predictors: a new predictor that leverages functionally informed fine-mapping to estimate causal effects (instead of tagging effects), addressing LD differences; and BOLT-LMM, a published predictor. In the special case where a large training sample is available in the non-European target population (or a closely related population), we propose PolyPred+, which further incorporates the non-European training data, addressing MAF differences and causal effect size differences. PolyPred and PolyPred+ require individual-level training data (for their BOLT-LMM component), but we also propose analogous methods that replace the BOLT-LMM component with summary statistic-based components if only summary statistics are available. We applied PolyPred to 49 diseases and complex traits in 4 UK Biobank populations using UK Biobank British training data (average N=325K), and observed statistically significant average relative improvements in prediction accuracy vs. BOLT-LMM ranging from +7% in South Asians to +32% in Africans (and vs. LDpruning + P-value thresholding (P+T) ranging from +77% to +164%), consistent with simulations. We applied PolyPred+ to 23 diseases and complex traits in UK Biobank East Asians using both UK Biobank British (average N=325K) and Biobank Japan (average N=124K) training data, and observed statistically significant average relative improvements in prediction accuracy of +24% vs. BOLT-LMM and +12% vs. PolyPred. The summary statistic-based analogues of PolyPred and PolyPred+ attained similar improvements. In conclusion, PolyPred and PolyPred+ improve cross-population polygenic prediction accuracy, ameliorating health disparities.

Quantifying portable genetic effects and improving cross-ancestry genetic prediction with GWAS summary statistics

“…In this work, we showcased its effectiveness in cross-population risk prediction using an annotation derived from local genetic correlations. But we note that it is a general framework that can incorporate arbitrary sets of annotation data, such as the epigenetic annotations used in the PRS literature or LD and allele frequencies which have been shown to improve heritability estimation 20,22,32,53 ( Supplementary Note ). It may also be applied to improve PRS portability across other non-ancestry-related demographic groups 54 .…”

“…There have been three types of approaches to improve cross-ancestry genetic prediction in the literature. First, prioritizing causal variants using functional genomic annotations can improve the portability of PRS based on European GWAS [20][21][22] . Second, several studies combine multiple PRS trained in various populations using linear regression to optimize the predictive performance in the target (non-European) population 15,22,23 .…”

“…First, prioritizing causal variants using functional genomic annotations can improve the portability of PRS based on European GWAS [20][21][22] . Second, several studies combine multiple PRS trained in various populations using linear regression to optimize the predictive performance in the target (non-European) population 15,22,23 . The third approach parametrizes the degree to which genetic effects are correlated across populations, and integrates GWAS summary statistics from multiple populations in a multivariate model to improve effect size estimation and prediction accuracy in each respective population 15,[24][25][26] .…”

“…These models have achieved moderately improved predictive performance compared to conventional single-population approaches, but several critical limitations and challenges still remain. First, previous studies used epigenetic regulatory annotations to prioritize variants for PRS [20][21][22] . While these annotations improved PRS portability for some traits, they are not designed to quantify the correlated genetic effects between populations 27 , and there is no guarantee that the same set of annotations will improve PRS performance for all complex traits.…”

Polygenic Health Index, General Health, Pleiotropy, Embryo Selection and Disease Risk

“…It is well-established that PRS are more accurate within a population ancestrally homogeneous and similar to the training population -however, generally a positive effect in one ancestry will persist in more distant ancestries. Research on this topic is ongoing and of high interest [7,[47][48][49][50]. The primary motivation for this paper is to investigate whether a composite genetic health index is reflective of general health in principle and we therefore focused 2…”