Theoretical and empirical quantification of the accuracy of polygenic scores in ancestry divergent populations

Managing differential performance of polygenic risk scores across groups: Real-world experience of the eMERGE Network

The differential performance of polygenic risk scores (PRSs) by group is one of the major ethical barriers to their clinical use. It is also one of the main practical challenges for any implementation effort. The social repercussions of how people are grouped in PRS research must be considered in communications with research participants, including return of results. Here, we outline the decisions faced and choices made by a large multi-site clinical implementation study returning PRSs to diverse participants in handling this issue of differential performance. Our approach to managing the complexities associated with the differential performance of PRSs serves as a case study that can help future implementers of PRSs to plot an anticipatory course in response to this issue.

Optimizing clinico-genomic disease prediction across ancestries: a machine learning strategy with Pareto improvement

BACKGROUND

Accurate prediction of an individual's predisposition to diseases is vital for preventive medicine and early intervention. Various statistical and machine learning models have been developed for disease prediction using clinico-genomic data. However, the accuracy of clinico-genomic prediction of diseases may vary significantly across ancestry groups due to their unequal representation in clinical genomic datasets.

METHODS

We introduced a deep transfer learning approach to improve the performance of clinico-genomic prediction models for data-disadvantaged ancestry groups. We conducted machine learning experiments on multi-ancestral genomic datasets of lung cancer, prostate cancer, and Alzheimer's disease, as well as on synthetic datasets with built-in data inequality and distribution shifts across ancestry groups.

RESULTS

Deep transfer learning significantly improved disease prediction accuracy for data-disadvantaged populations in our multi-ancestral machine learning experiments. In contrast, transfer learning based on linear frameworks did not achieve comparable improvements for these data-disadvantaged populations.

CONCLUSIONS

This study shows that deep transfer learning can enhance fairness in multi-ancestral machine learning by improving prediction accuracy for data-disadvantaged populations without compromising prediction accuracy for other populations, thus providing a Pareto improvement towards equitable clinico-genomic prediction of diseases.

Body mass index stratification optimizes polygenic prediction of type 2 diabetes in cross-biobank analyses

Type 2 diabetes (T2D) shows heterogeneous body mass index (BMI) sensitivity. Here, we performed stratification based on BMI to optimize predictions for BMI-related diseases. We obtained BMI-stratified datasets using data from more than 195,000 individuals (nT2D = 55,284) from BioBank Japan (BBJ) and UK Biobank. T2D heritability in the low-BMI group was greater than that in the high-BMI group. Polygenic predictions of T2D toward low-BMI targets had pseudo-R2 values that were more than 22% higher than BMI-unstratified targets. Polygenic risk scores (PRSs) from low-BMI discovery outperformed PRSs from high BMI, while PRSs from BMI-unstratified discovery performed best. Pathway-specific PRSs demonstrated the biological contributions of pathogenic pathways. Low-BMI T2D cases showed higher rates of neuropathy and retinopathy. Combining BMI stratification and a method integrating cross-population effects, T2D predictions showed greater than 37% improvements over unstratified-matched-population prediction. We replicated findings in the Tohoku Medical Megabank (n = 26,000) and the second BBJ cohort (n = 33,096). Our findings suggest that target stratification based on existing traits can improve the polygenic prediction of heterogeneous diseases.

Divorce, genetic risk, and suicidal thoughts and behaviors in a sample with recurrent major depressive disorder

BACKGROUND

Theories of risk for suicidal thoughts and behaviors (STB) implicate both interpersonal and biological factors. Divorce/separation and aggregate genetic liability are robustly associated with STB, but have seldom been evaluated in conjunction with one another. Furthermore, whether these factors are effective predictors in high-risk populations is not clear.

METHODS

Analyses were conducted in a sample of Han Chinese women with severe recurrent major depressive disorder (maximum N = 4380). Logistic regressions were used to evaluate the associations between divorce/separation and polygenic scores (PGS) for suicidal ideation or behavior with STB. Where appropriate, additive interactions between divorce and PGS were tested.

RESULTS

Divorce/separation was significantly associated with increased risk of suicidal ideation, plans, and attempts (odds ratios = 1.28-1.61). PGS for suicidal ideation were not associated with STB, while PGS for suicidal behavior were associated with ideation and plans (odds ratios = 1.08-1.09). There were no significant interactions between divorce/separation and PGS.

CONCLUSIONS

Consistent with theories of suicidality, the disruption or end of an important interpersonal relationship is an indicator of risk for STB. Aggregate genetic liability for suicidal behavior more modestly contributes to risk, but does not exacerbate the negative impact of divorce. Thus, even within a high-risk sample, interpersonal and biological exposures distinguish between those who do and do not experience STB, and could motivate targeted screening. Further research is necessary to evaluate whether and how the context of divorce contributes to variation in its effect on STB risk.

Aspiring toward equitable benefits from genomic advances to individuals of ancestrally diverse backgrounds

Advancements in genomic technologies have shown remarkable promise for improving health trajectories. The Human Genome Project has catalyzed the integration of genomic tools into clinical practice, such as disease risk assessment, prenatal testing and reproductive genomics, cancer diagnostics and prognostication, and therapeutic decision making. Despite the promise of genomic technologies, their full potential remains untapped without including individuals of diverse ancestries and integrating social determinants of health (SDOHs). The NHGRI launched the 2020 Strategic Vision with ten bold predictions by 2030, including "individuals from ancestrally diverse backgrounds will benefit equitably from advances in human genomics." Meeting this goal requires a holistic approach that brings together genomic advancements with careful consideration to healthcare access as well as SDOHs to ensure that translation of genetics research is inclusive, affordable, and accessible and ultimately narrows rather than widens health disparities. With this prediction in mind, this review delves into the two paramount applications of genetic testing-reproductive genomics and precision oncology. When discussing these applications of genomic advancements, we evaluate current accessibility limitations, highlight challenges in achieving representativeness, and propose paths forward to realize the ultimate goal of their equitable applications.

Towards fair and clinically relevant polygenic predictions

Lennon et al. recently proposed a clinical polygenic score (PGS) pipeline as part of the Electronic Medical Records and Genomics (eMERGE) network initiative. In this spotlight article we discuss the broader context for the use of PGS in preventive medicine and highlight key limitations and challenges facing their inclusion in prediction models.

Leveraging functional genomic annotations and genome coverage to improve polygenic prediction of complex traits within and between ancestries

We develop a method, SBayesRC, that integrates genome-wide association study (GWAS) summary statistics with functional genomic annotations to improve polygenic prediction of complex traits. Our method is scalable to whole-genome variant analysis and refines signals from functional annotations by allowing them to affect both causal variant probability and causal effect distribution. We analyze 50 complex traits and diseases using ∼7 million common single-nucleotide polymorphisms (SNPs) and 96 annotations. SBayesRC improves prediction accuracy by 14% in European ancestry and up to 34% in cross-ancestry prediction compared to the baseline method SBayesR, which does not use annotations, and outperforms other methods, including LDpred2, LDpred-funct, MegaPRS, PolyPred-S and PRS-CSx. Investigation of factors affecting prediction accuracy identifies a significant interaction between SNP density and annotation information, suggesting whole-genome sequence variants with annotations may further improve prediction. Functional partitioning analysis highlights a major contribution of evolutionary constrained regions to prediction accuracy and the largest per-SNP contribution from nonsynonymous SNPs.

Genetic modifiers of rare variants in monogenic developmental disorder loci

Rare damaging variants in a large number of genes are known to cause monogenic developmental disorders (DDs) and have also been shown to cause milder subclinical phenotypes in population cohorts. Here, we show that carrying multiple (2-5) rare damaging variants across 599 dominant DD genes has an additive adverse effect on numerous cognitive and socioeconomic traits in UK Biobank, which can be partially counterbalanced by a higher educational attainment polygenic score (EA-PGS). Phenotypic deviators from expected EA-PGS could be partly explained by the enrichment or depletion of rare DD variants. Among carriers of rare DD variants, those with a DD-related clinical diagnosis had a substantially lower EA-PGS and more severe phenotype than those without a clinical diagnosis. Our results suggest that the overall burden of both rare and common variants can modify the expressivity of a phenotype, which may then influence whether an individual reaches the threshold for clinical disease.

Multi-trait GWAS for diverse ancestries: mapping the knowledge gap

BACKGROUND

Approximately 95% of samples analyzed in univariate genome-wide association studies (GWAS) are of European ancestry. This bias toward European ancestry populations in association screening also exists for other analyses and methods that are often developed and tested on European ancestry only. However, existing data in non-European populations, which are often of modest sample size, could benefit from innovative approaches as recently illustrated in the context of polygenic risk scores.

METHODS

Here, we extend and assess the potential limitations and gains of our multi-trait GWAS pipeline, JASS (Joint Analysis of Summary Statistics), for the analysis of non-European ancestries. To this end, we conducted the joint GWAS of 19 hematological traits and glycemic traits across five ancestries (European (EUR), admixed American (AMR), African (AFR), East Asian (EAS), and South-East Asian (SAS)).

RESULTS

We detected 367 new genome-wide significant associations in non-European populations (15 in Admixed American (AMR), 72 in African (AFR) and 280 in East Asian (EAS)). New associations detected represent 5%, 17% and 13% of associations in the AFR, AMR and EAS populations, respectively. Overall, multi-trait testing increases the replication of European associated loci in non-European ancestry by 15%. Pleiotropic effects were highly similar at significant loci across ancestries (e.g. the mean correlation between multi-trait genetic effects of EUR and EAS ancestries was 0.88). For hematological traits, strong discrepancies in multi-trait genetic effects are tied to known evolutionary divergences: the ARKC1 loci, which is adaptive to overcome p.vivax induced malaria.

CONCLUSIONS

Multi-trait GWAS can be a valuable tool to narrow the genetic knowledge gap between European and non-European populations.

Risk factors affecting polygenic score performance across diverse cohorts

Apart from ancestry, personal or environmental covariates may contribute to differences in polygenic score (PGS) performance. We analyzed effects of covariate stratification and interaction on body mass index (BMI) PGS (PGS BMI ) across four cohorts of European (N=491,111) and African (N=21,612) ancestry. Stratifying on binary covariates and quintiles for continuous covariates, 18/62 covariates had significant and replicable R 2 differences among strata. Covariates with the largest differences included age, sex, blood lipids, physical activity, and alcohol consumption, with R 2 being nearly double between best and worst performing quintiles for certain covariates. 28 covariates had significant PGS BMI -covariate interaction effects, modifying PGS BMI effects by nearly 20% per standard deviation change. We observed overlap between covariates that had significant R 2 differences among strata and interaction effects - across all covariates, their main effects on BMI were correlated with their maximum R 2 differences and interaction effects (0.56 and 0.58, respectively), suggesting high-PGS BMI individuals have highest R 2 and increase in PGS effect. Using quantile regression, we show the effect of PGS BMI increases as BMI itself increases, and that these differences in effects are directly related to differences in R 2 when stratifying by different covariates. Given significant and replicable evidence for context-specific PGS BMI performance and effects, we investigated ways to increase model performance taking into account non-linear effects. Machine learning models (neural networks) increased relative model R 2 (mean 23%) across datasets. Finally, creating PGS BMI directly from GxAge GWAS effects increased relative R 2 by 7.8%. These results demonstrate that certain covariates, especially those most associated with BMI, significantly affect both PGS BMI performance and effects across diverse cohorts and ancestries, and we provide avenues to improve model performance that consider these effects.

Bayesian approach to assessing population differences in genetic risk of disease with application to prostate cancer

Population differences in risk of disease are common, but the potential genetic basis for these differences is not well understood. A standard approach is to compare genetic risk across populations by testing for mean differences in polygenic scores, but existing studies that use this approach do not account for statistical noise in effect estimates (i.e., the GWAS betas) that arise due to the finite sample size of GWAS training data. Here, we show using Bayesian polygenic score methods that the level of uncertainty in estimates of genetic risk differences across populations is highly dependent on the GWAS training sample size, the polygenicity (number of causal variants), and genetic distance (FST) between the populations considered. We derive a Wald test for formally assessing the difference in genetic risk across populations, which we show to have calibrated type 1 error rates under a simplified assumption that all SNPs are independent, which we achieve in practise using linkage disequilibrium (LD) pruning. We further provide closed-form expressions for assessing the uncertainty in estimates of relative genetic risk across populations under the special case of an infinitesimal genetic architecture. We suggest that for many complex traits and diseases, particularly those with more polygenic architectures, current GWAS sample sizes are insufficient to detect moderate differences in genetic risk across populations, though more substantial differences in relative genetic risk (relative risk > 1.5) can be detected. We show that conventional approaches that do not account for sampling error from the training sample, such as using a simple t-test, have very high type 1 error rates. When applying our approach to prostate cancer, we demonstrate a higher genetic risk in African Ancestry men, with lower risk in men of European followed by East Asian ancestry.

Phenomewide Association Study of Health Outcomes Associated With the Genetic Correlates of 25 Hydroxyvitamin D Concentration and Vitamin D Binding Protein Concentration

While it is known that vitamin D deficiency is associated with adverse bone outcomes, it remains unclear whether low vitamin D status may increase the risk of a wider range of health outcomes. We had the opportunity to explore the association between common genetic variants associated with both 25 hydroxyvitamin D (25OHD) and the vitamin D binding protein (DBP, encoded by the GC gene) with a comprehensive range of health disorders and laboratory tests in a large academic medical center. We used summary statistics for 25OHD and DBP to generate polygenic scores (PGS) for 66,482 participants with primarily European ancestry and 13,285 participants with primarily African ancestry from the Vanderbilt University Medical Center Biobank (BioVU). We examined the predictive properties of PGS25OHD, and two scores related to DBP concentration with respect to 1322 health-related phenotypes and 315 laboratory-measured phenotypes from electronic health records. In those with European ancestry: (a) the PGS25OHD and PGSDBP scores, and individual SNPs rs4588 and rs7041 were associated with both 25OHD concentration and 1,25 dihydroxyvitamin D concentrations; (b) higher PGS25OHD was associated with decreased concentrations of triglycerides and cholesterol, and reduced risks of vitamin D deficiency, disorders of lipid metabolism, and diabetes. In general, the findings for the African ancestry group were consistent with findings from the European ancestry analyses. Our study confirms the utility of PGS and two key variants within the GC gene (rs4588 and rs7041) to predict the risk of vitamin D deficiency in clinical settings and highlights the shared biology between vitamin D-related genetic pathways a range of health outcomes.

The association between DNA methylation and human height and a prospective model of DNA methylation-based height prediction

As a vital anthropometric characteristic, human height information not only helps to understand overall developmental status and genetic risk factors, but is also important for forensic DNA phenotyping. We utilized linear regression analysis to test the association between each CpG probe and the height phenotype. Next, we designed a methylation sequencing panel targeting 959 CpGs and subsequent height inference models were constructed for the Chinese population. A total of 11,730 height-associated sites were identified. By employing KPCA and deep neural networks, a prediction model was developed, of which the cross-validation RMSE, MAE and R2 were 5.62 cm, 4.45 cm and 0.64, respectively. Genetic factors could explain 39.4% of the methylation level variance of sites used in the height inference models. Collectively, we demonstrated an association between height and DNA methylation status through an EWAS analysis. Targeted methylation sequencing of only 959 CpGs combined with deep learning techniques could provide a model to estimate human height with higher accuracy than SNP-based prediction models.

Novel ancestry-specific primary open-angle glaucoma loci and shared biology with vascular mechanisms and cell proliferation

Primary open-angle glaucoma (POAG), a leading cause of irreversible blindness globally, shows disparity in prevalence and manifestations across ancestries. We perform meta-analysis across 15 biobanks (of the Global Biobank Meta-analysis Initiative) (n = 1,487,441: cases = 26,848) and merge with previous multi-ancestry studies, with the combined dataset representing the largest and most diverse POAG study to date (n = 1,478,037: cases = 46,325) and identify 17 novel significant loci, 5 of which were ancestry specific. Gene-enrichment and transcriptome-wide association analyses implicate vascular and cancer genes, a fifth of which are primary ciliary related. We perform an extensive statistical analysis of SIX6 and CDKN2B-AS1 loci in human GTEx data and across large electronic health records showing interaction between SIX6 gene and causal variants in the chr9p21.3 locus, with expression effect on CDKN2A/B. Our results suggest that some POAG risk variants may be ancestry specific, sex specific, or both, and support the contribution of genes involved in programmed cell death in POAG pathogenesis.

Recent advances in polygenic scores: translation, equitability, methods and FAIR tools

Polygenic scores (PGS) can be used for risk stratification by quantifying individuals' genetic predisposition to disease, and many potentially clinically useful applications have been proposed. Here, we review the latest potential benefits of PGS in the clinic and challenges to implementation. PGS could augment risk stratification through combined use with traditional risk factors (demographics, disease-specific risk factors, family history, etc.), to support diagnostic pathways, to predict groups with therapeutic benefits, and to increase the efficiency of clinical trials. However, there exist challenges to maximizing the clinical utility of PGS, including FAIR (Findable, Accessible, Interoperable, and Reusable) use and standardized sharing of the genomic data needed to develop and recalculate PGS, the equitable performance of PGS across populations and ancestries, the generation of robust and reproducible PGS calculations, and the responsible communication and interpretation of results. We outline how these challenges may be overcome analytically and with more diverse data as well as highlight sustained community efforts to achieve equitable, impactful, and responsible use of PGS in healthcare.

Highly parameterized polygenic scores tend to overfit to population stratification via random effects

Polygenic scores (PGSs), increasingly used in clinical settings, frequently include many genetic variants, with performance typically peaking at thousands of variants. Such highly parameterized PGSs often include variants that do not pass a genome-wide significance threshold. We propose a mathematical perspective that renders the effects of many of these non-significant variants random rather than causal, with the randomness capturing population structure. We devise methods to assess variant effect randomness and population stratification bias. Applying these methods to 141 traits from the UK Biobank, we find that, for many PGSs, the effects of non-significant variants are considerably random, with the extent of randomness associated with the degree of overfitting to population structure of the discovery cohort. Our findings explain why highly parameterized PGSs simultaneously have superior cohort-specific performance and limited generalizability, suggesting the critical need for variant randomness tests in PGS evaluation. Supporting code and a dashboard are available at https://github.com/songlab-cal/StratPGS.

Principles and methods for transferring polygenic risk scores across global populations

Polygenic risk scores (PRSs) summarize the genetic predisposition of a complex human trait or disease and may become a valuable tool for advancing precision medicine. However, PRSs that are developed in populations of predominantly European genetic ancestries can increase health disparities due to poor predictive performance in individuals of diverse and complex genetic ancestries. We describe genetic and modifiable risk factors that limit the transferability of PRSs across populations and review the strengths and weaknesses of existing PRS construction methods for diverse ancestries. Developing PRSs that benefit global populations in research and clinical settings provides an opportunity for innovation and is essential for health equity.

The Mexican Biobank Project promotes genetic discovery, inclusive science and local capacity building

Diversifying genotype-phenotype databases is essential to understanding complex trait and disease etiology across different environments and genetic ancestries. The rise of biobanks across the world is helping reveal the genetic and environmental architecture of multiple disease traits but the diversity they capture remains limited. To help close this gap, the Mexican Biobank (MXB) Project was recently generated, and has already revealed fine-scale genetic ancestries and demographic histories across the country, and their impact on trait-relevant genetic variation. This will help guide future genetic epidemiology and public health efforts, and has also improved polygenic prediction for several traits in Mexican populations compared with using data from other genome-wide association studies, such as the UK Biobank. The MXB illustrates the importance of transnational initiatives and funding calls that prioritize local leadership and capacity building to move towards inclusive genomic science.

BridgePRS leverages shared genetic effects across ancestries to increase polygenic risk score portability

Here we present BridgePRS, a novel Bayesian polygenic risk score (PRS) method that leverages shared genetic effects across ancestries to increase PRS portability. We evaluate BridgePRS via simulations and real UK Biobank data across 19 traits in individuals of African, South Asian and East Asian ancestry, using both UK Biobank and Biobank Japan genome-wide association study summary statistics; out-of-cohort validation is performed in the Mount Sinai (New York) BioMe biobank. BridgePRS is compared with the leading alternative, PRS-CSx, and two other PRS methods. Simulations suggest that the performance of BridgePRS relative to PRS-CSx increases as uncertainty increases: with lower trait heritability, higher polygenicity and greater between-population genetic diversity; and when causal variants are not present in the data. In real data, BridgePRS has a 61% larger average R2 than PRS-CSx in out-of-cohort prediction of African ancestry samples in BioMe (P = 6 × 10-5). BridgePRS is a computationally efficient, user-friendly and powerful approach for PRS analyses in non-European ancestries.

Pervasive correlations between causal disease effects of proximal SNPs vary with functional annotations and implicate stabilizing selection

The genetic architecture of human diseases and complex traits has been extensively studied, but little is known about the relationship of causal disease effect sizes between proximal SNPs, which have largely been assumed to be independent. We introduce a new method, LD SNP-pair effect correlation regression (LDSPEC), to estimate the correlation of causal disease effect sizes of derived alleles between proximal SNPs, depending on their allele frequencies, LD, and functional annotations; LDSPEC produced robust estimates in simulations across various genetic architectures. We applied LDSPEC to 70 diseases and complex traits from the UK Biobank (average N=306K), meta-analyzing results across diseases/traits. We detected significantly nonzero effect correlations for proximal SNP pairs (e.g., -0.37±0.09 for low-frequency positive-LD 0-100bp SNP pairs) that decayed with distance (e.g., -0.07±0.01 for low-frequency positive-LD 1-10kb), varied with allele frequency (e.g., -0.15±0.04 for common positive-LD 0-100bp), and varied with LD between SNPs (e.g., +0.12±0.05 for common negative-LD 0-100bp) (because we consider derived alleles, positive-LD and negative-LD SNP pairs may yield very different results). We further determined that SNP pairs with shared functions had stronger effect correlations that spanned longer genomic distances, e.g., -0.37±0.08 for low-frequency positive-LD same-gene promoter SNP pairs (average genomic distance of 47kb (due to alternative splicing)) and -0.32±0.04 for low-frequency positive-LD H3K27ac 0-1kb SNP pairs. Consequently, SNP-heritability estimates were substantially smaller than estimates of the sum of causal effect size variances across all SNPs (ratio of 0.87±0.02 across diseases/traits), particularly for certain functional annotations (e.g., 0.78±0.01 for common Super enhancer SNPs)-even though these quantities are widely assumed to be equal. We recapitulated our findings via forward simulations with an evolutionary model involving stabilizing selection, implicating the action of linkage masking, whereby haplotypes containing linked SNPs with opposite effects on disease have reduced effects on fitness and escape negative selection.

Pervasive correlations between causal disease effects of proximal SNPs vary with functional annotations and implicate stabilizing selection

The genetic architecture of human diseases and complex traits has been extensively studied, but little is known about the relationship of causal disease effect sizes between proximal SNPs, which have largely been assumed to be independent. We introduce a new method, LD SNP-pair effect correlation regression (LDSPEC), to estimate the correlation of causal disease effect sizes of derived alleles between proximal SNPs, depending on their allele frequencies, LD, and functional annotations; LDSPEC produced robust estimates in simulations across various genetic architectures. We applied LDSPEC to 70 diseases and complex traits from the UK Biobank (average N=306K), meta-analyzing results across diseases/traits. We detected significantly nonzero effect correlations for proximal SNP pairs (e.g., -0.37±0.09 for low-frequency positive-LD 0-100bp SNP pairs) that decayed with distance (e.g., -0.07±0.01 for low-frequency positive-LD 1-10kb), varied with allele frequency (e.g., -0.15±0.04 for common positive-LD 0-100bp), and varied with LD between SNPs (e.g., +0.12±0.05 for common negative-LD 0-100bp) (because we consider derived alleles, positive-LD and negative-LD SNP pairs may yield very different results). We further determined that SNP pairs with shared functions had stronger effect correlations that spanned longer genomic distances, e.g., -0.37±0.08 for low-frequency positive-LD same-gene promoter SNP pairs (average genomic distance of 47kb (due to alternative splicing)) and -0.32±0.04 for low-frequency positive-LD H3K27ac 0-1kb SNP pairs. Consequently, SNP-heritability estimates were substantially smaller than estimates of the sum of causal effect size variances across all SNPs (ratio of 0.87±0.02 across diseases/traits), particularly for certain functional annotations (e.g., 0.78±0.01 for common Super enhancer SNPs)-even though these quantities are widely assumed to be equal. We recapitulated our findings via forward simulations with an evolutionary model involving stabilizing selection, implicating the action of linkage masking, whereby haplotypes containing linked SNPs with opposite effects on disease have reduced effects on fitness and escape negative selection.

Genetic risk prediction in Hispanics/Latinos: milestones, challenges, and social-ethical considerations

Genome-wide association studies (GWAS) have allowed the identification of disease-associated variants, which can be leveraged to build polygenic scores (PGSs). Even though PGSs can be a valuable tool in personalized medicine, their predictive power is limited in populations of non-European ancestry, particularly in admixed populations. Recent efforts have focused on increasing racial and ethnic diversity in GWAS, thus, addressing some of the limitations of genetic risk prediction in these populations. Even with these efforts, few studies focus exclusively on Hispanics/Latinos. Additionally, Hispanic/Latino populations are often considered a single population despite varying admixture proportions between and within ethnic groups, diverse genetic heterogeneity, and demographic history. Combined with highly heterogeneous environmental and socioeconomic exposures, this diversity can reduce the transferability of genetic risk prediction models. Given the recent increase of genomic studies that include Hispanics/Latinos, we review the milestones and efforts that focus on genetic risk prediction, summarize the potential for improving PGS transferability, and highlight the challenges yet to be addressed. Additionally, we summarize social-ethical considerations and provide ideas to promote genetic risk prediction models that can be implemented equitably.

Applying polygenic risk score methods to pharmacogenomics GWAS: challenges and opportunities

Polygenic risk scores (PRSs) have emerged as promising tools for the prediction of human diseases and complex traits in disease genome-wide association studies (GWAS). Applying PRSs to pharmacogenomics (PGx) studies has begun to show great potential for improving patient stratification and drug response prediction. However, there are unique challenges that arise when applying PRSs to PGx GWAS beyond those typically encountered in disease GWAS (e.g. Eurocentric or trans-ethnic bias). These challenges include: (i) the lack of knowledge about whether PGx or disease GWAS/variants should be used in the base cohort (BC); (ii) the small sample sizes in PGx GWAS with corresponding low power and (iii) the more complex PRS statistical modeling required for handling both prognostic and predictive effects simultaneously. To gain insights in this landscape about the general trends, challenges and possible solutions, we first conduct a systematic review of both PRS applications and PRS method development in PGx GWAS. To further address the challenges, we propose (i) a novel PRS application strategy by leveraging both PGx and disease GWAS summary statistics in the BC for PRS construction and (ii) a new Bayesian method (PRS-PGx-Bayesx) to reduce Eurocentric or cross-population PRS prediction bias. Extensive simulations are conducted to demonstrate their advantages over existing PRS methods applied in PGx GWAS. Our systematic review and methodology research work not only highlights current gaps and key considerations while applying PRS methods to PGx GWAS, but also provides possible solutions for better PGx PRS applications and future research.

Causal interpretations of family GWAS in the presence of heterogeneous effects

Family-based genome-wide association studies (GWAS) have emerged as a gold standard for assessing causal effects of alleles and polygenic scores. Notably, family studies are often claimed to provide an unbiased estimate of the average causal effect (or average treatment effect; ATE) of an allele, on the basis of an analogy between the random transmission of alleles from parents to children and a randomized controlled trial. Here, we show that this interpretation does not hold in general. Because Mendelian segregation only randomizes alleles among children of heterozygotes, the effects of alleles in the children of homozygotes are not observable. Consequently, if an allele has different average effects in the children of homozygotes and heterozygotes, as can arise in the presence of gene-by-environment interactions, gene-by-gene interactions, or differences in LD patterns, family studies provide a biased estimate of the average effect in the sample. At a single locus, family-based association studies can be thought of as providing an unbiased estimate of the average effect in the children of heterozygotes (i.e., a local average treatment effect; LATE). This interpretation does not extend to polygenic scores, however, because different sets of SNPs are heterozygous in each family. Therefore, other than under specific conditions, the within-family regression slope of a PGS cannot be assumed to provide an unbiased estimate for any subset or weighted average of families. Instead, family-based studies can be reinterpreted as enabling an unbiased estimate of the extent to which Mendelian segregation at loci in the PGS contributes to the population-level variance in the trait. Because this estimate does not include the between-family variance, however, this interpretation applies to only (roughly) half of the sample PGS variance. In practice, the potential biases of a family-based GWAS are likely smaller than those arising from confounding in a standard, population-based GWAS, and so family studies remain important for the dissection of genetic contributions to phenotypic variation. Nonetheless, the causal interpretation of family-based GWAS estimates is less straightforward than has been widely appreciated.

Genome-wide association study analysis of disease severity in Acne reveals novel biological insights

Acne vulgaris is a common skin disease that affects >85% of teenage young adults among which >8% develop severe lesions that leaves permanent scars. Genetic heritability studies of acne in twin cohorts have estimated that the heritability for acne is 80%. Previous genome-wide association studies (GWAS) have identified 50 genetic loci associated with increased risk of developing acne when compared to healthy individuals. However only a few studies have investigated genetic association with disease severity. GWAS of disease progression may provide a more effective approach to unveil potential disease modifying therapeutic targets. Here, we performed a multi-ethnic GWAS analysis to capture disease severity in acne patients by using individuals with normal acne as a control. Our cohort consists of a total of 2,956 participants, including 290 severe acne cases and 930 normal acne controls from FinnGen, and 522 cases and 1,214 controls from BioVU. We also performed mendelian randomization (MR), colocalization analyses and transcriptome-wide association study (TWAS) to identify putative causal genes. Lastly, we performed gene-set enrichment analysis using MAGMA to implicate biological pathways that drive disease severity in Acne. We identified two new loci associated with acne severity at the genome-wide significance level, six novel associated genes by MR, colocalization and TWAS analyses, including genes CDC7, SLC7A1, ADAM23, TTLL10, CDK20 and DNAJA4 , and 5 novel pathways by MAGMA analyses. Our study suggests that the etiologies of acne susceptibility and severity have limited overlap, with only 26% of known acne risk loci presenting nominal association with acne severity and none of the novel severity associated genes reported as associated with acne risk in previous GWAS.

Molecular genetics of neuropsychiatric illness: some musings

Research into the genetic underpinnings of neuropsychiatric illness has occurred at many levels. As more information accumulates, it appears that many approaches may each offer their unique perspective. The search for low penetrance and common variants, that may mediate risk, has necessitated the formation of many international consortia, to pool resources, and achieve the large sample sizes needed to discover these variants. There has been the parallel development of statistical methods to analyse large datasets and present summary statistics which allows data comparison across studies. Even so, the results of studies on well-characterised clinical datasets of modest sizes can be enlightening and provide important clues to understanding these complex disorders. We describe the use of common variants, at multiallelic loci like TOMM40 and APOE to study dementia, weighted genetic risk scores for alcohol-induced liver cirrhosis and whole exome sequencing to identify rare variants in genes like PLA2G6 in familial psychoses and schizophrenia in our Indian population.

Polygenic risk scores for disease risk prediction in Africa: current challenges and future directions

Early identification of genetic risk factors for complex diseases can enable timely interventions and prevent serious outcomes, including mortality. While the genetics underlying many Mendelian diseases have been elucidated, it is harder to predict risk for complex diseases arising from the combined effects of many genetic variants with smaller individual effects on disease aetiology. Polygenic risk scores (PRS), which combine multiple contributing variants to predict disease risk, have the potential to influence the implementation for precision medicine. However, the majority of existing PRS were developed from European data with limited transferability to African populations. Notably, African populations have diverse genetic backgrounds, and a genomic architecture with smaller haplotype blocks compared to European genomes. Subsequently, growing evidence shows that using large-scale African ancestry cohorts as discovery for PRS development may generate more generalizable findings. Here, we (1) discuss the factors contributing to the poor transferability of PRS in African populations, (2) showcase the novel Africa genomic datasets for PRS development, (3) explore the potential clinical utility of PRS in African populations, and (4) provide insight into the future of PRS in Africa.

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

Polygenic risk scores (PRSs) developed from multi-ancestry genome-wide association studies (GWASs), PRSmulti, hold promise for improving PRS accuracy and generalizability across populations. To establish best practices for leveraging the increasing diversity of genomic studies, we investigated how various factors affect the performance of PRSmulti compared with PRSs constructed from single-ancestry GWASs (PRSsingle). Through extensive simulations and empirical analyses, we showed that PRSmulti overall outperformed PRSsingle in understudied populations, except when the understudied population represented a small proportion of the multi-ancestry GWAS. Furthermore, integrating PRSs based on local ancestry-informed GWASs and large-scale, European-based PRSs improved predictive performance in understudied African populations, especially for less polygenic traits with large-effect ancestry-enriched variants. Our work highlights the importance of diversifying genomic studies to achieve equitable PRS performance across ancestral populations and provides guidance for developing PRSs from multiple studies.

Integration of genetic fine-mapping and multi-omics data reveals candidate effector genes for hypertension

Genome-wide association studies of blood pressure (BP) have identified >1,000 loci, but the effector genes and biological pathways at these loci are mostly unknown. Using published association summary statistics, we conducted annotation-informed fine-mapping incorporating tissue-specific chromatin segmentation and colocalization to identify causal variants and candidate effector genes for systolic BP, diastolic BP, and pulse pressure. We observed 532 distinct signals associated with ≥2 BP traits and 84 with all three. For >20% of signals, a single variant accounted for >75% posterior probability, 65 were missense variants in known (SLC39A8, ADRB2, and DBH) and previously unreported BP candidate genes (NRIP1 and MMP14). In disease-relevant tissues, we colocalized >80 and >400 distinct signals for each BP trait with cis-eQTLs and regulatory regions from promoter capture Hi-C, respectively. Integrating mouse, human disorder, gene expression and tissue abundance data, and literature review, we provide consolidated evidence for 436 BP candidate genes for future functional validation and discover several potential drug targets.

Boosting the power of genome-wide association studies within and across ancestries by using polygenic scores

Genome-wide association studies (GWASs) have been mostly conducted in populations of European ancestry, which currently limits the transferability of their findings to other populations. Here, we show, through theory, simulations and applications to real data, that adjustment of GWAS analyses for polygenic scores (PGSs) increases the statistical power for discovery across all ancestries. We applied this method to analyze seven traits available in three large biobanks with participants of East Asian ancestry (n = 340,000 in total) and report 139 additional associations across traits. We also present a two-stage meta-analysis strategy whereby, in contributing cohorts, a PGS-adjusted GWAS is rerun using PGSs derived from a first round of a standard meta-analysis. On average, across traits, this approach yields a 1.26-fold increase in the number of detected associations (range 1.07- to 1.76-fold increase). Altogether, our study demonstrates the value of using PGSs to increase the power of GWASs in underrepresented populations and promotes such an analytical strategy for future GWAS meta-analyses.

Factors affecting the accuracy of genomic prediction in joint pig populations

Genomic prediction (GP) has greatly advanced animal and plant breeding over the past two decades. GP in joint populations is a feasible method to improve the accuracy of genomic estimated breeding values in small populations. However, there is still a need to understand the factors that influence GP in joint populations. This study used simulated data and real data from Duroc pig populations to examine the impact of linkage disequilibrium (LD), causal variants effect sizes (CVESs), and minor allele frequencies (MAF) of SNPs on the accuracy of genomic prediction in joint populations. Three prediction methods were used: genomic best linear unbiased prediction (GBLUP), single-step GBLUP and multi-trait GBLUP. Results from the simulated datasets showed that the accuracies of GP in joint populations were always higher than those in a single population when only LD inconsistencies existed. However, single-step GBLUP accuracy in joint populations decreased as the correlation of MAF between populations decreased, while the accuracy of GBLUP is consistently higher in joint populations than in a single population. As the correlation of CVES between populations decreased, the accuracy of both GBLUP and single-step GBLUP in joint populations declined. Analysis of real Duroc populations showed low genetic correlation, similar to the simulated relationship between the most distant populations. In most cases in Duroc populations, GP have higher accuracies in joint populations than in individual population. In conclusion, the consistency of CVES plays a more important role in multi-population GP. The genetic relatedness of the Duroc populations is so weak that the prediction accuracy of GP in joint populations is reduced in some traits. Multi-trait GBLUP is a competitive method for the joint breeding evaluation.

A new method for multiancestry polygenic prediction improves performance across diverse populations

Polygenic risk scores (PRSs) increasingly predict complex traits; however, suboptimal performance in non-European populations raise concerns about clinical applications and health inequities. We developed CT-SLEB, a powerful and scalable method to calculate PRSs, using ancestry-specific genome-wide association study summary statistics from multiancestry training samples, integrating clumping and thresholding, empirical Bayes and superlearning. We evaluated CT-SLEB and nine alternative methods with large-scale simulated genome-wide association studies (~19 million common variants) and datasets from 23andMe, Inc., the Global Lipids Genetics Consortium, All of Us and UK Biobank, involving 5.1 million individuals of diverse ancestry, with 1.18 million individuals from four non-European populations across 13 complex traits. Results demonstrated that CT-SLEB significantly improves PRS performance in non-European populations compared with simple alternatives, with comparable or superior performance to a recent, computationally intensive method. Moreover, our simulation studies offered insights into sample size requirements and SNP density effects on multiancestry risk prediction.

Polygenic scores in cancer

Since the publication of the first genome-wide association study for cancer in 2007, thousands of common alleles that are associated with the risk of cancer have been identified. The relative risk associated with individual variants is small and of limited clinical significance. However, the combined effect of multiple risk variants as captured by polygenic scores (PGSs) may be much greater and therefore provide risk discrimination that is clinically useful. We review the considerable research efforts over the past 15 years for developing statistical methods for PGSs and their application in large-scale genome-wide association studies to develop PGSs for various cancers. We review the predictive performance of these PGSs and the multiple challenges currently limiting the clinical application of PGSs. Despite this, PGSs are beginning to be incorporated into clinical multifactorial risk prediction models to stratify risk in both clinical trials and clinical implementation studies.

Testing for differences in polygenic scores in the presence of confounding

Polygenic scores have become an important tool in human genetics, enabling the prediction of individuals' phenotypes from their genotypes. Understanding how the pattern of differences in polygenic score predictions across individuals intersects with variation in ancestry can provide insights into the evolutionary forces acting on the trait in question, and is important for understanding health disparities. However, because most polygenic scores are computed using effect estimates from population samples, they are susceptible to confounding by both genetic and environmental effects that are correlated with ancestry. The extent to which this confounding drives patterns in the distribution of polygenic scores depends on patterns of population structure in both the original estimation panel and in the prediction/test panel. Here, we use theory from population and statistical genetics, together with simulations, to study the procedure of testing for an association between polygenic scores and axes of ancestry variation in the presence of confounding. We use a general model of genetic relatedness to describe how confounding in the estimation panel biases the distribution of polygenic scores in a way that depends on the degree of overlap in population structure between panels. We then show how this confounding can bias tests for associations between polygenic scores and important axes of ancestry variation in the test panel. Finally, we use the understanding gained from this analysis to develop a method that uses patterns of genetic similarity between the two panels to guard against these biases, and show that this method can provide better protection against confounding than the standard PCA-based approach.

Addressing the Challenge of Biomedical Data Inequality: An Artificial Intelligence Perspective

Artificial intelligence (AI) and other data-driven technologies hold great promise to transform healthcare and confer the predictive power essential to precision medicine. However, the existing biomedical data, which are a vital resource and foundation for developing medical AI models, do not reflect the diversity of the human population. The low representation in biomedical data has become a significant health risk for non-European populations, and the growing application of AI opens a new pathway for this health risk to manifest and amplify. Here we review the current status of biomedical data inequality and present a conceptual framework for understanding its impacts on machine learning. We also discuss the recent advances in algorithmic interventions for mitigating health disparities arising from biomedical data inequality. Finally, we briefly discuss the newly identified disparity in data quality among ethnic groups and its potential impacts on machine learning.

Calibrated prediction intervals for polygenic scores across diverse contexts

Polygenic scores (PGS) have emerged as the tool of choice for genomic prediction in a wide range of fields from agriculture to personalized medicine. We analyze data from two large biobanks in the US (All of Us) and the UK (UK Biobank) to find widespread variability in PGS performance across contexts. Many contexts, including age, sex, and income, impact PGS accuracies with similar magnitudes as genetic ancestry. PGSs trained in single versus multi-ancestry cohorts show similar context-specificity in their accuracies. We introduce trait prediction intervals that are allowed to vary across contexts as a principled approach to account for context-specific PGS accuracy in genomic prediction. We model the impact of all contexts in a joint framework to enable PGS-based trait predictions that are well-calibrated (contain the trait value with 90% probability in all contexts), whereas methods that ignore context are mis-calibrated. We show that prediction intervals need to be adjusted for all considered traits ranging from 10% for diastolic blood pressure to 80% for waist circumference. Adjustment of prediction intervals depends on the dataset; for example, prediction intervals for education years need to be adjusted by 90% in All of Us versus 8% in UK Biobank. Our results provide a path forward towards utilization of PGS as a prediction tool across all individuals regardless of their contexts while highlighting the importance of comprehensive profile of context information in study design and data collection.

Biobank-scale methods and projections for sparse polygenic prediction from machine learning

In this paper we characterize the performance of linear models trained via widely-used sparse machine learning algorithms. We build polygenic scores and examine performance as a function of training set size, genetic ancestral background, and training method. We show that predictor performance is most strongly dependent on size of training data, with smaller gains from algorithmic improvements. We find that LASSO generally performs as well as the best methods, judged by a variety of metrics. We also investigate performance characteristics of predictors trained on one genetic ancestry group when applied to another. Using LASSO, we develop a novel method for projecting AUC and correlation as a function of data size (i.e., for new biobanks) and characterize the asymptotic limit of performance. Additionally, for LASSO (compressed sensing) we show that performance metrics and predictor sparsity are in agreement with theoretical predictions from the Donoho-Tanner phase transition. Specifically, a future predictor trained in the Taiwan Precision Medicine Initiative for asthma can achieve an AUC of [Formula: see text] and for height a correlation of [Formula: see text] for a Taiwanese population. This is above the measured values of [Formula: see text] and [Formula: see text], respectively, for UK Biobank trained predictors applied to a European population.

Bridging the diversity gap: Analytical and study design considerations for improving the accuracy of trans-ancestry genetic prediction

Genetic prediction of common complex disease risk is an essential component of precision medicine. Currently, genome-wide association studies (GWASs) are mostly composed of European-ancestry samples and resulting polygenic scores (PGSs) have been shown to poorly transfer to other ancestries partly due to heterogeneity of allelic effects between populations. Fixed-effects (FETA) and random-effects (RETA) trans-ancestry meta-analyses do not model such ancestry-related heterogeneity, while ancestry-specific (AS) scores may suffer from low power due to low sample sizes. In contrast, trans-ancestry meta-regression (TAMR) builds ancestry-aware PGS that account for more complex trans-ancestry architectures. Here, we examine the predictive performance of these four PGSs under multiple genetic architectures and ancestry configurations. We show that the predictive performance of FETA and RETA is strongly affected by cross-ancestry genetic heterogeneity, while AS PGS performance decreases in under-represented target populations. TAMR PGS is also impacted by heterogeneity but maintains good prediction performance in most situations, especially in ancestry-diverse scenarios. In simulations of human complex traits, TAMR scores currently explain 25% more phenotypic variance than AS in triglyceride levels and 33% more phenotypic variance than FETA in type 2 diabetes in most non-European populations. Importantly, a high proportion of non-European-ancestry individuals is needed to reach prediction levels that are comparable in those populations to the one observed in European-ancestry studies. Our results highlight the need to rebalance the ancestral composition of GWAS to enable accurate prediction in non-European-ancestry groups, and demonstrate the relevance of meta-regression approaches for compensating some of the current population biases in GWAS.

150 risk variants for diverticular disease of intestine prioritize cell types and enable polygenic prediction of disease susceptibility

We conducted a genome-wide association study (GWAS) analysis of diverticular disease (DivD) of intestine within 724,372 individuals and identified 150 independent genome-wide significant DNA variants. Integration of the GWAS results with human gut single-cell RNA sequencing data implicated gut myocyte, mesothelial and stromal cells, and enteric neurons and glia in DivD development. Ninety-five genes were prioritized based on multiple lines of evidence, including SLC9A3, a drug target gene of tenapanor used for the treatment of the constipation subtype of irritable bowel syndrome. A DivD polygenic score (PGS) enables effective risk prediction (area under the curve [AUC], 0.688; 95% confidence interval [CI], 0.645-0.732) and the top 20% PGS was associated with ∼3.6-fold increased DivD risk relative to the remaining population. Our statistical and bioinformatic analyses suggest that the mechanism of DivD is through colon structure, gut motility, gastrointestinal mucus, and ionic homeostasis. Our analyses reinforce the link between gastrointestinal disorders and the enteric nervous system through genetics.

Optimal strategies for learning multi-ancestry polygenic scores vary across traits

Polygenic scores (PGSs) are individual-level measures that aggregate the genome-wide genetic predisposition to a given trait. As PGS have predominantly been developed using European-ancestry samples, trait prediction using such European ancestry-derived PGS is less accurate in non-European ancestry individuals. Although there has been recent progress in combining multiple PGS trained on distinct populations, the problem of how to maximize performance given a multiple-ancestry cohort is largely unexplored. Here, we investigate the effect of sample size and ancestry composition on PGS performance for fifteen traits in UK Biobank. For some traits, PGS estimated using a relatively small African-ancestry training set outperformed, on an African-ancestry test set, PGS estimated using a much larger European-ancestry only training set. We observe similar, but not identical, results when considering other minority-ancestry groups within UK Biobank. Our results emphasise the importance of targeted data collection from underrepresented groups in order to address existing disparities in PGS performance.

Polygenic scoring accuracy varies across the genetic ancestry continuum

Polygenic scores (PGSs) have limited portability across different groupings of individuals (for example, by genetic ancestries and/or social determinants of health), preventing their equitable use1-3. PGS portability has typically been assessed using a single aggregate population-level statistic (for example, R2)4, ignoring inter-individual variation within the population. Here, using a large and diverse Los Angeles biobank5 (ATLAS, n = 36,778) along with the UK Biobank6 (UKBB, n = 487,409), we show that PGS accuracy decreases individual-to-individual along the continuum of genetic ancestries7 in all considered populations, even within traditionally labelled 'homogeneous' genetic ancestries. The decreasing trend is well captured by a continuous measure of genetic distance (GD) from the PGS training data: Pearson correlation of -0.95 between GD and PGS accuracy averaged across 84 traits. When applying PGS models trained on individuals labelled as white British in the UKBB to individuals with European ancestries in ATLAS, individuals in the furthest GD decile have 14% lower accuracy relative to the closest decile; notably, the closest GD decile of individuals with Hispanic Latino American ancestries show similar PGS performance to the furthest GD decile of individuals with European ancestries. GD is significantly correlated with PGS estimates themselves for 82 of 84 traits, further emphasizing the importance of incorporating the continuum of genetic ancestries in PGS interpretation. Our results highlight the need to move away from discrete genetic ancestry clusters towards the continuum of genetic ancestries when considering PGSs.

Distinct genetic liability profiles define clinically relevant patient strata across common diseases

Genome-wide association studies have unearthed a wealth of genetic associations across many complex diseases. However, translating these associations into biological mechanisms contributing to disease etiology and heterogeneity has been challenging. Here, we hypothesize that the effects of disease-associated genetic variants converge onto distinct cell type specific molecular pathways within distinct subgroups of patients. In order to test this hypothesis, we develop the CASTom-iGEx pipeline to operationalize individual level genotype data to interpret personal polygenic risk and identify the genetic basis of clinical heterogeneity. The paradigmatic application of this approach to coronary artery disease and schizophrenia reveals a convergence of disease associated variant effects onto known and novel genes, pathways, and biological processes. The biological process specific genetic liabilities are not equally distributed across patients. Instead, they defined genetically distinct groups of patients, characterized by different profiles across pathways, endophenotypes, and disease severity. These results provide further evidence for a genetic contribution to clinical heterogeneity and point to the existence of partially distinct pathomechanisms across patient subgroups. Thus, the universally applicable approach presented here has the potential to constitute an important component of future personalized medicine concepts.

Amplification is the primary mode of gene-by-sex interaction in complex human traits

Sex differences in complex traits are suspected to be in part due to widespread gene-by-sex interactions (GxSex), but empirical evidence has been elusive. Here, we infer the mixture of ways in which polygenic effects on physiological traits covary between males and females. We find that GxSex is pervasive but acts primarily through systematic sex differences in the magnitude of many genetic effects ("amplification") rather than in the identity of causal variants. Amplification patterns account for sex differences in trait variance. In some cases, testosterone may mediate amplification. Finally, we develop a population-genetic test linking GxSex to contemporary natural selection and find evidence of sexually antagonistic selection on variants affecting testosterone levels. Our results suggest that amplification of polygenic effects is a common mode of GxSex that may contribute to sex differences and fuel their evolution.

Comparing Pruning and Thresholding with Continuous Shrinkage Polygenic Score Methods in a Large Sample of Ancestrally Diverse Adolescents from the ABCD Study®

Using individuals' genetic data researchers can generate Polygenic Scores (PS) that are able to predict risk for diseases, variability in different behaviors as well as anthropomorphic measures. This is achieved by leveraging models learned from previously published large Genome-Wide Association Studies (GWASs) associating locations in the genome with a phenotype of interest. Previous GWASs have predominantly been performed in European ancestry individuals. This is of concern as PS generated in samples with a different ancestry to the original training GWAS have been shown to have lower performance and limited portability, and many efforts are now underway to collect genetic databases on individuals of diverse ancestries. In this study, we compare multiple methods of generating PS, including pruning and thresholding and Bayesian continuous shrinkage models, to determine which of them is best able to overcome these limitations. To do this we use the ABCD Study, a longitudinal cohort with deep phenotyping on individuals of diverse ancestry. We generate PS for anthropometric and psychiatric phenotypes using previously published GWAS summary statistics and examine their performance in three subsamples of ABCD: African ancestry individuals (n = 811), European ancestry Individuals (n = 6703), and admixed ancestry individuals (n = 3664). We find that the single ancestry continuous shrinkage method, PRScs (CS), and the multi ancestry meta method, PRScsx Meta (CSx Meta), show the best performance across ancestries and phenotypes.

Low and differential polygenic score generalizability among African populations due largely to genetic diversity

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 (PRSs) in simulations across Africa and in empirical data from South Africa, Uganda, and the United Kingdom 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 African individuals, 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 United Kingdom 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.

Polygenic Risk Score in African populations: progress and challenges

Polygenic Risk Score (PRS) analysis is a method that predicts the genetic risk of an individual towards targeted traits. Even when there are no significant markers, it gives evidence of a genetic effect beyond the results of Genome-Wide Association Studies (GWAS). Moreover, it selects  single nucleotide polymorphisms (SNPs) that  contribute to the disease with low effect size  making it more precise at individual level risk prediction. PRS  analysis addresses the shortfall of GWAS by taking into account the SNPs/alleles with  low effect size but play an indispensable role to the observed phenotypic/trait variance.  PRS analysis has  applications that investigate the genetic basis of several traits, which includes rare diseases. However, the accuracy of PRS analysis depends on the genomic data of the underlying population. For instance, several studies  show   that obtaining higher prediction power of PRS analysis is challenging for non-Europeans. In this manuscript, we review the conventional PRS methods and their application to sub-Saharan African communities. We conclude that  lack of sufficient GWAS data and tools is  the limiting factor of applying PRS analysis to sub-Saharan populations.   We recommend developing Africa-specific PRS methods and tools for estimating and analyzing  African population data   for clinical  evaluation of PRSs of interest and predicting  rare diseases.

Polygenic Risk Score in African populations: progress and challenges

Polygenic Risk Score (PRS) analysis is a method that predicts the genetic risk of an individual towards targeted traits. Even when there are no significant markers, it gives evidence of a genetic effect beyond the results of Genome-Wide Association Studies (GWAS). Moreover, it selects  single nucleotide polymorphisms (SNPs) that  contribute to the disease with low effect size  making it more precise at individual level risk prediction. PRS  analysis addresses the shortfall of GWAS by taking into account the SNPs/alleles with  low effect size but play an indispensable role to the observed phenotypic/trait variance.  PRS analysis has  applications that investigate the genetic basis of several traits, which includes rare diseases. However, the accuracy of PRS analysis depends on the genomic data of the underlying population. For instance, several studies  show   that obtaining higher prediction power of PRS analysis is challenging for non-Europeans. In this manuscript, we review the conventional PRS methods and their application to sub-Saharan African communities. We conclude that  lack of sufficient GWAS data and tools is  the limiting factor of applying PRS analysis to sub-Saharan populations.   We recommend developing Africa-specific PRS methods and tools for estimating and analyzing  African population data   for clinical  evaluation of PRSs of interest and predicting  rare diseases.

Causal effects on complex traits are similar for common variants across segments of different continental ancestries within admixed individuals

Individuals of admixed ancestries (for example, African Americans) inherit a mosaic of ancestry segments (local ancestry) originating from multiple continental ancestral populations. This offers the unique opportunity of investigating the similarity of genetic effects on traits across ancestries within the same population. Here we introduce an approach to estimate correlation of causal genetic effects (radmix) across local ancestries and analyze 38 complex traits in African-European admixed individuals (N = 53,001) to observe very high correlations (meta-analysis radmix = 0.95, 95% credible interval 0.93-0.97), much higher than correlation of causal effects across continental ancestries. We replicate our results using regression-based methods from marginal genome-wide association study summary statistics. We also report realistic scenarios where regression-based methods yield inflated heterogeneity-by-ancestry due to ancestry-specific tagging of causal effects, and/or polygenicity. Our results motivate genetic analyses that assume minimal heterogeneity in causal effects by ancestry, with implications for the inclusion of ancestry-diverse individuals in studies.

Polygenic risk score prediction of multiple sclerosis in individuals of South Asian ancestry

Polygenic risk scores aggregate an individual's burden of risk alleles to estimate the overall genetic risk for a specific trait or disease. Polygenic risk scores derived from genome-wide association studies of European populations perform poorly for other ancestral groups. Given the potential for future clinical utility, underperformance of polygenic risk scores in South Asian populations has the potential to reinforce health inequalities. To determine whether European-derived polygenic risk scores underperform at multiple sclerosis prediction in a South Asian-ancestry population compared with a European-ancestry cohort, we used data from two longitudinal genetic cohort studies: Genes & Health (2015-present), a study of ∼50 000 British-Bangladeshi and British-Pakistani individuals, and UK Biobank (2006-present), which is comprised of ∼500 000 predominantly White British individuals. We compared individuals with and without multiple sclerosis in both studies (Genes & Health: N Cases = 42, N Control = 40 490; UK Biobank: N Cases = 2091, N Control = 374 866). Polygenic risk scores were calculated using clumping and thresholding with risk allele effect sizes obtained from the largest multiple sclerosis genome-wide association study to date. Scores were calculated with and without the major histocompatibility complex region, the most influential locus in determining multiple sclerosis risk. Polygenic risk score prediction was evaluated using Nagelkerke's pseudo-R 2 metric adjusted for case ascertainment, age, sex and the first four genetic principal components. We found that, as expected, European-derived polygenic risk scores perform poorly in the Genes & Health cohort, explaining 1.1% (including the major histocompatibility complex) and 1.5% (excluding the major histocompatibility complex) of disease risk. In contrast, multiple sclerosis polygenic risk scores explained 4.8% (including the major histocompatibility complex) and 2.8% (excluding the major histocompatibility complex) of disease risk in European-ancestry UK Biobank participants. These findings suggest that polygenic risk score prediction of multiple sclerosis based on European genome-wide association study results is less accurate in a South Asian population. Genetic studies of ancestrally diverse populations are required to ensure that polygenic risk scores can be useful across ancestries.

BridgePRS : A powerful trans-ancestry Polygenic Risk Score method

Polygenic Risk Scores (PRS) have huge potential to contribute to biomedical research and to a future of precision medicine, but to date their calculation relies largely on Europeanancestry GWAS data. This global bias makes most PRS substantially less accurate in individuals of non-European ancestry. Here we present BridgePRS , a novel Bayesian PRS method that leverages shared genetic effects across ancestries to increase the accuracy of PRS in non-European populations. The performance of BridgePRS is evaluated in simulated data and real UK Biobank (UKB) data across 19 traits in African, South Asian and East Asian ancestry individuals, using both UKB and Biobank Japan GWAS summary statistics. BridgePRS is compared to the leading alternative, PRS-CSx , and two single-ancestry PRS methods adapted for trans-ancestry prediction. PRS trained in the UK Biobank are then validated out-of-cohort in the independent Mount Sinai (New York) Bio Me Biobank. Simulations reveal that BridgePRS performance, relative to PRS-CSx , increases as uncertainty increases: with lower heritability, higher polygenicity, greater between-population genetic diversity, and when causal variants are not present in the data. Our simulation results are consistent with real data analyses in which BridgePRS has better predictive accuracy in African ancestry samples, especially in out-of-cohort prediction (into Bio Me ), which shows a 60% boost in mean R 2 compared to PRS-CSx ( P = 2 × 10 -6 ). BridgePRS performs the full PRS analysis pipeline, is computationally efficient, and is a powerful method for deriving PRS in diverse and under-represented ancestry populations.