Modeling Linkage Disequilibrium Increases Accuracy of Polygenic Risk Scores

Increased burden of ultra-rare protein-altering variants among 4,877 individuals with schizophrenia

By analyzing the exomes of 12,332 unrelated Swedish individuals – including 4,877 affected with schizophrenia – in ways informed by exome sequences from 45,376 other individuals, we identified 244,246 coding-sequence and splice-site ultra-rare variants (URVs) that were unique to individual Swedes. We found that gene-disruptive and putatively protein-damaging URVs (but not synonymous URVs) were more abundant in schizophrenia cases than controls (P = 1.3 × 10⁻¹⁰). This elevation of protein-compromising URVs was several times larger than an analogously elevated rate for de novo mutations, suggesting that most rare-variant effects on schizophrenia risk are inherited. Among individuals with schizophrenia, the elevated frequency of protein-compromising URVs was concentrated in brain-expressed genes, particularly in neuronally expressed genes; most of this genetic signal arose from large sets of genes whose RNAs have been found to interact with synaptically localized proteins. Our results suggest that synaptic dysfunction may mediate a large fraction of strong, individually rare genetic influences on schizophrenia risk.

Polygenic risk scores for cigarettes smoked per day do not generalize to a Native American population

BACKGROUND

Recent studies have demonstrated the utility of polygenic risk scores (PRSs) for exploring the genetic etiology of psychiatric phenotypes and the genetic correlations between them. To date, these studies have been conducted almost exclusively using participants of European ancestry, and thus, there is a need for similar studies conducted in other ancestral populations. However, given that the predictive ability of PRSs are sensitive to differences in linkage disequilibrium (LD) patterns and minor allele frequencies across discovery and target samples, the applicability of PRSs developed in European ancestry samples to other ancestral populations has yet to be determined. Therefore, the current study derived PRSs for cigarettes per day (CPD) from predominantly European-ancestry samples and examined their ability to predict nicotine dependence (ND) in a Native American (NA) population sample.

METHOD

Results from the Tobacco and Genetics Consortium's meta-analysis of genome-wide association studies of CPD were used to compute PRSs in a NA community sample (N=288). These scores were then used to predict ND diagnostic status.

RESULTS

The PRS was not significantly associated with liability for ND in the full sample. However, a significant interaction between PRS and percent NA ancestry was observed. Risk scores were positively associated with liability for ND at higher levels of European ancestry, but no association was observed at higher levels of NA ancestry.

CONCLUSION

These findings illustrate how differences in patterns of LD across discovery and target samples can reduce the predictive ability of PRSs for complex traits.

Multimarker analysis suggests the involvement of BDNF signaling and microRNA biosynthesis in suicidal behavior

Despite moderate heritability estimates the genetics of suicidal behavior remains unclear, genome-wide association and candidate gene studies focusing on single nucleotide associations reported inconsistent findings. Our study explored biologically informed, multimarker candidate gene associations with suicidal behavior in mood disorders. We analyzed the GAIN Whole Genome Association Study of Bipolar Disorder version 3 (n = 999, suicidal n = 358) and the GAIN Major Depression: Stage 1 Genomewide Association in Population-Based Samples (n = 1,753, suicidal n = 245) datasets. Suicidal behavior was defined as severe suicidal ideation or attempt. Candidate genes were selected based on literature search (Geneset1, n = 35), gene expression data of microRNA genes, (Geneset2, n = 68) and their target genes (Geneset3, n = 11,259). Quality control, dosage analyses were carried out with PLINK. Gene-based associations of Geneset1 were analyzed with KGG. Polygenic profile scores of suicidal behavior were computed in the major depression dataset both with PRSice and LDpred and validated in the bipolar disorder data. Several nominally significant gene-based associations were detected, but only DICER1 associated with suicidal behavior in both samples, while only the associations of NTRK2 in the depression sample reached family wise and experiment wise significance. Polygenic profile scores negatively predicted suicidal behavior in the bipolar sample for only Geneset2, with the strongest prediction by PRSice at Pt  < 0.03 (Nagelkerke R(2)  = 0.01, P < 0.007). Gene-based association results confirmed the potential involvement of the BDNF-NTRK2-CREB pathway in the pathogenesis of suicide and the cross-disorder association of DICER1. Polygenic risk prediction of the selected miRNA genes indicates that the miRNA system may play a mediating role, but with considerable pleiotropy. © 2016 Wiley Periodicals, Inc.

Accuracy of Gene Scores when Pruning Markers by Linkage Disequilibrium

OBJECTIVE

Gene scores are often used to model the combined effects of genetic variants. When variants are in linkage disequilibrium, it is common to prune all variants except the most strongly associated. This avoids duplicating information but discards information when variants have independent effects. However, joint modelling of correlated variants increases the sampling error in the gene score. In recent applications, joint modelling has offered only small improvements in accuracy over pruning. We aimed to quantify the relationship between pruning and joint modelling in relation to sample size.

METHODS

We derived the coefficient of determination R2 for a gene score constructed from pruned markers, and for one constructed from correlated markers with jointly estimated effects.

RESULTS

Pruned scores tend to have slightly lower R2 than jointly modelled scores, but the differences are small at sample sizes up to 100,000. If the proportion of correlated variants is high, joint modelling can obtain modest improvements asymptotically.

CONCLUSIONS

The small gains observed to date from joint modelling can be explained by sample size. As studies become larger, joint modelling will be useful for traits affected by many correlated variants, but the improvements may remain small. Pruning remains a useful heuristic for current studies.

Nationwide Genomic Study in Denmark Reveals Remarkable Population Homogeneity

Denmark has played a substantial role in the history of Northern Europe. Through a nationwide scientific outreach initiative, we collected genetic and anthropometrical data from ∼800 high school students and used them to elucidate the genetic makeup of the Danish population, as well as to assess polygenic predictions of phenotypic traits in adolescents. We observed remarkable homogeneity across different geographic regions, although we could still detect weak signals of genetic structure reflecting the history of the country. Denmark presented genomic affinity with primarily neighboring countries with overall resemblance of decreasing weight from Britain, Sweden, Norway, Germany, and France. A Polish admixture signal was detected in Zealand and Funen, and our date estimates coincided with historical evidence of Wend settlements in the south of Denmark. We also observed considerably diverse demographic histories among Scandinavian countries, with Denmark having the smallest current effective population size compared to Norway and Sweden. Finally, we found that polygenic prediction of self-reported adolescent height in the population was remarkably accurate (R² = 0.639 ± 0.015). The high homogeneity of the Danish population could render population structure a lesser concern for the upcoming large-scale gene-mapping studies in the country.

Age-Dependent Pleiotropy Between General Cognitive Function and Major Psychiatric Disorders

BACKGROUND

General cognitive function predicts psychiatric illness across the life course. This study examines the role of pleiotropy in explaining the link between cognitive function and psychiatric disorder.

METHODS

We used two large genome-wide association study data sets on cognitive function-one from older age, n = 53,949, and one from childhood, n = 12,441. We also used genome-wide association study data on educational attainment, n = 95,427, to examine the validity of its use as a proxy phenotype for cognitive function. Using a new method, linkage disequilibrium regression, we derived genetic correlations, free from the confounding of clinical state between psychiatric illness and cognitive function.

RESULTS

We found a genetic correlation of .711 (p = 2.26e-12) across the life course for general cognitive function. We also showed a positive genetic correlation between autism spectrum disorder and cognitive function in childhood (rg = .360, p = .0009) and for educational attainment (rg = .322, p = 1.37e-5) but not in older age. In schizophrenia, we found a negative genetic correlation between older age cognitive function (rg = -.231, p = 3.81e-12) but not in childhood or for educational attainment. For Alzheimer's disease, we found negative genetic correlations with childhood cognitive function (rg = -.341, p = .001), educational attainment (rg = -.324, p = 1.15e-5), and with older age cognitive function (rg = -.324, p = 1.78e-5).

CONCLUSIONS

The pleiotropy exhibited between cognitive function and psychiatric disorders changed across the life course. These age-dependent associations might explain why negative selection has not removed variants causally associated with autism spectrum disorder or schizophrenia.

Genetic evidence for natural selection in humans in the contemporary United States

Recent findings from molecular genetics now make it possible to test directly for natural selection by analyzing whether genetic variants associated with various phenotypes have been under selection. I leverage these findings to construct polygenic scores that use individuals' genotypes to predict their body mass index, educational attainment (EA), glucose concentration, height, schizophrenia, total cholesterol, and (in females) age at menarche. I then examine associations between these scores and fitness to test whether natural selection has been occurring. My study sample includes individuals of European ancestry born between 1931 and 1953 who participated in the Health and Retirement Study, a representative study of the US population. My results imply that natural selection has been slowly favoring lower EA in both females and males, and are suggestive that natural selection may have favored a higher age at menarche in females. For EA, my estimates imply a rate of selection of about -1.5 mo of education per generation (which pales in comparison with the increases in EA observed in contemporary times). Although they cannot be projected over more than one generation, my results provide additional evidence that humans are still evolving-albeit slowly, especially compared with the rapid changes that have occurred over the past few generations due to cultural and environmental factors.

Reflections on the Field of Human Genetics: A Call for Increased Disease Genetics Theory

Development of human genetics theoretical models and the integration of those models with experiment and statistical evaluation are critical for scientific progress. This perspective argues that increased effort in disease genetics theory, complementing experimental, and statistical efforts, will escalate the unraveling of molecular etiologies of complex diseases. In particular, the development of new, realistic disease genetics models will help elucidate complex disease pathogenesis, and the predicted patterns in genetic data made by these models will enable the concurrent, more comprehensive statistical testing of multiple aspects of disease genetics predictions, thereby better identifying disease loci. By theoretical human genetics, I intend to encompass all investigations devoted to modeling the heritable architecture underlying disease traits and studies of the resulting principles and dynamics of such models. Hence, the scope of theoretical disease genetics work includes construction and analysis of models describing how disease-predisposing alleles (1) arise, (2) are transmitted across families and populations, and (3) interact with other risk and protective alleles across both the genome and environmental factors to produce disease states. Theoretical work improves insight into viable genetic models of diseases consistent with empirical results from linkage, transmission, and association studies as well as population genetics. Furthermore, understanding the patterns of genetic data expected under realistic disease models will enable more powerful approaches to discover disease-predisposing alleles and additional heritable factors important in common diseases. In spite of the pivotal role of disease genetics theory, such investigation is not particularly vibrant.

A method to estimate the contribution of regional genetic associations to complex traits from summary association statistics

Despite considerable efforts, known genetic associations only explain a small fraction of predicted heritability. Regional associations combine information from multiple contiguous genetic variants and can improve variance explained at established association loci. However, regional associations are not easily amenable to estimation using summary association statistics because of sensitivity to linkage disequilibrium (LD). We now propose a novel method, LD Adjusted Regional Genetic Variance (LARGV), to estimate phenotypic variance explained by regional associations using summary statistics while accounting for LD. Our method is asymptotically equivalent to a multiple linear regression model when no interaction or haplotype effects are present. It has several applications, such as ranking of genetic regions according to variance explained or comparison of variance explained by two or more regions. Using height and BMI data from the Health Retirement Study (N = 7,776), we show that most genetic variance lies in a small proportion of the genome and that previously identified linkage peaks have higher than expected regional variance.

Genetic variants associated with subjective well-being, depressive symptoms, and neuroticism identified through genome-wide analyses

Very few genetic variants have been associated with depression and neuroticism, likely because of limitations on sample size in previous studies. Subjective well-being, a phenotype that is genetically correlated with both of these traits, has not yet been studied with genome-wide data. We conducted genome-wide association studies of three phenotypes: subjective well-being (n = 298,420), depressive symptoms (n = 161,460), and neuroticism (n = 170,911). We identify 3 variants associated with subjective well-being, 2 variants associated with depressive symptoms, and 11 variants associated with neuroticism, including 2 inversion polymorphisms. The two loci associated with depressive symptoms replicate in an independent depression sample. Joint analyses that exploit the high genetic correlations between the phenotypes (|ρ^| ≈ 0.8) strengthen the overall credibility of the findings and allow us to identify additional variants. Across our phenotypes, loci regulating expression in central nervous system and adrenal or pancreas tissues are strongly enriched for association.

Genetic determinants of polycystic ovary syndrome: progress and future directions

The field of the genetics of polycystic ovary syndrome (PCOS) has relatively recently moved into the era of genome-wide association studies. This has led to the discovery of 16 robust loci for PCOS. Some loci contain genes with clear roles in reproductive (LHCGR, FSHR, and FSHB) and metabolic (INSR and HMGA2) dysfunction in the syndrome. The next challenge facing the field is the identification of causal variants and genes and the role they play in PCOS pathophysiology. The potential for gene discovery to improve diagnosis and treatment of PCOS is promising, though there is much to be done in the field before the current findings can be translated to the clinic.

Identification of Common Genetic Variants Influencing Spontaneous Dizygotic Twinning and Female Fertility

Spontaneous dizygotic (DZ) twinning occurs in 1%-4% of women, with familial clustering and unknown physiological pathways and genetic origin. DZ twinning might index increased fertility and has distinct health implications for mother and child. We performed a GWAS in 1,980 mothers of spontaneous DZ twins and 12,953 control subjects. Findings were replicated in a large Icelandic cohort and tested for association across a broad range of fertility traits in women. Two SNPs were identified (rs11031006 near FSHB, p = 1.54 × 10(-9), and rs17293443 in SMAD3, p = 1.57 × 10(-8)) and replicated (p = 3 × 10(-3) and p = 1.44 × 10(-4), respectively). Based on ∼90,000 births in Iceland, the risk of a mother delivering twins increased by 18% for each copy of allele rs11031006-G and 9% for rs17293443-C. A higher polygenic risk score (PRS) for DZ twinning, calculated based on the results of the DZ twinning GWAS, was significantly associated with DZ twinning in Iceland (p = 0.001). A higher PRS was also associated with having children (p = 0.01), greater lifetime parity (p = 0.03), and earlier age at first child (p = 0.02). Allele rs11031006-G was associated with higher serum FSH levels, earlier age at menarche, earlier age at first child, higher lifetime parity, lower PCOS risk, and earlier age at menopause. Conversely, rs17293443-C was associated with later age at last child. We identified robust genetic risk variants for DZ twinning: one near FSHB and a second within SMAD3, the product of which plays an important role in gonadal responsiveness to FSH. These loci contribute to crucial aspects of reproductive capacity and health.

Developing and evaluating polygenic risk prediction models for stratified disease prevention

Knowledge of genetics and its implications for human health is rapidly evolving in accordance with recent events, such as discoveries of large numbers of disease susceptibility loci from genome-wide association studies, the US Supreme Court ruling of the non-patentability of human genes, and the development of a regulatory framework for commercial genetic tests. In anticipation of the increasing relevance of genetic testing for the assessment of disease risks, this Review provides a summary of the methodologies used for building, evaluating and applying risk prediction models that include information from genetic testing and environmental risk factors. Potential applications of models for primary and secondary disease prevention are illustrated through several case studies, and future challenges and opportunities are discussed.

Meta-analysis of Genome-Wide Association Studies for Extraversion: Findings from the Genetics of Personality Consortium

Extraversion is a relatively stable and heritable personality trait associated with numerous psychosocial, lifestyle and health outcomes. Despite its substantial heritability, no genetic variants have been detected in previous genome-wide association (GWA) studies, which may be due to relatively small sample sizes of those studies. Here, we report on a large meta-analysis of GWA studies for extraversion in 63,030 subjects in 29 cohorts. Extraversion item data from multiple personality inventories were harmonized across inventories and cohorts. No genome-wide significant associations were found at the single nucleotide polymorphism (SNP) level but there was one significant hit at the gene level for a long non-coding RNA site (LOC101928162). Genome-wide complex trait analysis in two large cohorts showed that the additive variance explained by common SNPs was not significantly different from zero, but polygenic risk scores, weighted using linkage information, significantly predicted extraversion scores in an independent cohort. These results show that extraversion is a highly polygenic personality trait, with an architecture possibly different from other complex human traits, including other personality traits. Future studies are required to further determine which genetic variants, by what modes of gene action, constitute the heritable nature of extraversion.

Autoimmune diseases - connecting risk alleles with molecular traits of the immune system

Genome-wide strategies have driven the discovery of more than 300 susceptibility loci for autoimmune diseases. However, for almost all loci, understanding of the mechanisms leading to autoimmunity remains limited, and most variants that are likely to be causal are in non-coding regions of the genome. A critical next step will be to identify the in vivo and ex vivo immunophenotypes that are affected by risk variants. To do this, key cell types and cell states that are implicated in autoimmune diseases will need to be defined. Functional genomic annotations from these cell types and states can then be used to resolve candidate genes and causal variants. Together with longitudinal studies, this approach may yield pivotal insights into how autoimmunity is triggered.

Are Genetic Tests for Atherosclerosis Ready for Routine Clinical Use?

In this review, we lay out 3 areas currently being evaluated for incorporation of genetic information into clinical practice related to atherosclerosis. The first, familial hypercholesterolemia, is the clearest case for utility of genetic testing in diagnosis and potentially guiding treatment. Already in use for confirmatory testing of familial hypercholesterolemia and for cascade screening of relatives, genetic testing is likely to expand to help establish diagnoses and facilitate research related to most effective therapies, including new agents, such as PCSK9 inhibitors. The second area, adding genetic information to cardiovascular risk prediction for primary prevention, is not currently recommended. Although identification of additional variants may add substantially to prediction in the future, combining known variants has not yet demonstrated sufficient improvement in prediction for incorporation into commonly used risk scores. The third area, pharmacogenetics, has utility for some therapies today. Future utility for pharmacogenetics will wax or wane depending on the nature of available drugs and therapeutic strategies.

Genome-Wide Estimates of Heritability for Social Demographic Outcomes

An increasing number of studies that are widely used in the demographic research community have collected genome-wide data from their respondents. It is therefore important that demographers have a proper understanding of some of the methodological tools needed to analyze such data. This article details the underlying methodology behind one of the most common techniques for analyzing genome-wide data, genome-wide complex trait analysis (GCTA). GCTA models provide heritability estimates for health, health behaviors, or indicators of attainment using data from unrelated persons. Our goal was to describe this model, highlight the utility of the model for biodemographic research, and demonstrate the performance of this approach under modifications to the underlying assumptions. The first set of modifications involved changing the nature of the genetic data used to compute genetic similarities between individuals (the genetic relationship matrix). We then explored the sensitivity of the model to heteroscedastic errors. In general, GCTA estimates are found to be robust to the modifications proposed here, but we also highlight potential limitations of GCTA estimates.

Polygenic Scores in Epidemiology: Risk Prediction, Etiology, and Clinical Utility

Genes account for a significant proportion of the risk for most common diseases. The genome-wide association scan (GWAS) era of genetic epidemiology has generated a massive amount of data, revolutionized our thinking on the genetic architecture of common diseases and positioned the field to realistically consider risk prediction for common polygenic diseases, such as non-familial cancers, and autoimmune, cardiovascular and psychiatric diseases. Polygenic scoring is an approach that shows promise for understanding the polygenic contribution to common human diseases. This is an approach typically relying on genome-wide SNP data, where a set of SNPs identified in a discovery GWAS are used to construct composite polygenic scores. These scores are then used in additional samples for association testing or risk prediction. This review summarizes the extant literature on the use, power, and accuracy of polygenic scores in studies of the etiology of disease and the promise for disease risk prediction.