Modeling Linkage Disequilibrium Increases Accuracy of Polygenic Risk Scores

Deciphering Sex-Specific Genetic Architectures Using Local Bayesian Regressions

Many complex human traits exhibit differences between sexes. While numerous factors likely contribute to this phenomenon, growing evidence from genome-wide studies suggest a partial explanation: that males and females from the same population possess differing genetic architectures. Despite this, mapping gene-by-sex (G×S) interactions remains a challenge likely because the magnitude of such an interaction is typically and exceedingly small; traditional genome-wide association techniques may be underpowered to detect such events, due partly to the burden of multiple test correction. Here, we developed a local Bayesian regression (LBR) method to estimate sex-specific SNP marker effects after fully accounting for local linkage-disequilibrium (LD) patterns. This enabled us to infer sex-specific effects and G×S interactions either at the single SNP level, or by aggregating the effects of multiple SNPs to make inferences at the level of small LD-based regions. Using simulations in which there was imperfect LD between SNPs and causal variants, we showed that aggregating sex-specific marker effects with LBR provides improved power and resolution to detect G×S interactions over traditional single-SNP-based tests. When using LBR to analyze traits from the UK Biobank, we detected a relatively large G×S interaction impacting bone mineral density within ABO, and replicated many previously detected large-magnitude G×S interactions impacting waist-to-hip ratio. We also discovered many new G×S interactions impacting such traits as height and body mass index (BMI) within regions of the genome where both male- and female-specific effects explain a small proportion of phenotypic variance (R2 < 1 × 10-4), but are enriched in known expression quantitative trait loci.

Pleiotropy of polygenic factors associated with focal and generalized epilepsy in the general population

Epilepsy is clinically heterogeneous, and neurological or psychiatric comorbidities are frequently observed in patients. It has not been tested whether common risk variants for generalized or focal epilepsy are enriched in people with other disorders or traits related to brain or cognitive function. Here, we perform two brain-focused phenome association studies of polygenic risk scores (PRS) for generalized epilepsy (GE-PRS) or focal epilepsy (FE-PRS) with all binary brain or cognitive function-related traits available for 334,310 European-ancestry individuals of the UK Biobank. Higher GE-PRS were associated with not having a college or university degree (P = 3.00x10-4), five neuroticism-related personality traits (P<2.51x10-4), and having ever smoked (P = 1.27x10-6). Higher FE-PRS were associated with several measures of low educational attainment (P<4.87x10-5), one neuroticism-related personality trait (P = 2.33x10-4), having ever smoked (P = 1.71x10-4), and having experienced events of anxiety or depression (P = 2.83x10-4). GE- and FE-PRS had the same direction of effect for each of the associated traits. Genetic factors associated with GE or FE showed similar patterns of correlation with genetic factors associated with cortical morphology in a subset of the UKB with 16,612 individuals and T1 magnetic resonance imaging data. In summary, our results suggest that genetic factors associated with epilepsies may confer risk for other neurological and psychiatric disorders in a population sample not enriched for epilepsy.

Polygenic Scores to Assess Atherosclerotic Cardiovascular Disease Risk: Clinical Perspectives and Basic Implications

An individual's susceptibility to atherosclerotic cardiovascular disease is influenced by numerous clinical and lifestyle factors, motivating the multifaceted approaches currently endorsed for primary and secondary cardiovascular disease prevention. With growing knowledge of the genetic basis of atherosclerotic cardiovascular disease-in particular, coronary artery disease-and its contribution to disease pathogenesis, there is increased interest in understanding the potential clinical utility of a genetic predictor that might further refine the assessment and management of atherosclerotic cardiovascular disease risk. Rapid scientific and technological advances have enabled widespread genotyping efforts and dynamic research in the field of coronary artery disease genetic risk prediction. In this review, we describe how genomic analyses of coronary artery disease have been leveraged to create polygenic risk scores. We then discuss evaluations of the clinical utility of these scores, pertinent mechanistic insights gleaned, and practical considerations relevant to the implementation of polygenic risk scores in the health care setting.

Genetic Vulnerability for Smoking and Cannabis Use: Associations With E-Cigarette and Water Pipe Use

INTRODUCTION

Cigarette smoking and cannabis use are heritable traits and share, at least in part, a common genetic substrate. In recent years, the prevalence of alternative methods of nicotine intakes, such as electronic cigarette (e-cigarette) and water pipe use, has risen substantially. We tested whether the genetic vulnerability underlying cigarettes smoking and cannabis use explained variability in e-cigarette and water pipe use phenotypes, as these vaping methods are alternatives for smoking tobacco cigarettes and joints.

METHODS

On the basis of the summary statistics of the International Cannabis Consortium and the Tobacco and Genetics Consortium, we generated polygenic risk scores (PRSs) for smoking and cannabis use traits, and used these to predict e-cigarette and water pipe use phenotypes in a sample of 5025 individuals from the Netherlands Twin Register.

RESULTS

PRSs for cigarettes per day were positively associated with lifetime e-cigarette use and early initiation of water pipe use, but only in ex-smokers (odds ratio = 1.43, R2 = 1.56%, p = .011) and never cigarette smokers (odds ratio = 1.35, R2 = 1.60%, p = .013) respectively.

CONCLUSIONS

Most associations of PRSs for cigarette smoking and cannabis use with e-cigarette and water pipe use were not significant, potentially due to a lack of power. The significant associations between genetic liability to smoking heaviness with e-cigarette and water pipe phenotypes are in line with studies indicating a common genetic background for substance-use phenotypes. These associations emerged only in nonsmokers, and future studies should investigate the nature of this observation.

IMPLICATIONS

Our study showed that genetic vulnerability to smoking heaviness is associated with lifetime e-cigarette use and age at initiation of water pipe use. This finding has implications for the current debate on whether alternative smoking methods, such as usage of vaping devices, predispose to smoking initiation and related behaviors.

Pathway-based polygene risk for severe depression implicates drug metabolism in CONVERGE

BACKGROUND

The Psychiatric Genomics Consortium (PGC) has made major advances in the molecular etiology of MDD, confirming that MDD is highly polygenic. Pathway enrichment results from PGC meta-analyses can also be used to help inform molecular drug targets. Prior to any knowledge of molecular biomarkers for MDD, drugs targeting molecular pathways (MPs) proved successful in treating MDD. It is possible that examining polygenicity within specific MPs implicated in MDD can further refine molecular drug targets.

METHODS

Using a large case-control GWAS based on low-coverage whole genome sequencing (N = 10 640) in Han Chinese women, we derived polygenic risk scores (PRS) for MDD and for MDD specific to each of over 300 MPs previously shown to be relevant to psychiatric diagnoses. We then identified sets of PRSs, accounting for critical covariates, significantly predictive of case status.

RESULTS

Over and above global MDD polygenic risk, polygenic risk within the GO: 0017144 drug metabolism pathway significantly predicted recurrent depression after multiple testing correction. Secondary transcriptomic analysis suggests that among genes in this pathway, CYP2C19 (family of Cytochrome P450) and CBR1 (Carbonyl Reductase 1) might be most relevant to MDD. Within the cases, pathway-based risk was additionally associated with age at onset of MDD.

CONCLUSIONS

Results indicate that pathway-based risk might inform etiology of recurrent major depression. Future research should examine whether polygenicity of the drug metabolism gene pathway has any association with clinical presentation or treatment response. We discuss limitations to the generalizability of these preliminary findings, and urge replication in future research.

Substance use: Interplay between polygenic risk and neighborhood environment

BACKGROUND

Tobacco, alcohol, and cannabis use are prevalent behaviors that pose considerable health risks. Genetic vulnerability and characteristics of the neighborhood of residence form important risk factors for substance use. Possibly, these factors do not act in isolation. This study tested the interaction between neighborhood characteristics and genetic risk (gene-environment interaction, GxE) and the association between these classes of risk factors (gene-environment correlation, rGE) in substance use.

METHODS

Two polygenic scores (PGS) each (based on different discovery datasets) were created for smoking initiation, cigarettes per day, and glasses of alcohol per week based on summary statistics of different genome-wide association studies (GWAS). For cannabis initiation one PGS was created. These PGS were used to predict their respective phenotype in a large population-based sample from the Netherlands Twin Register (N = 6,567). Neighborhood characteristics as retrieved from governmental registration systems were factor analyzed and resulting measures of socioeconomic status (SES) and metropolitanism were used as predictors.

RESULTS

There were (small) main effects of neighborhood characteristics and PGS on substance use. One of the 14 tested GxE effects was significant, such that the PGS was more strongly associated with alcohol use in individuals with high SES. This was effect was only significant for one out of two PGS. There were weak indications of rGE, mainly with age and cohort covariates.

CONCLUSION

We conclude that both genetic and neighborhood-level factors are predictors for substance use. More research is needed to establish the robustness of the findings on the interplay between these factors.

A gene-diet interaction-based score predicts response to dietary fat in the Women's Health Initiative

BACKGROUND

Although diet response prediction for cardiometabolic risk factors (CRFs) has been demonstrated using single genetic variants and main-effect genetic risk scores, little investigation has gone into the development of genome-wide diet response scores.

OBJECTIVE

We sought to leverage the multistudy setup of the Women's Health Initiative cohort to generate and test genetic scores for the response of 6 CRFs (BMI, systolic blood pressure, LDL cholesterol, HDL cholesterol, triglycerides, and fasting glucose) to dietary fat.

METHODS

A genome-wide interaction study was undertaken for each CRF in women (n ∼ 9000) not participating in the dietary modification (DM) trial, which focused on the reduction of dietary fat. Genetic scores based on these analyses were developed using a pruning-and-thresholding approach and tested for the prediction of 1-y CRF changes as well as long-term chronic disease development in DM trial participants (n ∼ 5000).

RESULTS

Only 1 of these genetic scores, for LDL cholesterol, predicted changes in the associated CRF. This 1760-variant score explained 3.7% (95% CI: 0.09, 11.9) of the variance in 1-y LDL cholesterol changes in the intervention arm but was unassociated with changes in the control arm. In contrast, a main-effect genetic risk score for LDL cholesterol was not useful for predicting dietary fat response. Further investigation of this score with respect to downstream disease outcomes revealed suggestive differential associations across DM trial arms, especially with respect to coronary heart disease and stroke subtypes.

CONCLUSIONS

These results lay the foundation for the combination of many genome-wide gene-diet interactions for diet response prediction while highlighting the need for further research and larger samples in order to achieve robust biomarkers for use in personalized nutrition.

Efficient Estimation and Applications of Cross-Validated Genetic Predictions to Polygenic Risk Scores and Linear Mixed Models

Large-scale cohorts with combined genetic and phenotypic data, coupled with methodological advances, have produced increasingly accurate genetic predictors of complex human phenotypes called polygenic risk scores (PRSs). In addition to the potential translational impacts of identifying at-risk individuals, PRS are being utilized for a growing list of scientific applications, including causal inference, identifying pleiotropy and genetic correlation, and powerful gene-based and mixed-model association tests. Existing PRS approaches rely on external large-scale genetic cohorts that have also measured the phenotype of interest. They further require matching on ancestry and genotyping platform or imputation quality. In this work, we present a novel reference-free method to produce a PRS that does not rely on an external cohort. We show that naive implementations of reference-free PRS either result in substantial overfitting or prohibitive increases in computational time. We show that our algorithm avoids both of these issues and can produce informative in-sample PRSs over a single cohort without overfitting. We then demonstrate several novel applications of reference-free PRSs, including detection of pleiotropy across 246 metabolic traits and efficient mixed-model association testing.

Genetic Endowments and Wealth Inequality

We show that genetic endowments linked to educational attainment strongly and robustly predict wealth at retirement. The estimated relationship is not fully explained by flexibly controlling for education and labor income. We therefore investigate a host of additional mechanisms that could account for the gene-wealth gradient, including inheritances, mortality, risk preferences, portfolio decisions, beliefs about the probabilities of macroeconomic events, and planning horizons. We provide evidence that genetic endowments related to human capital accumulation are associated with wealth not only through educational attainment and labor income, but also through a facility with complex financial decision-making.

Electronic health records and polygenic risk scores for predicting disease risk

Accurate prediction of disease risk based on the genetic make-up of an individual is essential for effective prevention and personalized treatment. Nevertheless, to date, individual genetic variants from genome-wide association studies have achieved only moderate prediction of disease risk. The aggregation of genetic variants under a polygenic model shows promising improvements in prediction accuracies. Increasingly, electronic health records (EHRs) are being linked to patient genetic data in biobanks, which provides new opportunities for developing and applying polygenic risk scores in the clinic, to systematically examine and evaluate patient susceptibilities to disease. However, the heterogeneous nature of EHR data brings forth many practical challenges along every step of designing and implementing risk prediction strategies. In this Review, we present the unique considerations for using genotype and phenotype data from biobank-linked EHRs for polygenic risk prediction.

Genome-wide association identifies seven loci for pelvic organ prolapse in Iceland and the UK Biobank

Pelvic organ prolapse (POP) is a downward descent of one or more of the pelvic organs, resulting in a protrusion of the vaginal wall and/or uterus. We performed a genome-wide association study of POP using data from Iceland and the UK Biobank, a total of 15,010 cases with hospital-based diagnosis code and 340,734 female controls, and found eight sequence variants at seven loci associating with POP (P < 5 × 10-8); seven common (minor allele frequency >5%) and one with minor allele frequency of 4.87%. Some of the variants associating with POP also associated with traits of similar pathophysiology. Of these, rs3820282, which may alter the estrogen-based regulation of WNT4, also associates with leiomyoma of uterus, gestational duration and endometriosis. Rs3791675 at EFEMP1, a gene involved in connective tissue homeostasis, also associates with hernias and carpal tunnel syndrome. Our results highlight the role of connective tissue metabolism and estrogen exposure in the etiology of POP.

The emerging landscape of health research based on biobanks linked to electronic health records: Existing resources, statistical challenges, and potential opportunities

Biobanks linked to electronic health records provide rich resources for health-related research. With improvements in administrative and informatics infrastructure, the availability and utility of data from biobanks have dramatically increased. In this paper, we first aim to characterize the current landscape of available biobanks and to describe specific biobanks, including their place of origin, size, and data types. The development and accessibility of large-scale biorepositories provide the opportunity to accelerate agnostic searches, expedite discoveries, and conduct hypothesis-generating studies of disease-treatment, disease-exposure, and disease-gene associations. Rather than designing and implementing a single study focused on a few targeted hypotheses, researchers can potentially use biobanks' existing resources to answer an expanded selection of exploratory questions as quickly as they can analyze them. However, there are many obvious and subtle challenges with the design and analysis of biobank-based studies. Our second aim is to discuss statistical issues related to biobank research such as study design, sampling strategy, phenotype identification, and missing data. We focus our discussion on biobanks that are linked to electronic health records. Some of the analytic issues are illustrated using data from the Michigan Genomics Initiative and UK Biobank, two biobanks with two different recruitment mechanisms. We summarize the current body of literature for addressing these challenges and discuss some standing open problems. This work complements and extends recent reviews about biobank-based research and serves as a resource catalog with analytical and practical guidance for statisticians, epidemiologists, and other medical researchers pursuing research using biobanks.

Polygenic risk scores in coronary artery disease

PURPOSE OF REVIEW

Large genome-wide association studies (GWAS) have identified variants accounting for a substantial portion of the heritable risk for coronary artery disease (CAD). These studies have catalyzed drug discovery and generated the possibility of improved risk prediction and stratification. Here, we review the current state-of-the art in polygenic risk scores (PRSs) and look to the future, as these scores move towards clinical application.

RECENT FINDINGS

Over the last decade, multilocus PRSs for CAD have expanded to include millions of variants and demonstrated strong association with CAD outcomes, even when adjusted for traditional risk factors. Recently, PRSs have shown better prediction of CAD outcomes than any single traditional risk factor alone. Advances in statistical methods used to generate PRSs have improved their predictive ability and transferability between populations with varied ancestries. Initial clinical studies have also demonstrated the potential of genetic information to impact shared decision-making between patients and providers, leading to improved outcomes.

SUMMARY

PRSs can improve risk stratification for CAD especially in white/European populations and have the potential to alter routine clinical care. However, unlocking this potential will require additional research in PRSs in nonwhite populations and substantial investment in clinical implementation studies.

Assessing thyroid cancer risk using polygenic risk scores

Genome-wide association studies (GWASs) have identified at least 10 single-nucleotide polymorphisms (SNPs) associated with papillary thyroid cancer (PTC) risk. Most of these SNPs are common variants with small to moderate effect sizes. Here we assessed the combined genetic effects of these variants on PTC risk by using summarized GWAS results to build polygenic risk score (PRS) models in three PTC study groups from Ohio (1,544 patients and 1,593 controls), Iceland (723 patients and 129,556 controls), and the United Kingdom (534 patients and 407,945 controls). A PRS based on the 10 established PTC SNPs showed a stronger predictive power compared with the clinical factors model, with a minimum increase of area under the receiver-operating curve of 5.4 percentage points (P ≤ 1.0 × 10^-9). Adding an extended PRS based on 592,475 common variants did not significantly improve the prediction power compared with the 10-SNP model, suggesting that most of the remaining undiscovered genetic risk in thyroid cancer is due to rare, moderate- to high-penetrance variants rather than to common low-penetrance variants. Based on the 10-SNP PRS, individuals in the top decile group of PRSs have a close to sevenfold greater risk (95% CI, 5.4-8.8) compared with the bottom decile group. In conclusion, PRSs based on a small number of common germline variants emphasize the importance of heritable low-penetrance markers in PTC.

Genome-Wide Polygenic Score and the Risk of Ischemic Stroke in a Prospective Cohort

Background and Purpose—

Environmental and genetic factors contribute to the development of ischemic stroke (IS). We recently developed a genome-wide polygenic risk score (PRS) for IS using case-control datasets from 4 large-scale observational studies conducted in Japan. Our objective in the present study was to confirm the association between the PRS and the risk of IS with data from an independent prospective cohort recruited from the general Japanese population.

Methods—

A total of 3038 subjects aged ≥40 years were followed up for 10 years (2002–2012). The genome-wide PRS was calculated using genotype data from >350 000 single-nucleotide polymorphisms. The PRS levels were divided into quintiles. High and low genetic risk groups were defined as top 60% and bottom 40% of PRS, respectively. The hazard ratio (HR) for the development of IS was estimated using a Cox proportional hazards model.

Results—

During the follow-up period, 91 cases developed first-ever IS. The age- and sex-adjusted HR for IS increased with higher PRS levels ( P for trend, 0.03). Subjects with the highest quintile level of PRS had a 2.44-fold (95% CI, 1.16–5.12) greater risk for IS than those with the lowest quintile level after adjusting for age and sex. A similar association was observed after adjusting for environmental risk factors ( P for trend, 0.03). As compared with low genetic risk group, the age- and sex-adjusted HR in high genetic risk group was 1.63 (95% CI, 1.04–2.55), which was comparable to the HR of hypertension (HR, 1.41), diabetes mellitus (HR, 1.72), and smoking (HR, 1.54). The age- and sex-adjusted HR increased with the number of environmental risk factors in both high and low genetic risk groups without significant interaction.

Conclusions—

A high genome-wide PRS was a significant risk factor for IS independent of environmental risk factors in a general Japanese population. This finding suggests that PRS may be useful to identify individuals at a high risk of IS.

Measurement and genetic architecture of lifetime depression in the Netherlands as assessed by LIDAS (Lifetime Depression Assessment Self-report)

BACKGROUND

Major depressive disorder (MDD) is a common mood disorder, with a heritability of around 34%. Molecular genetic studies made significant progress and identified genetic markers associated with the risk of MDD; however, progress is slowed down by substantial heterogeneity as MDD is assessed differently across international cohorts. Here, we used a standardized online approach to measure MDD in multiple cohorts in the Netherlands and evaluated whether this approach can be used in epidemiological and genetic association studies of depression.

METHODS

Within the Biobank Netherlands Internet Collaboration (BIONIC) project, we collected MDD data in eight cohorts involving 31 936 participants, using the online Lifetime Depression Assessment Self-report (LIDAS), and estimated the prevalence of current and lifetime MDD in 22 623 unrelated individuals. In a large Netherlands Twin Register (NTR) twin-family dataset (n ≈ 18 000), we estimated the heritability of MDD, and the prediction of MDD in a subset (n = 4782) through Polygenic Risk Score (PRS).

RESULTS

Estimates of current and lifetime MDD prevalence were 6.7% and 18.1%, respectively, in line with population estimates based on validated psychiatric interviews. In the NTR heritability estimates were 0.34/0.30 (s.e. = 0.02/0.02) for current/lifetime MDD, respectively, showing that the LIDAS gives similar heritability rates for MDD as reported in the literature. The PRS predicted risk of MDD (OR 1.23, 95% CI 1.15-1.32, R2 = 1.47%).

CONCLUSIONS

By assessing MDD status in the Netherlands using the LIDAS instrument, we were able to confirm previously reported MDD prevalence and heritability estimates, which suggests that this instrument can be used in epidemiological and genetic association studies of depression.

Predictive Accuracy of a Polygenic Risk Score-Enhanced Prediction Model vs a Clinical Risk Score for Coronary Artery Disease

Importance

The incremental value of polygenic risk scores in addition to well-established risk prediction models for coronary artery disease (CAD) is uncertain.

Objective

To examine whether a polygenic risk score for CAD improves risk prediction beyond pooled cohort equations.

Design, Setting, and Participants

Observational study of UK Biobank participants enrolled from 2006 to 2010. A case-control sample of 15 947 prevalent CAD cases and equal number of age and sex frequency-matched controls was used to optimize the predictive performance of a polygenic risk score for CAD based on summary statistics from published genome-wide association studies. A separate cohort of 352 660 individuals (with follow-up to 2017) was used to evaluate the predictive accuracy of the polygenic risk score, pooled cohort equations, and both combined for incident CAD.

Exposures

Polygenic risk score for CAD, pooled cohort equations, and both combined.

Main Outcomes and Measures

CAD (myocardial infarction and its related sequelae). Discrimination, calibration, and reclassification using a risk threshold of 7.5% were assessed.

Results

In the cohort of 352 660 participants (mean age, 55.9 years; 205 297 women [58.2%]) used to evaluate the predictive accuracy of the examined models, there were 6272 incident CAD events over a median of 8 years of follow-up. CAD discrimination for polygenic risk score, pooled cohort equations, and both combined resulted in C statistics of 0.61 (95% CI, 0.60 to 0.62), 0.76 (95% CI, 0.75 to 0.77), and 0.78 (95% CI, 0.77 to 0.79), respectively. The change in C statistic between the latter 2 models was 0.02 (95% CI, 0.01 to 0.03). Calibration of the models showed overestimation of risk by pooled cohort equations, which was corrected after recalibration. Using a risk threshold of 7.5%, addition of the polygenic risk score to pooled cohort equations resulted in a net reclassification improvement of 4.4% (95% CI, 3.5% to 5.3%) for cases and -0.4% (95% CI, -0.5% to -0.4%) for noncases (overall net reclassification improvement, 4.0% [95% CI, 3.1% to 4.9%]).

Conclusions and Relevance

The addition of a polygenic risk score for CAD to pooled cohort equations was associated with a statistically significant, yet modest, improvement in the predictive accuracy for incident CAD and improved risk stratification for only a small proportion of individuals. The use of genetic information over the pooled cohort equations model warrants further investigation before clinical implementation.

Predictive Accuracy of a Polygenic Risk Score Compared With a Clinical Risk Score for Incident Coronary Heart Disease

IMPORTANCE

Polygenic risk scores comprising millions of single-nucleotide polymorphisms (SNPs) could be useful for population-wide coronary heart disease (CHD) screening.

OBJECTIVE

To determine whether a polygenic risk score improves prediction of CHD compared with a guideline-recommended clinical risk equation.

DESIGN, SETTING, AND PARTICIPANTS

A retrospective cohort study of the predictive accuracy of a previously validated polygenic risk score was assessed among 4847 adults of white European ancestry, aged 45 through 79 years, participating in the Atherosclerosis Risk in Communities (ARIC) study and 2390 participating in the Multi-Ethnic Study of Atherosclerosis (MESA) from 1996 through December 31, 2015, the final day of follow-up. The performance of the polygenic risk score was compared with that of the 2013 American College of Cardiology and American Heart Association pooled cohort equations.

EXPOSURES

Genetic risk was computed for each participant by summing the product of the weights and allele dosage across 6 630 149 SNPs. Weights were based on an international genome-wide association study.

MAIN OUTCOMES AND MEASURES

Prediction of 10-year first CHD events (including myocardial infarctions, fatal coronary events, silent infarctions, revascularization procedures, or resuscitated cardiac arrest) assessed using measures of model discrimination, calibration, and net reclassification improvement (NRI).

RESULTS

The study population included 4847 adults from the ARIC study (mean [SD] age, 62.9 [5.6] years; 56.4% women) and 2390 adults from the MESA cohort (mean [SD] age, 61.8 [9.6] years; 52.2% women). Incident CHD events occurred in 696 participants (14.4%) and 227 participants (9.5%), respectively, over median follow-up of 15.5 years (interquartile range [IQR], 6.3 years) and 14.2 (IQR, 2.5 years) years. The polygenic risk score was significantly associated with 10-year CHD incidence in ARIC with hazard ratios per SD increment of 1.24 (95% CI, 1.15 to 1.34) and in MESA, 1.38 (95% CI, 1.21 to 1.58). Addition of the polygenic risk score to the pooled cohort equations did not significantly increase the C statistic in either cohort (ARIC, change in C statistic, -0.001; 95% CI, -0.009 to 0.006; MESA, 0.021; 95% CI, -0.0004 to 0.043). At the 10-year risk threshold of 7.5%, the addition of the polygenic risk score to the pooled cohort equations did not provide significant improvement in reclassification in either ARIC (NRI, 0.018, 95% CI, -0.012 to 0.036) or MESA (NRI, 0.001, 95% CI, -0.038 to 0.076). The polygenic risk score did not significantly improve calibration in either cohort.

CONCLUSIONS AND RELEVANCE

In this analysis of 2 cohorts of US adults, the polygenic risk score was associated with incident coronary heart disease events but did not significantly improve discrimination, calibration, or risk reclassification compared with conventional predictors. These findings suggest that a polygenic risk score may not enhance risk prediction in a general, white middle-aged population.

Mathematical Properties of Linkage Disequilibrium Statistics Defined by Normalization of the Coefficient D = pAB - pApB

BACKGROUND

Many statistics for measuring linkage disequilibrium (LD) take the form of a normalization of the LD coefficient D. Different normalizations produce statistics with different ranges, interpretations, and arguments favoring their use.

METHODS

Here, to compare the mathematical properties of these normalizations, we consider 5 of these normalized statistics, describing their upper bounds, the mean values of their maxima over the set of possible allele frequency pairs, and the size of the allele frequency regions accessible given specified values of the statistics.

RESULTS

We produce detailed characterizations of these properties for the statistics d and ρ, analogous to computations previously performed for r2. We examine the relationships among the statistics, uncovering conditions under which some of them have close connections.

CONCLUSION

The results contribute insight into LD measurement, particularly the understanding of differences in the features of different LD measures when computed on the same data.

Polygenic score for neuroticism is related to sleep difficulties

Neuroticism, a broad trait measure of the tendency to experience negative emotions and vulnerability to stress, is consistently related to poor sleep quality. Less is known about potential pleiotropy in the genetic risk for high neuroticism and poor sleep. Therefore, the present study examined whether polygenic score (PGS) for neuroticism is related to sleep quality in two large samples of adults. In addition, depressive symptoms, anxiety and phenotypical neuroticism were tested as mediators in both samples. Participants were 8316 individuals aged from 50 to 101 years (mean age = 68.29, SD = 9.83) from the Health and Retirement Study, and 4973 individuals aged from 63 to 67 years (mean age = 64.30, SD = 0.68) from the Wisconsin Longitudinal Study. Participants from both samples were genotyped and answered questions on sleep quality. A higher PGS for neuroticism was related to lower sleep quality concurrently and over time in both samples. Anxiety, depressive symptoms and neuroticism mediated these relationships in the two samples. Although effect sizes were small, the present study provides replicable evidence that individuals with a higher genetic predisposition to experience negative emotions and distress are at risk of sleep difficulties.

A powerful fine-mapping method for transcriptome-wide association studies

Transcriptome-wide association studies (TWAS) have been recently applied to successfully identify many novel genes associated with complex traits. While appealing, TWAS tend to identify multiple significant genes per locus, and many of them may not be causal due to confounding through linkage disequilibrium (LD) among SNPs. Here we introduce a powerful fine-mapping method that prioritizes putative causal genes by accounting for local LD. We apply a weighted adaptive test with eQTL-derived weights to maintain high power across various scenarios. Through simulations, we show that our new approach yielded a well-controlled Type I error rate while achieving higher power and AUC than competing methods. We applied our approach to a schizophrenia GWAS summary dataset and successfully prioritized some well-known schizophrenia-related genes, such as C4A. Importantly, our approach identified some putative causal genes (e.g., B3GAT1 and RGS6) that were missed by competing methods and TWAS. Our results suggest that our approach is a useful tool to prioritize putative causal genes, gaining insights into the mechanisms of complex traits.

Leveraging effect size distributions to improve polygenic risk scores derived from summary statistics of genome-wide association studies

Genetic risk prediction is an important problem in human genetics, and accurate prediction can facilitate disease prevention and treatment. Calculating polygenic risk score (PRS) has become widely used due to its simplicity and effectiveness, where only summary statistics from genome-wide association studies are needed in the standard method. Recently, several methods have been proposed to improve standard PRS by utilizing external information, such as linkage disequilibrium and functional annotations. In this paper, we introduce EB-PRS, a novel method that leverages information for effect sizes across all the markers to improve prediction accuracy. Compared to most existing genetic risk prediction methods, our method does not need to tune parameters nor external information. Real data applications on six diseases, including asthma, breast cancer, celiac disease, Crohn's disease, Parkinson's disease and type 2 diabetes show that EB-PRS achieved 307.1%, 42.8%, 25.5%, 3.1%, 74.3% and 49.6% relative improvements in terms of predictive r2 over standard PRS method with optimally tuned parameters. Besides, compared to LDpred that makes use of LD information, EB-PRS also achieved 37.9%, 33.6%, 8.6%, 36.2%, 40.6% and 10.8% relative improvements. We note that our method is not the first method leveraging effect size distributions. Here we first justify our method by presenting theoretical optimal property over existing methods in this class of methods, and substantiate our theoretical result with extensive simulation results. The R-package EBPRS that implements our method is available on CRAN.

Variable prediction accuracy of polygenic scores within an ancestry group

Fields as diverse as human genetics and sociology are increasingly using polygenic scores based on genome-wide association studies (GWAS) for phenotypic prediction. However, recent work has shown that polygenic scores have limited portability across groups of different genetic ancestries, restricting the contexts in which they can be used reliably and potentially creating serious inequities in future clinical applications. Using the UK Biobank data, we demonstrate that even within a single ancestry group (i.e., when there are negligible differences in linkage disequilibrium or in causal alleles frequencies), the prediction accuracy of polygenic scores can depend on characteristics such as the socio-economic status, age or sex of the individuals in which the GWAS and the prediction were conducted, as well as on the GWAS design. Our findings highlight both the complexities of interpreting polygenic scores and underappreciated obstacles to their broad use.

Polygenic risk for coronary heart disease acts through atherosclerosis in type 2 diabetes

Background

Type 2 diabetes increases the risk of coronary heart disease (CHD), yet the mechanisms involved remain poorly described. Polygenic risk scores (PRS) provide an opportunity to understand risk factors since they reflect etiologic pathways from the entire genome. We therefore tested whether a PRS for CHD influenced risk of CHD in individuals with type 2 diabetes and which risk factors were associated with this PRS.

Methods

We tested the association of a CHD PRS with CHD and its traditional clinical risk factors amongst individuals with type 2 diabetes in UK Biobank (N = 21,102). We next tested the association of the CHD PRS with atherosclerotic burden in a cohort of 352 genome-wide genotyped participants with type 2 diabetes who had undergone coronary angiograms.

Results

In the UK Biobank we found that the CHD PRS was strongly associated with CHD amongst individuals with type 2 diabetes (OR per standard deviation increase = 1.50; p = 1.5 × 10^− 59). But this CHD PRS was, at best, only weakly associated with traditional clinical risk factors, such as hypertension, hyperlipidemia, glycemic control, obesity and smoking. Conversely, in the angiographic cohort, the CHD PRS was strongly associated with multivessel stenosis (OR = 1.65; p = 4.9 × 10^− 4) and increased number of major stenotic lesions (OR = 1.35; p = 9.4 × 10^− 3).

Conclusions

Polygenic predisposition to CHD is strongly associated with atherosclerotic burden in individuals with type 2 diabetes and this effect is largely independent of traditional clinical risk factors. This suggests that genetic risk for CHD acts through atherosclerosis with little effect on most traditional risk factors, providing the opportunity to explore new biological pathways.

Variable prediction accuracy of polygenic scores within an ancestry group

Fields as diverse as human genetics and sociology are increasingly using polygenic scores based on genome-wide association studies (GWAS) for phenotypic prediction. However, recent work has shown that polygenic scores have limited portability across groups of different genetic ancestries, restricting the contexts in which they can be used reliably and potentially creating serious inequities in future clinical applications. Using the UK Biobank data, we demonstrate that even within a single ancestry group (i.e., when there are negligible differences in linkage disequilibrium or in causal alleles frequencies), the prediction accuracy of polygenic scores can depend on characteristics such as the socio-economic status, age or sex of the individuals in which the GWAS and the prediction were conducted, as well as on the GWAS design. Our findings highlight both the complexities of interpreting polygenic scores and underappreciated obstacles to their broad use.

Genetic Risk of Coronary Artery Disease, Features of Atherosclerosis, and Coronary Plaque Burden

Background

Polygenic risk scores (PRSs) based on risk variants from genome‐wide association studies predict coronary artery disease (CAD) risk. However, it is unknown whether the PRS is associated with specific CAD characteristics.

Methods and Results

We consecutively included 1645 patients with suspected stable CAD undergoing coronary computed tomography angiography. A multilocus PRS was calculated as the weighted sum of CAD risk variants. Plaques were evaluated using an 18‐segment model and characterized by stenosis severity and composition (soft [0%‐19% calcified], mixed‐soft [20%‐49% calcified], mixed‐calcified [50%‐79% calcified], or calcified [≥80% calcified]). Coronary artery calcium score and segment stenosis score were used to characterize plaque burden. For each standard deviation increase in the PRS, coronary artery calcium score increased by 78% (P=4.1e‐26) and segment stenosis score increased by 16% (P=2.4e‐29) in the fully adjusted model. The PRS was associated with a higher prevalence of obstructive plaques (odds ratio [OR]: 1.78, P=5.6e‐16), calcified (OR: 1.69, P=6.5e‐17), mixed‐calcified (OR: 1.67, P=7.3e‐9), mixed‐soft (OR: 1.45, P=1.6e‐6), and soft plaques (OR: 1.49, P=2.5e‐6), and a higher prevalence of plaque in each coronary vessel (all P<1.0e‐4). However, when analyzing data on a plaque level (3007 segments with plaque in 849 patients) the PRS was not associated with stenosis severity, plaque composition, or localization (all P>0.05).

Conclusions

Our results suggest that polygenic risk based on large genome‐wide association studies increases CAD risk through an increased burden of coronary atherosclerosis rather than promoting specific plaque features.

Clinical Trial Registration

URL: https://www.clinicaltrials.gov. Unique identifier: NCT02264717.

Genetic and gene expression signatures in multiple sclerosis

Multiple sclerosis (MS) exhibits a well-documented increased incidence in individuals with respective family history, that is, is a heritable disease. In the last decade, genome-wide association studies have enabled the agnostic interrogation of the whole genome at a large scale. To date, over 200 genetic associations have been described at the strict level of genome-wide significance. Our current understanding of MS genetics can explain up to half of the disease's heritability, raising the important question of whether this is enough information to leverage toward improving diagnosis in MS. Parallel advancements in technologies that allow the characterization of the full transcriptome down to the single-cell level have enabled the generation of an unprecedented wealth of information. Transcriptional changes of putative causal cells could be utilized to identify early signs of disease onset. These recent findings in genetics and genomics, coupled with new technologies and deeply phenotyped cohorts, have the potential to improve the diagnosis of MS.

Genomic prediction of alcohol-related morbidity and mortality

While polygenic risk scores (PRS) have been shown to predict many diseases and risk factors, the potential of genomic prediction in harm caused by alcohol use has not yet been extensively studied. Here, we built a novel polygenic risk score of 1.1 million variants for alcohol consumption and studied its predictive capacity in 96,499 participants from the FinnGen study and 39,695 participants from prospective cohorts with detailed baseline data and up to 25 years of follow-up time. A 1 SD increase in the PRS was associated with 11.2 g (=0.93 drinks) higher weekly alcohol consumption (CI = 9.85-12.58 g, p = 2.3 × 10-58). The PRS was associated with alcohol-related morbidity (4785 incident events) and the risk estimate between the highest and lowest quintiles of the PRS was 1.83 (95% CI = 1.66-2.01, p = 1.6 × 10-36). When adjusted for self-reported alcohol consumption, education, marital status, and gamma-glutamyl transferase blood levels in 28,639 participants with comprehensive baseline data from prospective cohorts, the risk estimate between the highest and lowest quintiles of the PRS was 1.58 (CI = 1.26-1.99, p = 8.2 × 10-5). The PRS was also associated with all-cause mortality with a risk estimate of 1.33 between the highest and lowest quintiles (CI = 1.20-1.47, p = 4.5 × 10-8) in the adjusted model. In conclusion, the PRS for alcohol consumption independently associates for both alcohol-related morbidity and all-cause mortality. Together, these findings underline the importance of heritable factors in alcohol-related health burden while highlighting how measured genetic risk for an important behavioral risk factor can be used to predict related health outcomes.

Multitrait analysis of glaucoma identifies new risk loci and enables polygenic prediction of disease susceptibility and progression

Glaucoma, a disease characterized by progressive optic nerve degeneration, can be prevented through timely diagnosis and treatment. We characterize optic nerve photographs of 67,040 UK Biobank participants and use a multitrait genetic model to identify risk loci for glaucoma. A glaucoma polygenic risk score (PRS) enables effective risk stratification in unselected glaucoma cases and modifies penetrance of the MYOC variant encoding p.Gln368Ter, the most common glaucoma-associated myocilin variant. In the unselected glaucoma population, individuals in the top PRS decile reach an absolute risk for glaucoma 10 years earlier than the bottom decile and are at 15-fold increased risk of developing advanced glaucoma (top 10% versus remaining 90%, odds ratio = 4.20). The PRS predicts glaucoma progression in prospectively monitored, early manifest glaucoma cases (P = 0.004) and surgical intervention in advanced disease (P = 3.6 × 10^-6). This glaucoma PRS will facilitate the development of a personalized approach for earlier treatment of high-risk individuals, with less intensive monitoring and treatment being possible for lower-risk groups.

Association Between Polygenic Risk Score and Risk of Myopia

Importance

Myopia is a leading cause of untreatable visual impairment and is increasing in prevalence worldwide. Interventions for slowing childhood myopia progression have shown success in randomized clinical trials; hence, there is a need to identify which children would benefit most from treatment intervention.

Objectives

To examine whether genetic information alone can identify children at risk of myopia development and whether including a child's genetic predisposition to educational attainment is associated with improved genetic prediction of the risk of myopia.

Design, Setting, and Participants

Meta-analysis of 3 genome-wide association studies (GWAS) including a total of 711 984 individuals. These were a published GWAS for educational attainment and 2 GWAS for refractive error in the UK Biobank, which is a multisite cohort study that recruited participants between January 2006 and October 2010. A polygenic risk score was applied in a population-based validation sample examined between September 1998 and September 2000 (Avon Longitudinal Study of Parents and Children [ALSPAC] mothers). Data analysis was performed from February 2018 to May 2019.

Main Outcomes and Measures

The primary outcome was the area under the receiver operating characteristic curve (AUROC) in analyses for predicting myopia, using noncycloplegic autorefraction measurements for myopia severity levels of less than or equal to -0.75 diopter (D) (any), less than or equal to -3.00 D (moderate), or less than or equal to -5.00 D (high). The predictor variable was a polygenic risk score (PRS) derived from genome-wide association study data for refractive error (n = 95 619), age of onset of spectacle wear (n = 287 448), and educational attainment (n = 328 917).

Results

A total of 383 067 adults aged 40 to 69 years from the UK Biobank were included in the new GWAS analyses. The PRS was evaluated in 1516 adults aged 24 to 51 years from the ALSPAC mothers cohort. The PRS had an AUROC of 0.67 (95% CI, 0.65-0.70) for myopia, 0.75 (95% CI, 0.70-0.79) for moderate myopia, and 0.73 (95% CI, 0.66-0.80) for high myopia. Inclusion in the PRS of information associated with genetic predisposition to educational attainment marginally improved the AUROC for myopia (AUROC, 0.674 vs 0.668; P = .02), but not those for moderate and high myopia. Individuals with a PRS in the top 10% were at 6.1-fold higher risk (95% CI, 3.4-10.9) of high myopia.

Conclusions and Relevance

A personalized medicine approach may be feasible for detecting very young children at risk of myopia. However, accuracy must improve further to merit uptake in clinical practice; currently, cycloplegic autorefraction remains a better indicator of myopia risk (AUROC, 0.87).

Methodological challenges in constructing DNA methylation risk scores

Polygenic approaches often access more variance of complex traits than is possible by single variant approaches. For genotype data, genetic risk scores (GRS) are widely used for risk prediction as well as in association and interaction studies. Recently, interest has been growing in transferring GRS approaches to DNA methylation data (methylation risk scores, MRS), which can be used 1) as biomarkers for environmental exposures, 2) in association analyses in which single CpG sites do not achieve significance, 3) as dimension reduction approach in interaction and mediation analyses, and 4) to predict individual risks of disease or treatment success. Most GRS approaches can directly be transferred to methylation data. However, since methylation data is more sensitive to confounding, e.g. by age and tissue, it is more complex to find appropriate external weights. In this review, we will outline the adaption of current GRS approaches to methylation data and highlight occurring challenges.

Pathways Between a Polygenic Score for Educational Attainment and Higher Educational Attainment in an African American Sample

We investigated the extent to which performance on standardized achievement tests, executive function (EF), and aggression in childhood and adolescence accounted for the relationship between a polygenic score for educational attainment (EA PGS) and years of education in a community sample of African Americans. Participants (N = 402; 49.9% female) were initially recruited for an elementary school-based prevention trial in a Mid-Atlantic city and followed into adulthood. In first and twelfth grade, participants completed math and reading standardized tests and teachers reported on participants' aggression and EF, specifically impulsivity and concentration problems. At age 20, participants reported on their years of education and post-secondary degrees attained and their genotype was assayed from blood or buccal swabs. An EA PGS was created using results from a large-scale GWAS on EA. A higher EA PGS was associated with higher education indirectly via adolescent achievement. No other mediating mechanisms were significant. Adolescent academic achievement is thus one mechanism through which polygenic propensity for EA influences post-secondary education among urban, African American youth.

Broad- and narrow-sense validity performance of three polygenic risk score methods for prostate cancer risk assessment

BACKGROUND

Several polygenic risk score (PRS) methods are available for measuring the cumulative effect of multiple risk-associated single nucleotide polymorphisms (SNPs). Their performance in predicting risk at the individual level has not been well studied.

METHODS

We compared the performance of three PRS methods for prostate cancer risk assessment in a clinical trial cohort, including genetic risk score (GRS), pruning and thresholding (P + T), and linkage disequilibrium prediction (LDpred). Performance was evaluated for score deciles (broad-sense validity) and score values (narrow-sense validity).

RESULTS

A training process was required to identify the best P + T model (397 SNPs) and LDpred model (3 011 362 SNPs). In contrast, GRS was directly calculated based on 110 established risk-associated SNPs. For broad-sense validity in the testing population, higher deciles were significantly associated with higher observed risk; P_trend was 7.40 × 10^-11 , 7.64 × 10^-13 , and 7.51 × 10^-10 for GRS, P + T, and LDpred, respectively. For narrow-sense validity, the calibration slope (1 is best) was 1.03, 0.77, and 0.87, and mean bias score (0 is best) was 0.09, 0.21, and 0.10 for GRS, P + T, and LDpred, respectively.

CONCLUSIONS

The performance of GRS was better than P + T and LDpred. Fewer and well-established SNPs of GRS also make it more feasible and interpretable for genetic testing at the individual level.

The p factor: genetic analyses support a general dimension of psychopathology in childhood and adolescence

BACKGROUND

Diverse behaviour problems in childhood correlate phenotypically, suggesting a general dimension of psychopathology that has been called the p factor. The shared genetic architecture between childhood psychopathology traits also supports a genetic p. This study systematically investigates the manifestation of this common dimension across self-, parent- and teacher-rated measures in childhood and adolescence.

METHODS

The sample included 7,026 twin pairs from the Twins Early Development Study (TEDS). First, we employed multivariate twin models to estimate common genetic and environmental influences on p based on diverse measures of behaviour problems rated by children, parents and teachers at ages 7, 9, 12 and 16 (depressive traits, emotional problems, peer problems, autism traits, hyperactivity, antisocial behaviour, conduct problems and psychopathic tendencies). Second, to assess the stability of genetic and environmental influences on p across time, we conducted longitudinal twin modelling of the first phenotypic principal components of childhood psychopathological measures across each of the four ages. Third, we created a genetic p factor in 7,026 unrelated genotyped individuals based on eight polygenic scores for psychiatric disorders to estimate how a general polygenic predisposition to mostly adult psychiatric disorders relates to childhood p.

RESULTS

Behaviour problems were consistently correlated phenotypically and genetically across ages and raters. The p factor is substantially heritable (50%-60%) and manifests consistently across diverse ages and raters. However, residual variation in the common factor models indicates unique contributions as well. Genetic correlations of p components across childhood and adolescence suggest stability over time (49%-78%). A polygenic general psychopathology factor derived from studies of psychiatric disorders consistently predicted a general phenotypic p factor across development (0.3%-0.9%).

CONCLUSIONS

Diverse forms of psychopathology generally load on a common p factor, which is highly heritable. There are substantial genetic influences on the stability of p across childhood. Our analyses indicate genetic overlap between general risk for psychiatric disorders in adulthood and p in childhood, even as young as age 7. The p factor has far-reaching implications for genomic research and, eventually, for diagnosis and treatment of behaviour problems.

Maternal environmental risk factors and the development of internalizing and externalizing problems in childhood: The complex role of genetic factors

The development of problem behavior in children is associated with exposure to environmental factors, including the maternal environment. Both are influenced by genetic factors, which may also be correlated, that is, environmental risk and problem behavior in children might be influenced by partly the same genetic factors. In addition, environmental and genetic factors could interact with each other increasing the risk of problem behavior in children. To date, limited research investigated these mechanisms in a genome-wide approach. Therefore, the goal of this study was to investigate the association between genetic risk for psychiatric and related traits, as indicated by polygenetic risk scores (PRSs), exposure to previously identified maternal risk factors, and problem behavior in a sample of 1,154 children from the Amsterdam Born Children and their Development study at ages 5-6 and 11-12 years old. The PRSs were derived from genome-wide association studies (GWASs) on schizophrenia, major depressive disorder, neuroticism, and wellbeing. Regression analysis showed that the PRSs were associated with exposure to multiple environmental risk factors, suggesting passive gene-environment correlation. In addition, the PRS based on the schizophrenia GWAS was associated with externalizing behavior problems in children at age 5-6. We did not find any association with problem behavior for the other PRSs. Our results indicate that genetic predispositions for psychiatric disorders and wellbeing are associated with early environmental risk factors for children's problem behavior.

Predicting educational achievement from genomic measures and socioeconomic status

The two best predictors of children's educational achievement available from birth are parents’ socioeconomic status (SES) and, recently, children's inherited DNA differences that can be aggregated in genome‐wide polygenic scores (GPS). Here, we chart for the first time the developmental interplay between these two predictors of educational achievement at ages 7, 11, 14 and 16 in a sample of almost 5,000 UK school children. We show that the prediction of educational achievement from both GPS and SES increases steadily throughout the school years. Using latent growth curve models, we find that GPS and SES not only predict educational achievement in the first grade but they also account for systematic changes in achievement across the school years. At the end of compulsory education at age 16, GPS and SES, respectively, predict 14% and 23% of the variance of educational achievement. Analyses of the extremes of GPS and SES highlight their influence and interplay: In children who have high GPS and come from high SES families, 77% go to university, whereas 21% of children with low GPS and from low SES backgrounds attend university. We find that the associations of GPS and SES with educational achievement are primarily additive, suggesting that their joint influence is particularly dramatic for children at the extreme ends of the distribution.

A comprehensive analysis of methods for assessing polygenic burden on Alzheimer's disease pathology and risk beyond APOE

Genome-wide association studies have identified dozens of loci that alter the risk to develop Alzheimer's disease. However, with the exception of the APOE-ε4 allele, most variants bear only little individual effect and have, therefore, limited diagnostic and prognostic value. Polygenic risk scores aim to collate the disease risk distributed across the genome in a single score. Recent works have demonstrated that polygenic risk scores designed for Alzheimer's disease are predictive of clinical diagnosis, pathology confirmed diagnosis and changes in imaging biomarkers. Methodological innovations in polygenic risk modelling include the polygenic hazard score, which derives effect estimates for individual single nucleotide polymorphisms from survival analysis, and methods that account for linkage disequilibrium between genomic loci. In this work, using data from the Alzheimer's disease neuroimaging initiative, we compared different approaches to quantify polygenic disease burden for Alzheimer's disease and their association (beyond the APOE locus) with a broad range of Alzheimer's disease-related traits: cross-sectional CSF biomarker levels, cross-sectional cortical amyloid burden, clinical diagnosis, clinical progression, longitudinal loss of grey matter and longitudinal decline in cognitive function. We found that polygenic scores were associated beyond APOE with clinical diagnosis, CSF-tau levels and, to a minor degree, with progressive atrophy. However, for many other tested traits such as clinical disease progression, CSF amyloid, cognitive decline and cortical amyloid load, the additional effects of polygenic burden beyond APOE were of minor nature. Overall, polygenic risk scores and the polygenic hazard score performed equally and given the ease with which polygenic risk scores can be derived; they constitute the more practical choice in comparison with polygenic hazard scores. Furthermore, our results demonstrate that incomplete adjustment for the APOE locus, i.e. only adjusting for APOE-ε4 carrier status, can lead to overestimated effects of polygenic scores due to APOE-ε4 homozygous participants. Lastly, on many of the tested traits, the major driving factor remained the APOE locus, with the exception of quantitative CSF-tau and p-tau measures.

Benefits and limitations of genome-wide association studies

Genome-wide association studies (GWAS) involve testing genetic variants across the genomes of many individuals to identify genotype-phenotype associations. GWAS have revolutionized the field of complex disease genetics over the past decade, providing numerous compelling associations for human complex traits and diseases. Despite clear successes in identifying novel disease susceptibility genes and biological pathways and in translating these findings into clinical care, GWAS have not been without controversy. Prominent criticisms include concerns that GWAS will eventually implicate the entire genome in disease predisposition and that most association signals reflect variants and genes with no direct biological relevance to disease. In this Review, we comprehensively assess the benefits and limitations of GWAS in human populations and discuss the relevance of performing more GWAS.

Making the Most of Clumping and Thresholding for Polygenic Scores

Polygenic prediction has the potential to contribute to precision medicine. Clumping and thresholding (C+T) is a widely used method to derive polygenic scores. When using C+T, several p value thresholds are tested to maximize predictive ability of the derived polygenic scores. Along with this p value threshold, we propose to tune three other hyper-parameters for C+T. We implement an efficient way to derive thousands of different C+T scores corresponding to a grid over four hyper-parameters. For example, it takes a few hours to derive 123K different C+T scores for 300K individuals and 1M variants using 16 physical cores. We find that optimizing over these four hyper-parameters improves the predictive performance of C+T in both simulations and real data applications as compared to tuning only the p value threshold. A particularly large increase can be noted when predicting depression status, from an AUC of 0.557 (95% CI: [0.544-0.569]) when tuning only the p value threshold to an AUC of 0.592 (95% CI: [0.580-0.604]) when tuning all four hyper-parameters we propose for C+T. We further propose stacked clumping and thresholding (SCT), a polygenic score that results from stacking all derived C+T scores. Instead of choosing one set of hyper-parameters that maximizes prediction in some training set, SCT learns an optimal linear combination of all C+T scores by using an efficient penalized regression. We apply SCT to eight different case-control diseases in the UK biobank data and find that SCT substantially improves prediction accuracy with an average AUC increase of 0.035 over standard C+T.

Novel genome-wide associations for anhedonia, genetic correlation with psychiatric disorders, and polygenic association with brain structure

Anhedonia is a core symptom of several psychiatric disorders but its biological underpinnings are poorly understood. We performed a genome-wide association study of state anhedonia in 375,275 UK Biobank participants and assessed for genetic correlation between anhedonia and neuropsychiatric conditions (major depressive disorder, schizophrenia, bipolar disorder, obsessive compulsive disorder and Parkinson's Disease). We then used a polygenic risk score approach to test for association between genetic loading for anhedonia and both brain structure and brain function. This included: magnetic resonance imaging (MRI) assessments of total grey matter volume, white matter volume, cerebrospinal fluid volume, and 15 cortical/subcortical regions of interest; diffusion tensor imaging (DTI) measures of white matter tract integrity; and functional MRI activity during an emotion processing task. We identified 11 novel loci associated at genome-wide significance with anhedonia, with a SNP heritability estimate (h2SNP) of 5.6%. Strong positive genetic correlations were found between anhedonia and major depressive disorder, schizophrenia and bipolar disorder; but not with obsessive compulsive disorder or Parkinson's Disease. Polygenic risk for anhedonia was associated with poorer brain white matter integrity, smaller total grey matter volume, and smaller volumes of brain regions linked to reward and pleasure processing, including orbito-frontal cortex. In summary, the identification of novel anhedonia-associated loci substantially expands our current understanding of the biological basis of state anhedonia and genetic correlations with several psychiatric disorders confirm the utility of this phenotype as a transdiagnostic marker of vulnerability to mental illness. We also provide the first evidence that genetic risk for state anhedonia influences brain structure, including in regions associated with reward and pleasure processing.

Polygenic score for educational attainment captures DNA variants shared between personality traits and educational achievement

Genome-wide polygenic scores (GPS) can be used to predict individual genetic risk and resilience. For example, a GPS for years of education (EduYears) explains substantial variance in cognitive traits such as general cognitive ability and educational achievement. Personality traits are also known to contribute to individual differences in educational achievement. However, the association between EduYears GPS and personality traits remains largely unexplored. Here, we test the relation between GPS for EduYears, neuroticism, and well-being, and 6 personality and motivation domains: Academic Motivation, Extraversion, Openness, Conscientiousness, Neuroticism, and Agreeableness. The sample was drawn from a U.K.-representative sample of up to 8,322 individuals assessed at age 16. We find that EduYears GPS was positively associated with Openness, Conscientiousness, Agreeableness, and Academic Motivation, predicting between 0.6% and 3% of the variance. In addition, we find that EduYears GPS explains between 8% and 16% of the association between personality domains and educational achievement at the end of compulsory education. In contrast, both the neuroticism and well-being GPS significantly accounted for between 0.3% and 0.7% of the variance in a subset of personality domains. Furthermore, they did not significantly account for any of the covariance between the personality domains and achievement, with the exception of the neuroticism GPS explaining 5% of the covariance between Neuroticism and achievement. These results demonstrate that the genetic effects of educational attainment relate to personality traits, highlighting the multifaceted nature of EduYears GPS. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Genetic Risk Scores

Genome-wide variation data with millions of genetic markers have become commonplace. However, the potential for interpretation and application of these data for clinical assessment of outcomes of interest, and prediction of disease risk, is currently not fully realized. Many common complex diseases now have numerous, well-established risk loci and likely harbor many genetic determinants with effects too small to be detected at genome-wide levels of statistical significance. A simple and intuitive approach for converting genetic data to a predictive measure of disease susceptibility is to aggregate the effects of these loci into a single measure, the genetic risk score. Here, we describe some common methods and software packages for calculating genetic risk scores and polygenic risk scores, with focus on studies of common complex diseases. We review the basic information needed, as well as important considerations for constructing genetic risk scores, including specific requirements for phenotypic and genetic data, and limitations in their application. © 2019 by John Wiley & Sons, Inc.

Screening Human Embryos for Polygenic Traits Has Limited Utility

The increasing proportion of variance in human complex traits explained by polygenic scores, along with progress in preimplantation genetic diagnosis, suggests the possibility of screening embryos for traits such as height or cognitive ability. However, the expected outcomes of embryo screening are unclear, which undermines discussion of associated ethical concerns. Here, we use theory, simulations, and real data to evaluate the potential gain of embryo screening, defined as the difference in trait value between the top-scoring embryo and the average embryo. The gain increases very slowly with the number of embryos but more rapidly with the variance explained by the score. Given current technology, the average gain due to screening would be ≈2.5 cm for height and ≈2.5 IQ points for cognitive ability. These mean values are accompanied by wide prediction intervals, and indeed, in large nuclear families, the majority of children top-scoring for height are not the tallest.

Validity of polygenic risk scores: are we measuring what we think we are?

Polygenic risk scores (PRSs) have become the standard for quantifying genetic liability in the prediction of disease risks. PRSs are generally constructed as weighted sum scores of risk alleles using effect sizes from genome-wide association studies as their weights. The construction of PRSs is being improved with more appropriate selection of independent single-nucleotide polymorphisms (SNPs) and optimized estimation of their weights but is rarely reflected upon from a theoretical perspective, focusing on the validity of the risk score. Borrowing from psychometrics, this paper discusses the validity of PRSs and introduces the three main types of validity that are considered in the evaluation of tests and measurements: construct, content, and criterion validity. This introduction is followed by a discussion of three topics that challenge the validity of PRS, namely, their claimed independence of clinical risk factors, the consequences of relaxing SNP inclusion thresholds and the selection of SNP weights. This discussion of the validity of PRS reminds us that we need to keep questioning if weighted sums of risk alleles are measuring what we think they are in the various scenarios in which PRSs are used and that we need to keep exploring alternative modeling strategies that might better reflect the underlying biological pathways.

Patients With High Genome-Wide Polygenic Risk Scores for Coronary Artery Disease May Receive Greater Clinical Benefit From Alirocumab Treatment in the ODYSSEY OUTCOMES Trial

BACKGROUND

Alirocumab, an antibody that blocks PCSK9 (proprotein convertase subtilisin/kexin type 9), was associated with reduced major adverse cardiovascular events (MACE) and death in the ODYSSEY OUTCOMES trial (Evaluation of Cardiovascular Outcomes After an Acute Coronary Syndrome During Treatment With Alirocumab). In this study, higher baseline levels of low-density lipoprotein cholesterol (LDL-C) predicted greater benefit from alirocumab treatment. Recent studies indicate high polygenic risk scores (PRS) for coronary artery disease (CAD) identify individuals at higher risk who derive increased benefit from statins. We performed post hoc analyses to determine whether high PRS for CAD identifies higher-risk individuals, independent of baseline LDL-C and other known risk factors, who might derive greater benefit from alirocumab treatment.

METHODS

ODYSSEY OUTCOMES was a randomized, double-blind, placebo-controlled trial comparing alirocumab or placebo in 18 924 patients with acute coronary syndrome and elevated atherogenic lipoproteins despite optimized statin treatment. The primary endpoint (MACE) comprised death of CAD, nonfatal myocardial infarction, ischemic stroke, or unstable angina requiring hospitalization. A genome-wide PRS for CAD comprising 6 579 025 genetic variants was evaluated in 11 953 patients with available DNA samples. Analysis of MACE risk was performed in placebo-treated patients, whereas treatment benefit analysis was performed in all patients.

RESULTS

The incidence of MACE in the placebo group was related to PRS for CAD: 17.0% for high PRS patients (>90th percentile) and 11.4% for lower PRS patients (≤90th percentile; P<0.001); this PRS relationship was not explained by baseline LDL-C or other established risk factors. Both the absolute and relative reduction of MACE by alirocumab compared with placebo was greater in high versus low PRS patients. There was an absolute reduction by alirocumab in high versus low PRS groups of 6.0% and 1.5%, respectively, and a relative risk reduction by alirocumab of 37% in the high PRS group (hazard ratio, 0.63 [95% CI, 0.46-0.86]; P=0.004) versus a 13% reduction in the low PRS group (hazard ratio, 0.87 [95% CI, 0.78-0.98]; P=0.022; interaction P=0.04).

CONCLUSIONS

A high PRS for CAD is associated with elevated risk for recurrent MACE after acute coronary syndrome and a larger absolute and relative risk reduction with alirocumab treatment, providing an independent tool for risk stratification and precision medicine.

Improved polygenic prediction by Bayesian multiple regression on summary statistics

Accurate prediction of an individual’s phenotype from their DNA sequence is one of the great promises of genomics and precision medicine. We extend a powerful individual-level data Bayesian multiple regression model (BayesR) to one that utilises summary statistics from genome-wide association studies (GWAS), SBayesR. In simulation and cross-validation using 12 real traits and 1.1 million variants on 350,000 individuals from the UK Biobank, SBayesR improves prediction accuracy relative to commonly used state-of-the-art summary statistics methods at a fraction of the computational resources. Furthermore, using summary statistics for variants from the largest GWAS meta-analysis (n ≈ 700, 000) on height and BMI, we show that on average across traits and two independent data sets that SBayesR improves prediction R² by 5.2% relative to LDpred and by 26.5% relative to clumping and p value thresholding.

Various approaches are being used for polygenic prediction including Bayesian multiple regression methods that require access to individual-level genotype data. Here, the authors extend BayesR to utilise GWAS summary statistics (SBayesR) and show that it outperforms other summary statistic-based methods.

Monogenic and Polygenic Contributions to Atrial Fibrillation Risk: Results From a National Biobank

RATIONALE

Genome-wide association studies have identified over 100 genetic loci for atrial fibrillation (AF); recent work described an association between loss-of-function (LOF) variants in TTN and early-onset AF.

OBJECTIVE

We sought to determine the contribution of rare and common genetic variation to AF risk in the general population.

METHODS

The UK Biobank is a population-based study of 500 000 individuals including a subset with genome-wide genotyping and exome sequencing. In this case-control study, we included AF cases and controls of genetically determined white-European ancestry; analyses were performed using a logistic mixed-effects model adjusting for age, sex, the first 4 principal components of ancestry, empirical relationships, and case-control imbalance. An exome-wide, gene-based burden analysis was performed to examine the relationship between AF and rare, high-confidence LOF variants in genes with ≥10 LOF carriers. A polygenic risk score for AF was estimated using the LDpred algorithm. We then compared the contribution of AF polygenic risk score and LOF variants to AF risk.

RESULTS

The study included 1546 AF cases and 41 593 controls. In an analysis of 9099 genes with sufficient LOF variant carriers, a significant association between AF and rare LOF variants was observed in a single gene, TTN (odds ratio, 2.71, P=2.50×10^-8). The association with AF was more significant (odds ratio, 6.15, P=3.26×10^-14) when restricting to LOF variants located in exons highly expressed in cardiac tissue (TTN_LOF). Overall, 0.44% of individuals carried TTN_LOF variants, of whom 14% had AF. Among individuals in the highest 0.44% of the AF polygenic risk score only 9.3% had AF. In contrast, the AF polygenic risk score explained 4.7% of the variance in AF susceptibility, while TTN_LOF variants only accounted for 0.2%.

CONCLUSIONS

Both monogenic and polygenic factors contribute to AF risk in the general population. While rare TTN_LOF variants confer a substantial AF penetrance, the additive effect of many common variants explains a larger proportion of genetic susceptibility to AF.