Analysis of polygenic risk score usage and performance in diverse human populations

Duncan, Laramie; Shen, Hanyang; Gelaye, Bizu; Meijsen, Joeri; Ressler, Kerry J.; Feldman, Marcus W.; Peterson, Roseann E.; Domingue, Benjamin W.

doi:10.1038/s41467-019-11112-0

Cited by 845 publications

(818 citation statements)

References 51 publications

Supporting

Mentioning

772

Contrasting

Unclassified

Order By: Relevance

“…One major limitation is that the majority of participants in GWAS used to derive PRS are of European ancestry 8,9 . Although many genome-wide significant GWAS hits do replicate in non-European ancestry cohorts 10–14 , overall the predictive power of PRS is lower and decreases with genetic distance from Europeans 15–18 . There are many possible explanations for this, including inter-cohort differences in data collection, phenotype or environment, differences in linkage disequilibrium (LD) structure or allele frequencies across populations 19 , differences in causal or marginal effect sizes, and epistatic or gene-environment interactions 20 .…”

Section: Introductionmentioning

confidence: 99%

Polygenic scores for height in admixed populations

Bitarello

Mathieson

2020

Preprint

View full text Add to dashboard Cite

11Polygenic risk scores (PRS) use the results of genome-wide association studies (GWAS) to predict quantitative 12 phenotypes or disease risk at an individual level. This provides a potential route to the use of genetic data in 13 personalized medical care. However, a major barrier to the use of PRS is that the majority of GWAS come from 14 cohorts of European ancestry. The predictive power of PRS constructed from these studies is substantially 15 lower in non-European ancestry cohorts, although the reasons for this are unclear. To address this question, 16we investigate the performance of PRS for height in cohorts with admixed African and European ancestry, 17 allowing us to evaluate ancestry-related differences in PRS predictive accuracy while controlling for 18 environment and cohort differences. We first show that that the predictive accuracy of height PRS increases 19 linearly with European ancestry and is largely explained by European ancestry segments of the admixed 20 genomes. We show that differences in allele frequencies, recombination rate, and marginal effect sizes across 21 ancestries all contribute to the decrease in predictive power, but none of these effects explain the decrease on 22 its own. Finally, we demonstrate that prediction for admixed individuals can be improved by using a linear 23 combination of PRS that includes ancestry-specific effect sizes, although this approach is at present limited by 24 the small size of non-European ancestry discovery cohorts. 60we lifted over SNP positions to hg19 using liftOver. For WHI, JHS and HRS, we flipped alleles to the positive 61 strand using the appropriate strand files from https://www.well.ox.ac.uk/~wrayner/strand/. We identified 62 individuals with admixed African ancestry in each cohort using a combination of genetic clustering and self-63 reported ancestry as follows: 64 65 3 UKB: This dataset contains several ancestry groups. We selected 8,813 individuals with African or admixed 66 African and European ancestry based on PCA ( Figure S1) and refer to them as UKB_afr. We further filtered this 67 set to contain individuals with at least 5% of African ancestry, resulting in 8,700 individuals ( Table 1). We 68 randomly selected 9,998 European ancestry individuals from the "White British" subset to use as a comparison 69 sample and refer to them as "UKB_eur". 71WHI: This dataset contains both African American and Hispanic participants. We ran unsupervised 72 ADMIXTURE 27 with k=3 and identified 7,285 individuals with self-reported "African American" ancestry with at 73 most 0.8 of the first ADMIXTURE component (which we interpret as reflecting European ancestry), and at most 74 0.05 of the second (which we interpret as reflecting Native American ancestry; Figure S2). We further filtered 75 this set to contain individuals with at least 5% of African ancestry and height between ±2 standard deviations 76 (sd) from the mean (see Figure S12), resulting in 6,863 individuals ( Table 1). We refer to them as "WHI_afr". 77 78 HRS: This dataset contains multip...

show abstract

Section: Introductionmentioning

confidence: 99%

Polygenic scores for height in admixed populations

Bitarello

Mathieson

2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Polygenic risk scores (PRSs) have successfully shown their predictive ability to idenitify those with a several-fold higher inherited risk of a given disease or condition 2 . Both an increase in statistical power and ethnic diversity in genetic studies—accelerated by nation-wide biobanks—have been instrumental in accurately predicting disease onset by PRSs 3–6 . Stratification of health risks based on PRSs would be one of the strategies to improve population health through targeted prevention.…”

Section: Introductionmentioning

confidence: 99%

Trans-biobank analysis with 676,000 individuals elucidates the association of polygenic risk scores of complex traits with human lifespan

Sakaue

Kanai

Karjalainen

et al. 2019

Preprint

View full text Add to dashboard Cite

50Human genetics seeks a way to improve human health on a global scale. Expectations are 51 running high for polygenic risk scores (PRSs) to be translated into clinical practice to predict 52 an inborn susceptibility to health risks. While risk stratification based on PRS is one way to 53 promote population health, a strategy to utilize genetics to prioritize modifiable risk factors 54 and biomarkers driving heath outcome is also warranted. To this end, here we utilized PRSs 55 to comprehensively investigate the association of the genetic susceptibility to complex traits 56 with human lifespan in collaboration with three worldwide biobanks (ntotal = 675,898). First, 57we conducted genome-wide association studies for 45 quantitative clinical phenotypes, 58 constructed the individual PRSs, and associated them with the age at death of 179,066 59 participants in BioBank Japan. The PRSs revealed that the genetic susceptibility of high 60 systolic blood pressure (sBP) was strongly associated with a shorter lifespan (hazard ratio 61[HR] = 1.03, P = 1.4×10 -7 ). Next, we sought to replicate these associations in individuals of 62European ancestry in UK Biobank (n = 361,194) and FinnGen (n = 135,638). Among the 63 investigated traits, the individuals with higher blood pressure-related PRSs were trans-64 ethnically associated with a shorter lifespan (HR = 1.03, Pmeta = 3.9×10 -13 for sBP) and 65 parental lifespan (HR = 1.06, PUKBB = 2.0×10 -86 for sBP). Further, our trans-biobank study 66 identified additional complex traits associated with lifespan (e.g., obesity, height, serum lipids, 67 and platelet counts). Of them, obesity-related traits showed strikingly heterogeneous effects 68 on lifespan between Japanese and European populations (Pheterogeneity = 9.5×10 -8 for body 69 mass index). Through trans-ethnic biobank collaboration, we elucidated the novel value of 70 the PRS study in genetics-driven prioritization of risk factors and biomarkers which can be 71 medically intervened to improve population health. 72 73 5 Main 74 differences among populations, trans-ethnic comparison is also warranted. A collaboration 108 with three trans-ethnic nation-wide biobanks collecting genotype, phenotype and survival 109 data (ntotal = 675,898) has enabled us to uncover the modifiable risk factors and monitorable 110 biomarkers affecting human lifespan across the populations, on an unprecedented scale and 111 without any clinical intervention. 112 113 7 Results 114 Study overview 115

show abstract

“…Common variant gene scoring approaches have limited trans-ancestral predictive value and clinical implications remain highest when discovery and validation sets are ethnically matched, which is largely the case only for European populations 18,19,23 . Moreover, empirical and simulation models have demonstrated that directional selection across ancestries can severely impair predictive accuracy in non-European validation populations, even after accounting for principal components and proportion of causal variants 18,19 .…”

Section: Discussionmentioning

confidence: 99%

Calibrated rare variant genetic risk scores for complex disease prediction using large exome sequence repositories

Lali

Chong

Omidi

et al. 2020

Preprint

View full text Add to dashboard Cite

Rare variants are collectively numerous and may underlie a considerable proportion of complex disease risk. However, identifying genuine rare variant associations is challenging due to small effect sizes, presence of technical artefacts, and heterogeneity in population structure. We hypothesized that rare variant burden over a large number of genes can be combined into predictive rare variant genetic risk score (RVGRS). We propose a novel method (RV-EXCALIBER) that leverages summary-level data from a large public exome sequencing database (gnomAD) as controls and robustly calibrates rare variant burden to account for the aforementioned biases. A RVGRS was found to strongly associate with coronary artery disease (CAD) in European and South Asian populations. Calibrated RVGRS capture the aggregate effect of rare variants through a polygenic model of inheritance, identifies 1.5% of the population with substantial risk of early CAD, and confers risk even when adjusting for known Mendelian CAD genes, clinical risk factors, and common variant gene scores.

show abstract

Analysis of polygenic risk score usage and performance in diverse human populations

Cited by 845 publications

References 51 publications

Polygenic scores for height in admixed populations

Polygenic scores for height in admixed populations

Trans-biobank analysis with 676,000 individuals elucidates the association of polygenic risk scores of complex traits with human lifespan

Calibrated rare variant genetic risk scores for complex disease prediction using large exome sequence repositories

Contact Info

Product

Resources

About