TATES: efficient multivariate genotype-phenotype analysis for genome-wide association studies

Tree-based quantitative trait mapping in the presence of external covariates

A central goal in biological and biomedical sciences is to identify the molecular basis of variation in morphological and behavioral traits. Over the last decade, improvements in sequencing technologies coupled with the active development of association mapping methods have made it possible to link single nucleotide polymorphisms (SNPs) and quantitative traits. However, a major limitation of existing methods is that they are often unable to consider complex, but biologically-realistic, scenarios. Previous work showed that association mapping method performance can be improved by using the evolutionary history within each SNP to estimate the covariance structure among randomly-sampled individuals. Here, we propose a method that can be used to analyze a variety of data types, such as data including external covariates, while considering the evolutionary history among SNPs, providing an advantage over existing methods. Existing methods either do so at a computational cost, or fail to model these relationships altogether. By considering the broad-scale relationships among SNPs, the proposed approach is both computationally-feasible and informed by the evolutionary history among SNPs. We show that incorporating an approximate covariance structure during analysis of complex data sets increases performance in quantitative trait mapping, and apply the proposed method to deer mice data.

Finding structure in data using multivariate tree boosting

Technology and collaboration enable dramatic increases in the size of psychological and psychiatric data collections, but finding structure in these large data sets with many collected variables is challenging. Decision tree ensembles such as random forests (Strobl, Malley, & Tutz, 2009) are a useful tool for finding structure, but are difficult to interpret with multiple outcome variables which are often of interest in psychology. To find and interpret structure in data sets with multiple outcomes and many predictors (possibly exceeding the sample size), we introduce a multivariate extension to a decision tree ensemble method called gradient boosted regression trees (Friedman, 2001). Our extension, multivariate tree boosting, is a method for nonparametric regression that is useful for identifying important predictors, detecting predictors with nonlinear effects and interactions without specification of such effects, and for identifying predictors that cause 2 or more outcome variables to covary. We provide the R package "mvtboost" to estimate, tune, and interpret the resulting model, which extends the implementation of univariate boosting in the R package "gbm" (Ridgeway, 2015) to continuous, multivariate outcomes. To illustrate the approach, we analyze predictors of psychological well-being (Ryff & Keyes, 1995). Simulations verify that our approach identifies predictors with nonlinear effects and achieves high prediction accuracy, exceeding or matching the performance of (penalized) multivariate multiple regression and multivariate decision trees over a wide range of conditions. (PsycINFO Database Record

Heritability of Behavioral Problems in 7-Year Olds Based on Shared and Unique Aspects of Parental Views

In studies of child psychopathology, phenotypes of interest are often obtained by parental ratings. When behavioral ratings are obtained in the context of a twin study, this allows for the decomposition of the phenotypic variance, into a genetic and a non-genetic part. If a phenotype is assessed by a single rater, heritability is based on the child’s behavior as expressed in the presence of that particular rater, whereas heritability based on assessments by multiple raters allows for the estimation of the heritability of the phenotype based on rater agreement, as well as the heritability of the rater specific view of the behavior. The aim of this twin study was to quantify the rater common and rater specific contributions to the variation in children’s behavioral problems. We estimated the heritability of maternal and paternal ratings of the Child Behavior Checklist (CBCL) 6–18 empirical emotional and behavioral problem scales in a large sample of 12,310 7-year old Dutch twin pairs. Between 30 and 59% of variation in the part of the phenotype parents agree upon was explained by genetic effects. Common environmental effects that make children in the same family similar explained less variance, ranging between 0 and 32%. For unique views of their children’s behavioral problems, heritability ranged between 0 and 20% for maternal and between 0 and 22% for paternal views. Between 7 and 24% of the variance was accounted for by common environmental factors specific to mother and father’s views. The proportion of rater shared and rater specific heritability can be translated into genetic correlations between parental views and inform the design and interpretation of results of molecular genetic studies. Genetic correlations were nearly or above 0.7 for all CBCL based psychopathology scales. Such large genetic correlations suggest two practical guidelines for genome-wide association studies (GWAS): when studies have collected data from either fathers or mothers, the shared genetic aetiology in parental ratings indicates that is possible to analyze paternal and maternal assessments in a single GWAS or meta-analysis. Secondly, if a study has collected information from both parents, a gain in statistical power may be realized in GWAS by the simultaneous analysis of the data.

Pleiotropic Meta-Analyses of Longitudinal Studies Discover Novel Genetic Variants Associated with Age-Related Diseases

Age-related diseases may result from shared biological mechanisms in intrinsic processes of aging. Genetic effects on age-related diseases are often modulated by environmental factors due to their little contribution to fitness or are mediated through certain endophenotypes. Identification of genetic variants with pleiotropic effects on both common complex diseases and endophenotypes may reveal potential conflicting evolutionary pressures and deliver new insights into shared genetic contribution to healthspan and lifespan. Here, we performed pleiotropic meta-analyses of genetic variants using five NIH-funded datasets by integrating univariate summary statistics for age-related diseases and endophenotypes. We investigated three groups of traits: (1) endophenotypes such as blood glucose, blood pressure, lipids, hematocrit, and body mass index, (2) time-to-event outcomes such as the age-at-onset of diabetes mellitus (DM), cancer, cardiovascular diseases (CVDs) and neurodegenerative diseases (NDs), and (3) both combined. In addition to replicating previous findings, we identify seven novel genome-wide significant loci (< 5e-08), out of which five are low-frequency variants. Specifically, from Group 2, we find rs7632505 on 3q21.1 in SEMA5B, rs460976 on 21q22.3 (1 kb from TMPRSS2) and rs12420422 on 11q24.1 predominantly associated with a variety of CVDs, rs4905014 in ITPK1 associated with stroke and heart failure, rs7081476 on 10p12.1 in ANKRD26 associated with multiple diseases including DM, CVDs, and NDs. From Group 3, we find rs8082812 on 18p11.22 and rs1869717 on 4q31.3 associated with both endophenotypes and CVDs. Our follow-up analyses show that rs7632505, rs4905014, and rs8082812 have age-dependent effects on coronary heart disease or stroke. Functional annotation suggests that most of these SNPs are within regulatory regions or DNase clusters and in linkage disequilibrium with expression quantitative trait loci, implying their potential regulatory influence on the expression of nearby genes. Our mediation analyses suggest that the effects of some SNPs are mediated by specific endophenotypes. In conclusion, these findings indicate that loci with pleiotropic effects on age-related disorders tend to be enriched in genes involved in underlying mechanisms potentially related to nervous, cardiovascular and immune system functions, stress resistance, inflammation, ion channels and hematopoiesis, supporting the hypothesis of shared pathological role of infection, and inflammation in chronic age-related diseases.

An Adaptive Fisher's Combination Method for Joint Analysis of Multiple Phenotypes in Association Studies

Currently, the analyses of most genome-wide association studies (GWAS) have been performed on a single phenotype. There is increasing evidence showing that pleiotropy is a widespread phenomenon in complex diseases. Therefore, using only one single phenotype may lose statistical power to identify the underlying genetic mechanism. There is an increasing need to develop and apply powerful statistical tests to detect association between multiple phenotypes and a genetic variant. In this paper, we develop an Adaptive Fisher’s Combination (AFC) method for joint analysis of multiple phenotypes in association studies. The AFC method combines p-values obtained in standard univariate GWAS by using the optimal number of p-values which is determined by the data. We perform extensive simulations to evaluate the performance of the AFC method and compare the power of our method with the powers of TATES, Tippett’s method, Fisher’s combination test, MANOVA, MultiPhen, and SUMSCORE. Our simulation studies show that the proposed method has correct type I error rates and is either the most powerful test or comparable with the most powerful test. Finally, we illustrate our proposed methodology by analyzing whole-genome genotyping data from a lung function study.

Accelerating Gene Discovery by Phenotyping Whole-Genome Sequenced Multi-mutation Strains and Using the Sequence Kernel Association Test (SKAT)

Forward genetic screens represent powerful, unbiased approaches to uncover novel components in any biological process. Such screens suffer from a major bottleneck, however, namely the cloning of corresponding genes causing the phenotypic variation. Reverse genetic screens have been employed as a way to circumvent this issue, but can often be limited in scope. Here we demonstrate an innovative approach to gene discovery. Using C. elegans as a model system, we used a whole-genome sequenced multi-mutation library, from the Million Mutation Project, together with the Sequence Kernel Association Test (SKAT), to rapidly screen for and identify genes associated with a phenotype of interest, namely defects in dye-filling of ciliated sensory neurons. Such anomalies in dye-filling are often associated with the disruption of cilia, organelles which in humans are implicated in sensory physiology (including vision, smell and hearing), development and disease. Beyond identifying several well characterised dye-filling genes, our approach uncovered three genes not previously linked to ciliated sensory neuron development or function. From these putative novel dye-filling genes, we confirmed the involvement of BGNT-1.1 in ciliated sensory neuron function and morphogenesis. BGNT-1.1 functions at the trans-Golgi network of sheath cells (glia) to influence dye-filling and cilium length, in a cell non-autonomous manner. Notably, BGNT-1.1 is the orthologue of human B3GNT1/B4GAT1, a glycosyltransferase associated with Walker-Warburg syndrome (WWS). WWS is a multigenic disorder characterised by muscular dystrophy as well as brain and eye anomalies. Together, our work unveils an effective and innovative approach to gene discovery, and provides the first evidence that B3GNT1-associated Walker-Warburg syndrome may be considered a ciliopathy.

Model organisms are useful tools for uncovering new genes involved in a biological process via genetic screens. Such an approach is powerful, but suffers from drawbacks that can slow down gene discovery. In forward genetics screens, difficult-to-map phenotypes present daunting challenges, and whole-genome coverage can be equally challenging for reverse genetic screens where typically only a single gene’s function is assayed per strain. Here, we show a different approach which includes positive aspects of forward (high-coverage, randomly-induced mutations) and reverse genetics (prior knowledge of gene disruption) to accelerate gene discovery. We paired a whole-genome sequenced multi-mutation C. elegans library with a rare-variant associated test to rapidly identify genes associated with a phenotype of interest: defects in sensory neurons bearing sensory organelles called cilia, via a simple dye-filling assay to probe the form and function of these cells. We found two well characterised dye-filling genes and three genes, not previously linked to ciliated sensory neuron development or function, that were associated with dye-filling defects. We reveal that disruption of one of these (BGNT-1.1), whose human orthologue is associated with Walker-Warburg syndrome, results in abrogated uptake of dye and cilia length defects. We believe that our novel approach is useful for any organism with a small genome that can be quickly sequenced and where many mutant strains can be easily isolated and phenotyped, such as Drosophila and Arabidopsis.

metaCCA: summary statistics-based multivariate meta-analysis of genome-wide association studies using canonical correlation analysis

MOTIVATION

A dominant approach to genetic association studies is to perform univariate tests between genotype-phenotype pairs. However, analyzing related traits together increases statistical power, and certain complex associations become detectable only when several variants are tested jointly. Currently, modest sample sizes of individual cohorts, and restricted availability of individual-level genotype-phenotype data across the cohorts limit conducting multivariate tests.

RESULTS

We introduce metaCCA, a computational framework for summary statistics-based analysis of a single or multiple studies that allows multivariate representation of both genotype and phenotype. It extends the statistical technique of canonical correlation analysis to the setting where original individual-level records are not available, and employs a covariance shrinkage algorithm to achieve robustness.Multivariate meta-analysis of two Finnish studies of nuclear magnetic resonance metabolomics by metaCCA, using standard univariate output from the program SNPTEST, shows an excellent agreement with the pooled individual-level analysis of original data. Motivated by strong multivariate signals in the lipid genes tested, we envision that multivariate association testing using metaCCA has a great potential to provide novel insights from already published summary statistics from high-throughput phenotyping technologies.

AVAILABILITY AND IMPLEMENTATION

Code is available at https://github.com/aalto-ics-kepaco

CONTACTS

anna.cichonska@helsinki.fi or matti.pirinen@helsinki.fi

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Power Comparisons of Methods for Joint Association Analysis of Multiple Phenotypes

BACKGROUND/AIMS

Genome-wide association studies (GWAS) have identified many variants that each affect multiple phenotypes, which suggests that pleiotropic effects on human complex phenotypes may be widespread. Therefore, statistical methods that can jointly analyze multiple phenotypes in GWAS may have advantages over analyzing each phenotype individually. Several statistical methods have been developed to utilize such multivariate phenotypes in genetic association studies; however, the performance of these methods under different scenarios is largely unknown. Our goal was to provide researchers with useful guidelines on selecting statistical methods for the application of real data to multiple phenotypes.

METHODS

In this study, we evaluated the performance of some of the existing methods for association studies using multiple phenotypes. These methods included the O'Brien method (OB), cross-validation method (CV), optimal weight method (OW), Trait-based Association Test that uses Extended Simes procedure (TATES), principal components of heritability (PCH), canonical correlation analysis (CCA), multivariate analysis of variance (MANOVA), and a joint model of multiple phenotypes (MultiPhen). We used simulation studies to compare the powers of these methods under a variety of scenarios, including different numbers of phenotypes, different values of between-phenotype correlation, different minor allele frequencies, and different mean and variance models.

RESULTS AND CONCLUSION

Our simulation results show that there is no single method with consistently good performance among all the scenarios. Each method has its own advantages and disadvantages.

Family-Based Rare Variant Association Analysis: A Fast and Efficient Method of Multivariate Phenotype Association Analysis

Family-based designs have been repeatedly shown to be powerful in detecting the significant rare variants associated with human diseases. Furthermore, human diseases are often defined by the outcomes of multiple phenotypes, and thus we expect multivariate family-based analyses may be very efficient in detecting associations with rare variants. However, few statistical methods implementing this strategy have been developed for family-based designs. In this report, we describe one such implementation: the multivariate family-based rare variant association tool (mFARVAT). mFARVAT is a quasi-likelihood-based score test for rare variant association analysis with multiple phenotypes, and tests both homogeneous and heterogeneous effects of each variant on multiple phenotypes. Simulation results show that the proposed method is generally robust and efficient for various disease models, and we identify some promising candidate genes associated with chronic obstructive pulmonary disease. The software of mFARVAT is freely available at http://healthstat.snu.ac.kr/software/mfarvat/, implemented in C++ and supported on Linux and MS Windows.

Phenome-wide association study maps new diseases to the human major histocompatibility complex region

BACKGROUND

Over 160 disease phenotypes have been mapped to the major histocompatibility complex (MHC) region on chromosome 6 by genome-wide association study (GWAS), suggesting that the MHC region as a whole may be involved in the aetiology of many phenotypes, including unstudied diseases. The phenome-wide association study (PheWAS), a powerful and complementary approach to GWAS, has demonstrated its ability to discover and rediscover genetic associations. The objective of this study is to comprehensively investigate the MHC region by PheWAS to identify new phenotypes mapped to this genetically important region.

METHODS

In the current study, we systematically explored the MHC region using PheWAS to associate 2692 MHC-linked variants (minor allele frequency ≥0.01) with 6221 phenotypes in a cohort of 7481 subjects from the Marshfield Clinic Personalized Medicine Research Project.

RESULTS

Findings showed that expected associations previously identified by GWAS could be identified by PheWAS (eg, psoriasis, ankylosing spondylitis, type I diabetes and coeliac disease) with some having strong cross-phenotype associations potentially driven by pleiotropic effects. Importantly, novel associations with eight diseases not previously assessed by GWAS (eg, lichen planus) were also identified and replicated in an independent population. Many of these associated diseases appear to be immune-related disorders. Further assessment of these diseases in 16 484 Marshfield Clinic twins suggests that some of these diseases, including lichen planus, may have genetic aetiologies.

CONCLUSIONS

These results demonstrate that the PheWAS approach is a powerful and novel method to discover SNP-disease associations, and is ideal when characterising cross-phenotype associations, and further emphasise the importance of the MHC region in human health and disease.

The CHRM3 gene is implicated in abnormal thalamo-orbital frontal cortex functional connectivity in first-episode treatment-naive patients with schizophrenia

BACKGROUND

The genetic influences in human brain structure and function and impaired functional connectivities are the hallmarks of the schizophrenic brain. To explore how common genetic variants affect the connectivities in schizophrenia, we applied genome-wide association studies assaying the abnormal neural connectivities in schizophrenia as quantitative traits.

METHOD

We recruited 161 first-onset and treatment-naive patients with schizophrenia and 150 healthy controls. All the participants underwent scanning with a 3 T-magnetic resonance imaging scanner to acquire structural and functional imaging data and genotyping using the HumanOmniZhongHua-8 BeadChip. The brain-wide association study approach was employed to account for the inherent modular nature of brain connectivities.

RESULTS

We found differences in four abnormal functional connectivities [left rectus to left thalamus (REC.L-THA.L), left rectus to right thalamus (REC.L-THA.R), left superior orbital cortex to left thalamus (ORBsup.L-THA.L) and left superior orbital cortex to right thalamus (ORBsup.L-THA.R)] between the two groups. Univariate single nucleotide polymorphism (SNP)-based association revealed that the SNP rs6800381, located nearest to the CHRM3 (cholinergic receptor, muscarinic 3) gene, reached genomic significance (p = 1.768 × 10-8) using REC.L-THA.R as the phenotype. Multivariate gene-based association revealed that the FAM12A (family with sequence similarity 12, member A) gene nearly reached genomic significance (nominal p = 2.22 × 10-6, corrected p = 0.05).

CONCLUSIONS

Overall, we identified the first evidence that the CHRM3 gene plays a role in abnormal thalamo-orbital frontal cortex functional connectivity in first-episode treatment-naive patients with schizophrenia. Identification of these genetic variants using neuroimaging genetics provides insights into the causes of variability in human brain development, and may help us determine the mechanisms of dysfunction in schizophrenia.

Multivariate Gene-Based Association Test on Family Data in MGAS

In analyses of unrelated individuals, the program multivariate gene-based association test by extended Simes (MGAS), which facilitates multivariate gene-based association testing, was shown to have correct Type I error rate and superior statistical power compared to other multivariate gene-based approaches. Here we show, through simulation, that MGAS can also be applied to data including genetically related subjects (e.g., family data), by using p value information obtained in Plink or in generalized estimating equations (with the 'exchangeable' working correlation matrix), both of which account for the family structure on a univariate single nucleotide polymorphism-based level by applying a sandwich correction of standard errors. We show that when applied to family-data, MGAS has correct Type I error rate, and given the details of the simulation setup, adequate power. Application of MGAS to seven eye measurement phenotypes showed statistically significant association with two genes that were not discovered in previous univariate analyses of a composite score. We conclude that MGAS is a useful and convenient tool for multivariate gene-based genome-wide association analysis in both unrelated and related individuals.

Joint Analysis of Multiple Traits in Rare Variant Association Studies

The joint analysis of multiple traits has recently become popular since it can increase statistical power to detect genetic variants and there is increasing evidence showing that pleiotropy is a widespread phenomenon in complex diseases. Currently, the majority of existing methods for the joint analysis of multiple traits test association between one common variant and multiple traits. However, the variant-by-variant methods for common variant association studies may not be optimal for rare variant association studies due to the allelic heterogeneity as well as the extreme rarity of individual variants. Current statistical methods for rare variant association studies are for one single trait only. In this paper, we propose an adaptive weighting reverse regression (AWRR) method to test association between multiple traits and rare variants in a genomic region. AWRR is robust to the directions of effects of causal variants and is also robust to the directions of association of traits. Using extensive simulation studies, we compare the performance of AWRR with canonical correlation analysis (CCA), Single-TOW, and the weighted sum reverse regression (WSRR). Our results show that, in all of the simulation scenarios, AWRR is consistently more powerful than CCA. In most scenarios, AWRR is more powerful than Single-TOW and WSRR.

Joint Analysis of Multiple Traits Using "Optimal" Maximum Heritability Test

The joint analysis of multiple traits has recently become popular since it can increase statistical power to detect genetic variants and there is increasing evidence showing that pleiotropy is a widespread phenomenon in complex diseases. Currently, most of existing methods use all of the traits for testing the association between multiple traits and a single variant. However, those methods for association studies may lose power in the presence of a large number of noise traits. In this paper, we propose an “optimal” maximum heritability test (MHT-O) to test the association between multiple traits and a single variant. MHT-O includes a procedure of deleting traits that have weak or no association with the variant. Using extensive simulation studies, we compare the performance of MHT-O with MHT, Trait-based Association Test uses Extended Simes procedure (TATES), SUM_SCORE and MANOVA. Our results show that, in all of the simulation scenarios, MHT-O is either the most powerful test or comparable to the most powerful test among the five tests we compared.

Unravelling the human genome-phenome relationship using phenome-wide association studies

Advances in genotyping technology have, over the past decade, enabled the focused search for common genetic variation associated with human diseases and traits. With the recently increased availability of detailed phenotypic data from electronic health records and epidemiological studies, the impact of one or more genetic variants on the phenome is starting to be characterized both in clinical and population-based settings using phenome-wide association studies (PheWAS). These studies reveal a number of challenges that will need to be overcome to unlock the full potential of PheWAS for the characterization of the complex human genome-phenome relationship.

Sequence Kernel Association Test of Multiple Continuous Phenotypes

Genetic studies often collect multiple correlated traits, which could be analyzed jointly to increase power by aggregating multiple weak effects and provide additional insights into the etiology of complex human diseases. Existing methods for multiple trait association tests have primarily focused on common variants. There is a surprising dearth of published methods for testing the association of rare variants with multiple correlated traits. In this paper, we extend the commonly used sequence kernel association test (SKAT) for single-trait analysis to test for the joint association of rare variant sets with multiple traits. We investigate the performance of the proposed method through extensive simulation studies. We further illustrate its usefulness with application to the analysis of diabetes-related traits in the Atherosclerosis Risk in Communities (ARIC) Study. We identified an exome-wide significant rare variant set in the gene YAP1 worthy of further investigations.

An efficient genome-wide association test for multivariate phenotypes based on the Fisher combination function

BACKGROUND

In genome-wide association studies (GWAS) for complex diseases, the association between a SNP and each phenotype is usually weak. Combining multiple related phenotypic traits can increase the power of gene search and thus is a practically important area that requires methodology work. This study provides a comprehensive review of existing methods for conducting GWAS on complex diseases with multiple phenotypes including the multivariate analysis of variance (MANOVA), the principal component analysis (PCA), the generalizing estimating equations (GEE), the trait-based association test involving the extended Simes procedure (TATES), and the classical Fisher combination test. We propose a new method that relaxes the unrealistic independence assumption of the classical Fisher combination test and is computationally efficient. To demonstrate applications of the proposed method, we also present the results of statistical analysis on the Study of Addiction: Genetics and Environment (SAGE) data.

RESULTS

Our simulation study shows that the proposed method has higher power than existing methods while controlling for the type I error rate. The GEE and the classical Fisher combination test, on the other hand, do not control the type I error rate and thus are not recommended. In general, the power of the competing methods decreases as the correlation between phenotypes increases. All the methods tend to have lower power when the multivariate phenotypes come from long tailed distributions. The real data analysis also demonstrates that the proposed method allows us to compare the marginal results with the multivariate results and specify which SNPs are specific to a particular phenotype or contribute to the common construct.

CONCLUSIONS

The proposed method outperforms existing methods in most settings and also has great applications in GWAS on complex diseases with multiple phenotypes such as the substance abuse disorders.

USAT: A Unified Score-Based Association Test for Multiple Phenotype-Genotype Analysis

Genome-wide association studies (GWASs) for complex diseases often collect data on multiple correlated endo-phenotypes. Multivariate analysis of these correlated phenotypes can improve the power to detect genetic variants. Multivariate analysis of variance (MANOVA) can perform such association analysis at a GWAS level, but the behavior of MANOVA under different trait models has not been carefully investigated. In this paper, we show that MANOVA is generally very powerful for detecting association but there are situations, such as when a genetic variant is associated with all the traits, where MANOVA may not have any detection power. In these situations, marginal model based methods, however, perform much better than multivariate methods. We investigate the behavior of MANOVA, both theoretically and using simulations, and derive the conditions where MANOVA loses power. Based on our findings, we propose a unified score-based test statistic USAT that can perform better than MANOVA in such situations and nearly as well as MANOVA elsewhere. Our proposed test reports an approximate asymptotic P-value for association and is computationally very efficient to implement at a GWAS level. We have studied through extensive simulations the performance of USAT, MANOVA, and other existing approaches and demonstrated the advantage of using the USAT approach to detect association between a genetic variant and multivariate phenotypes. We applied USAT to data from three correlated traits collected on 5, 816 Caucasian individuals from the Atherosclerosis Risk in Communities (ARIC, The ARIC Investigators []) Study and detected some interesting associations.

Powerful Tests for Multi-Marker Association Analysis Using Ensemble Learning

Multi-marker approaches have received a lot of attention recently in genome wide association studies and can enhance power to detect new associations under certain conditions. Gene-, gene-set- and pathway-based association tests are increasingly being viewed as useful supplements to the more widely used single marker association analysis which have successfully uncovered numerous disease variants. A major drawback of single-marker based methods is that they do not look at the joint effects of multiple genetic variants which individually may have weak or moderate signals. Here, we describe novel tests for multi-marker association analyses that are based on phenotype predictions obtained from machine learning algorithms. Instead of assuming a linear or logistic regression model, we propose the use of ensembles of diverse machine learning algorithms for prediction. We show that phenotype predictions obtained from ensemble learning algorithms provide a new framework for multi-marker association analysis. They can be used for constructing tests for the joint association of multiple variants, adjusting for covariates and testing for the presence of interactions. To demonstrate the power and utility of this new approach, we first apply our method to simulated SNP datasets. We show that the proposed method has the correct Type-1 error rates and can be considerably more powerful than alternative approaches in some situations. Then, we apply our method to previously studied asthma-related genes in 2 independent asthma cohorts to conduct association tests.

Summaries of plenary, symposia, and oral sessions at the XXII World Congress of Psychiatric Genetics, Copenhagen, Denmark, 12-16 October 2014

The XXII World Congress of Psychiatric Genetics, sponsored by the International Society of Psychiatric Genetics, took place in Copenhagen, Denmark, on 12-16 October 2014. A total of 883 participants gathered to discuss the latest findings in the field. The following report was written by student and postdoctoral attendees. Each was assigned one or more sessions as a rapporteur. This manuscript represents topics covered in most, but not all of the oral presentations during the conference, and contains some of the major notable new findings reported.

Dimensionality and Genetic Correlates of Problem Behavior in Low-Income African American Adolescents

Researchers have long observed that problem behaviors tend to cluster together, particularly among adolescents. Epidemiological studies have suggested that this covariation is due, in part, to common genetic influences, and a number of plausible candidates have emerged as targets for investigation. To date, however, genetic association studies of these behaviors have focused mostly on unidimensional models of individual phenotypes within European American samples. Herein, we compared a series of confirmatory factor models to best characterize the structure of problem behavior (alcohol and marijuana use, sexual behavior, and disruptive behavior) within a representative community-based sample of 592 low-income African American adolescents (50.3% female), ages 13 to 18. We further explored the extent to which 3 genes previously implicated for their role in similar behavioral dimensions (CHRM2, GABRA2, and OPRM1) independently accounted for variance within factors specified in the best-fitting model. Supplementary analyses were conducted to derive comparative estimates for the predictive utility of these genes in more traditional unidimensional models. Findings provide initial evidence for a bifactor structure of problem behavior among African American adolescents and highlight novel genetic correlates of specific behavioral dimensions otherwise undetected in an orthogonal syndromal factor. Implications of this approach include increased precision in the assessment of problem behavior, with corresponding increases in the reliability and validity of identified genetic associations. As a corollary, the comparison of primary and supplementary association analyses illustrates the potential for overlooking and/or overinterpreting meaningful genetic effects when failing to adequately account for phenotypic complexity.

Semiparametric Allelic Tests for Mapping Multiple Phenotypes: Binomial Regression and Mahalanobis Distance

Binary phenotypes commonly arise due to multiple underlying quantitative precursors and genetic variants may impact multiple traits in a pleiotropic manner. Hence, simultaneously analyzing such correlated traits may be more powerful than analyzing individual traits. Various genotype-level methods, e.g., MultiPhen (O'Reilly et al. []), have been developed to identify genetic factors underlying a multivariate phenotype. For univariate phenotypes, the usefulness and applicability of allele-level tests have been investigated. The test of allele frequency difference among cases and controls is commonly used for mapping case-control association. However, allelic methods for multivariate association mapping have not been studied much. In this article, we explore two allelic tests of multivariate association: one using a Binomial regression model based on inverted regression of genotype on phenotype (Binomial regression-based Association of Multivariate Phenotypes [BAMP]), and the other employing the Mahalanobis distance between two sample means of the multivariate phenotype vector for two alleles at a single-nucleotide polymorphism (Distance-based Association of Multivariate Phenotypes [DAMP]). These methods can incorporate both discrete and continuous phenotypes. Some theoretical properties for BAMP are studied. Using simulations, the power of the methods for detecting multivariate association is compared with the genotype-level test MultiPhen's. The allelic tests yield marginally higher power than MultiPhen for multivariate phenotypes. For one/two binary traits under recessive mode of inheritance, allelic tests are found to be substantially more powerful. All three tests are applied to two different real data and the results offer some support for the simulation study. We propose a hybrid approach for testing multivariate association that implements MultiPhen when Hardy-Weinberg Equilibrium (HWE) is violated and BAMP otherwise, because the allelic approaches assume HWE.

An Adaptive Association Test for Multiple Phenotypes with GWAS Summary Statistics

We study the problem of testing for single marker-multiple phenotype associations based on genome-wide association study (GWAS) summary statistics without access to individual-level genotype and phenotype data. For most published GWASs, because obtaining summary data is substantially easier than accessing individual-level phenotype and genotype data, while often multiple correlated traits have been collected, the problem studied here has become increasingly important. We propose a powerful adaptive test and compare its performance with some existing tests. We illustrate its applications to analyses of a meta-analyzed GWAS dataset with three blood lipid traits and another with sex-stratified anthropometric traits, and further demonstrate its potential power gain over some existing methods through realistic simulation studies. We start from the situation with only one set of (possibly meta-analyzed) genome-wide summary statistics, then extend the method to meta-analysis of multiple sets of genome-wide summary statistics, each from one GWAS. We expect the proposed test to be useful in practice as more powerful than or complementary to existing methods.

Phenome-Wide Association Studies: Embracing Complexity for Discovery

The inherent complexity of biological systems can be leveraged for a greater understanding of the impact of genetic architecture on outcomes, traits, and pharmacological response. The genome-wide association study (GWAS) approach has well-developed methods and relatively straight-forward methodologies; however, the bigger picture of the impact of genetic architecture on phenotypic outcome still remains to be elucidated even with an ever-growing number of GWAS performed. Greater consideration of the complexity of biological processes, using more data from the phenome, exposome, and diverse -omic resources, including considering the interplay of pleiotropy and genetic interactions, may provide additional leverage for making the most of the incredible wealth of information available for study. Here, we describe how incorporating greater complexity into analyses through the use of additional phenotypic data and widespread deployment of phenome-wide association studies may provide new insights into genetic factors influencing diseases, traits, and pharmacological response.

Psychometric precision in phenotype definition is a useful step in molecular genetic investigation of psychiatric disorders

Affective disorders are highly heritable, but few genetic risk variants have been consistently replicated in molecular genetic association studies. The common method of defining psychiatric phenotypes in molecular genetic research is either a summation of symptom scores or binary threshold score representing the risk of diagnosis. Psychometric latent variable methods can improve the precision of psychiatric phenotypes, especially when the data structure is not straightforward. Using data from the British 1946 birth cohort, we compared summary scores with psychometric modeling based on the General Health Questionnaire (GHQ-28) scale for affective symptoms in an association analysis of 27 candidate genes (249 single-nucleotide polymorphisms (SNPs)). The psychometric method utilized a bi-factor model that partitioned the phenotype variances into five orthogonal latent variable factors, in accordance with the multidimensional data structure of the GHQ-28 involving somatic, social, anxiety and depression domains. Results showed that, compared with the summation approach, the affective symptoms defined by the bi-factor psychometric model had a higher number of associated SNPs of larger effect sizes. These results suggest that psychometrically defined mental health phenotypes can reflect the dimensions of complex phenotypes better than summation scores, and therefore offer a useful approach in genetic association investigations.

Statistical methods for association tests of multiple continuous traits in genome-wide association studies

Multiple correlated traits are often collected in genetic studies. The joint analysis of multiple traits could have increased power by aggregating multiple weak effects and offer additional insights into the aetiology of complex human diseases by revealing pleiotropic variants. We propose to study multivariate test statistics to detect single nucleotide polymorphism (SNP) association with multiple correlated traits. Most existing methods have been based on the generalized estimating equation (GEE) approach without explicitly modelling the trait correlations. In this article, we explore an alternative likelihood-based framework to test the multiple trait associations. It is based on the familiar multinomial logistic regression modelling of genotypes, which can be readily implemented using widely available software, and offers very competitive performance. We demonstrate through extensive numerical studies that the proposed method has competitive performance. Its usefulness is further illustrated with application to association analysis of diabetes-related traits in the Atherosclerosis Risk in Communities (ARIC) Study.

The effects of chromatin organization on variation in mutation rates in the genome

The variation in local rates of mutations can affect both the evolution of genes and their function in normal and cancer cells. Deciphering the molecular determinants of this variation will be aided by the elucidation of distinct types of mutations, as they differ in regional preferences and in associations with genomic features. Chromatin organization contributes to regional variation in mutation rates, but its contribution differs among mutation types. In both germline and somatic mutations, base substitutions are more abundant in regions of closed chromatin, perhaps reflecting error accumulation late in replication. By contrast, a distinctive mutational state with very high levels of insertions and deletions (indels) and substitutions is enriched in regions of open chromatin. These associations indicate an intricate interplay between the nucleotide sequence of DNA and its dynamic packaging into chromatin, and have important implications for current biomedical research. This Review focuses on recent studies showing associations between chromatin state and mutation rates, including pairwise and multivariate investigations of germline and somatic (particularly cancer) mutations.

Hippocampal transcriptome-guided genetic analysis of correlated episodic memory phenotypes in Alzheimer's disease

As the most common type of dementia, Alzheimer's disease (AD) is a neurodegenerative disorder initially manifested by impaired memory performances. While the diagnosis information indicates a dichotomous status of a patient, memory scores have the potential to capture the continuous nature of the disease progression and may provide more insights into the underlying mechanism. In this work, we performed a targeted genetic study of memory scores on an AD cohort to identify the associations between a set of genes highly expressed in the hippocampal region and seven cognitive scores related to episodic memory. Both main effects and interaction effects of the targeted genetic markers on these correlated memory scores were examined. In addition to well-known AD genetic markers APOE and TOMM40, our analysis identified a new risk gene NAV2 through the gene-level main effect analysis. NAV2 was found to be significantly and consistently associated with all seven episodic memory scores. Genetic interaction analysis also yielded a few promising hits warranting further investigation, especially for the RAVLT list B Score.

Family-based association analysis: a fast and efficient method of multivariate association analysis with multiple variants

BACKGROUND

Many disease phenotypes are outcomes of the complicated interplay between multiple genes, and multiple phenotypes are affected by a single or multiple genotypes. Therefore, joint analysis of multiple phenotypes and multiple markers has been considered as an efficient strategy for genome-wide association analysis, and in this work we propose an omnibus family-based association test for the joint analysis of multiple genotypes and multiple phenotypes.

RESULTS

The proposed test can be applied for both quantitative and dichotomous phenotypes, and it is robust under the presence of population substructure, as long as large-scale genomic data is available. Using simulated data, we showed that our method is statistically more efficient than the existing methods, and the practical relevance is illustrated by application of the approach to obesity-related phenotypes.

CONCLUSIONS

The proposed method may be more statistically efficient than the existing methods. The application was developed in C++ and is available at the following URL: http://healthstat.snu.ac.kr/software/mfqls/ .

Common genetic variants influence human subcortical brain structures

The highly complex structure of the human brain is strongly shaped by genetic influences. Subcortical brain regions form circuits with cortical areas to coordinate movement, learning, memory and motivation, and altered circuits can lead to abnormal behaviour and disease. To investigate how common genetic variants affect the structure of these brain regions, here we conduct genome-wide association studies of the volumes of seven subcortical regions and the intracranial volume derived from magnetic resonance images of 30,717 individuals from 50 cohorts. We identify five novel genetic variants influencing the volumes of the putamen and caudate nucleus. We also find stronger evidence for three loci with previously established influences on hippocampal volume and intracranial volume. These variants show specific volumetric effects on brain structures rather than global effects across structures. The strongest effects were found for the putamen, where a novel intergenic locus with replicable influence on volume (rs945270; P = 1.08 × 10(-33); 0.52% variance explained) showed evidence of altering the expression of the KTN1 gene in both brain and blood tissue. Variants influencing putamen volume clustered near developmental genes that regulate apoptosis, axon guidance and vesicle transport. Identification of these genetic variants provides insight into the causes of variability in human brain development, and may help to determine mechanisms of neuropsychiatric dysfunction.

Meta-analysis of correlated traits via summary statistics from GWASs with an application in hypertension

Genome-wide association studies (GWASs) have identified many genetic variants underlying complex traits. Many detected genetic loci harbor variants that associate with multiple-even distinct-traits. Most current analysis approaches focus on single traits, even though the final results from multiple traits are evaluated together. Such approaches miss the opportunity to systemically integrate the phenome-wide data available for genetic association analysis. In this study, we propose a general approach that can integrate association evidence from summary statistics of multiple traits, either correlated, independent, continuous, or binary traits, which might come from the same or different studies. We allow for trait heterogeneity effects. Population structure and cryptic relatedness can also be controlled. Our simulations suggest that the proposed method has improved statistical power over single-trait analysis in most of the cases we studied. We applied our method to the Continental Origins and Genetic Epidemiology Network (COGENT) African ancestry samples for three blood pressure traits and identified four loci (CHIC2, HOXA-EVX1, IGFBP1/IGFBP3, and CDH17; p < 5.0 × 10(-8)) associated with hypertension-related traits that were missed by a single-trait analysis in the original report. Six additional loci with suggestive association evidence (p < 5.0 × 10(-7)) were also observed, including CACNA1D and WNT3. Our study strongly suggests that analyzing multiple phenotypes can improve statistical power and that such analysis can be executed with the summary statistics from GWASs. Our method also provides a way to study a cross phenotype (CP) association by using summary statistics from GWASs of multiple phenotypes.

Pathway-based association study of multiple candidate genes and multiple traits using structural equation models

There is increasing interest in the joint analysis of multiple genetic variants from multiple genes and multiple correlated quantitative traits in association studies. The classical approach involves testing univariate associations between genotypes and phenotypes and correcting for multiple testing that results in loss of power to detect associations. In this paper, we propose modeling complex relationships between genetic variants in candidate genes and measured correlated traits using structural equation models (SEM), taking advantage of prior knowledge on clinical and genetic pathways. We adopt generalized structured component analysis (GSCA) as an approach to SEM and develop a single association test between multiple genetic variants in a gene and a set of correlated traits, taking into account all available data from other genes and other traits. The performance of this test is investigated by simulations. We apply the proposed method to the Quebec Child and Adolescent Health and Social Survey (1999) data to investigate genetic associations with cardiovascular disease-related traits.

MGAS: a powerful tool for multivariate gene-based genome-wide association analysis

Motivation: Standard genome-wide association studies, testing the association between one phenotype and a large number of single nucleotide polymorphisms (SNPs), are limited in two ways: (i) traits are often multivariate, and analysis of composite scores entails loss in statistical power and (ii) gene-based analyses may be preferred, e.g. to decrease the multiple testing problem.

Results: Here we present a new method, multivariate gene-based association test by extended Simes procedure (MGAS), that allows gene-based testing of multivariate phenotypes in unrelated individuals. Through extensive simulation, we show that under most trait-generating genotype–phenotype models MGAS has superior statistical power to detect associated genes compared with gene-based analyses of univariate phenotypic composite scores (i.e. GATES, multiple regression), and multivariate analysis of variance (MANOVA). Re-analysis of metabolic data revealed 32 False Discovery Rate controlled genome-wide significant genes, and 12 regions harboring multiple genes; of these 44 regions, 30 were not reported in the original analysis.

Conclusion: MGAS allows researchers to conduct their multivariate gene-based analyses efficiently, and without the loss of power that is often associated with an incorrectly specified genotype–phenotype models.

Availability and implementation: MGAS is freely available in KGG v3.0 (http://statgenpro.psychiatry.hku.hk/limx/kgg/download.php). Access to the metabolic dataset can be requested at dbGaP (https://dbgap.ncbi.nlm.nih.gov/). The R-simulation code is available from http://ctglab.nl/people/sophie_van_der_sluis.

Contact: mxli@hku.hk

Supplementary information: Supplementary data are available at Bioinformatics online.

An alternative approach to multiple testing for methylation QTL mapping reduces the proportion of falsely identified CpGs

INTRODUCTION

An increasing number of studies investigates the influence of local genetic variation on DNA methylation levels, so-called in cis methylation quantitative trait loci (meQTLs). A common multiple testing approach in genome-wide cis meQTL studies limits the false discovery rate (FDR) among all CpG-SNP pairs to 0.05 and reports on CpGs from the significant CpG-SNP pairs. However, a statistical test for each CpG is not performed, potentially increasing the proportion of CpGs falsely reported on. Here, we presented an alternative approach that properly control for multiple testing at the CpG level.

RESULTS

We performed cis meQTL mapping for varying window sizes using publicly available single-nucleotide polymorphism (SNP) and 450 kb data, extracting the CpGs from the significant CpG-SNP pairs ([Formula: see text]). Using a new bait-and-switch simulation approach, we show that up to 50% of the CpGs found in the simulated data may be false-positive results. We present an alternative two-step multiple testing approach using the Simes and Benjamini-Hochberg procedures that does control the FDR among the CpGs, as confirmed by the bait-and-switch simulation. This approach indicates the use of window sizes in cis meQTL mapping studies that are significantly smaller than commonly adopted.

DISCUSSION

Our approach to cis meQTL mapping properly controls the FDR at the CpG level, is computationally fast and can also be applied to cis eQTL studies.

AVAILABILITY AND IMPLEMENTATION

An examplary R script for performing the Simes procedure is available as supplementary material.

CONTACT

e.w.van_zwet@lumc.nl or b.t.heijmans@lumc.nl

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

High-Performance Mixed Models Based Genome-Wide Association Analysis with omicABEL software

To raise the power of genome-wide association studies (GWAS) and avoid false-positive results in structured populations, one can rely on mixed model based tests. When large samples are used, and when multiple traits are to be studied in the 'omics' context, this approach becomes computationally challenging. Here we consider the problem of mixed-model based GWAS for arbitrary number of traits, and demonstrate that for the analysis of single-trait and multiple-trait scenarios different computational algorithms are optimal. We implement these optimal algorithms in a high-performance computing framework that uses state-of-the-art linear algebra kernels, incorporates optimizations, and avoids redundant computations, increasing throughput while reducing memory usage and energy consumption. We show that, compared to existing libraries, our algorithms and software achieve considerable speed-ups. The OmicABEL software described in this manuscript is available under the GNU GPL v. 3 license as part of the GenABEL project for statistical genomics at http: //www.genabel.org/packages/OmicABEL.

Testing for association with multiple traits in generalized estimation equations, with application to neuroimaging data

There is an increasing need to develop and apply powerful statistical tests to detect multiple traits-single locus associations, as arising from neuroimaging genetics and other studies. For example, in the Alzheimer's Disease Neuroimaging Initiative (ADNI), in addition to genome-wide single nucleotide polymorphisms (SNPs), thousands of neuroimaging and neuropsychological phenotypes as intermediate phenotypes for Alzheimer's disease, have been collected. Although some classic methods like MANOVA and newly proposed methods may be applied, they have their own limitations. For example, MANOVA cannot be applied to binary and other discrete traits. In addition, the relationships among these methods are not well understood. Importantly, since these tests are not data adaptive, depending on the unknown association patterns among multiple traits and between multiple traits and a locus, these tests may or may not be powerful. In this paper we propose a class of data-adaptive weights and the corresponding weighted tests in the general framework of generalized estimation equations (GEE). A highly adaptive test is proposed to select the most powerful one from this class of the weighted tests so that it can maintain high power across a wide range of situations. Our proposed tests are applicable to various types of traits with or without covariates. Importantly, we also analytically show relationships among some existing and our proposed tests, indicating that many existing tests are special cases of our proposed tests. Extensive simulation studies were conducted to compare and contrast the power properties of various existing and our new methods. Finally, we applied the methods to an ADNI dataset to illustrate the performance of the methods. We conclude with the recommendation for the use of the GEE-based Score test and our proposed adaptive test for their high and complementary performance.

Genome-wide analyses of borderline personality features

The heritability of borderline personality (BP) features has been established in multiple twin and family studies. Using data from the borderline subscale of the Personality Assessment Inventory Borderline Features Scale (PAI-BOR) collected in two Dutch cohorts (N=7125), the Netherlands Twin Register and The Netherlands Study of Depression and Anxiety, we show that heritability of the PAI-BOR total score using genome-wide single-nucleotide polymorphism (SNPs) is estimated at 23%, and that the genetic variance is substantially higher in affect instability items compared with the other three subscales of the PAI-BOR (42.7% vs non-significant estimates for self-harm, negative relations and identity problems). We present results from a first genome-wide association study of BP features, which shows a promising signal on chromosome 5 corresponding to SERINC5, a protein involved in myelination. Reduced myelination has been suggested as possibly having a role in the development of psychiatric disorders characterized by lack of social interaction. The signal was confirmed in a third independent Dutch cohort drawn from the Erasmus Rucphen Family study (N=1301). Our analyses were complemented by investigating the heterogeneity that was implied by the differences in genetic variance components in the four subscales of the PAI-BOR. These analyses show that the association of SNPs tagging SERINC5 differs substantially across the 24 items of the PAI-BOR. Further, using reverse regression we showed that the effects were present only in subjects with higher scores on the PAI-BOR. Taken together, these results suggest that future genome-wide analyses can benefit substantially by taking into account the phenotypic and genetic heterogeneity of BP features.

Genetics of the human metabolome, what is next?

Increases in throughput and decreases in costs have facilitated large scale metabolomics studies, the simultaneous measurement of large numbers of biochemical components in biological samples. Initial large scale studies focused on biomarker discovery for disease or disease progression and helped to understand biochemical pathways underlying disease. The first population-based studies that combined metabolomics and genome wide association studies (mGWAS) have increased our understanding of the (genetic) regulation of biochemical conversions. Measurements of metabolites as intermediate phenotypes are a potentially very powerful approach to uncover how genetic variation affects disease susceptibility and progression. However, we still face many hurdles in the interpretation of mGWAS data. Due to the composite nature of many metabolites, single enzymes may affect the levels of multiple metabolites and, conversely, levels of single metabolites may be affected by multiple enzymes. Here, we will provide a global review of the current status of mGWAS. We will specifically discuss the application of prior biological knowledge present in databases to the interpretation of mGWAS results and discuss the potential of mathematical models. As the technology continuously improves to detect metabolites and to measure genetic variation, it is clear that comprehensive systems biology based approaches are required to further our insight in the association between genes, metabolites and disease. This article is part of a Special Issue entitled: From Genome to Function.

Genetic and Environmental Stability of Intelligence in Childhood and Adolescence

The present study examined the genetic and environmental contributions to the temporal stability of verbal, non-verbal and general intelligence across a developmental period spanning childhood and adolescence (5–18 years). Longitudinal twin data collected in four different studies on a total of 1,748 twins, comprising 4,641 measurement points in total, were analyzed using genetic adaptations of the simplex model. The heterogeneity in the type of instrument used to assess psychometric intelligence across the different subsamples and ages allowed us to address the auxiliary question of how to optimally utilize the existing longitudinal data in the context of gene-finding studies. The results were consistent across domains (verbal, non-verbal and general intelligence), and indicated that phenotypic stability was driven primarily by the high stability of additive genetic factors, that the stability of common environment was moderate, and that the unique environment contributed primarily to change. The cross-subscale stability was consistently low, indicating a small overlap between different domains of intelligence over time. The high stability of additive genetic factors justifies the use of a linear combination of scores across the different ages in the context of gene-finding studies.

Unraveling genetic origin of aging-related traits: evolving concepts

Discovering the genetic origin of aging-related traits could greatly advance strategies aiming to extend health span. The results of genome-wide association studies (GWAS) addressing this problem are controversial, and new genetic concepts have been fostered to advance the progress in the field. A limitation of GWAS and new genetic concepts is that they do not thoroughly address specifics of aging-related traits. Integration of theoretical concepts in genetics and aging research with empirical evidence from different disciplines highlights the conceptual problems in studies of genetic origin of aging-related traits. To address these problems, novel approaches of systemic nature are required. These approaches should adopt the non-deterministic nature of linkage of genes with aging-related traits and, consequently, reinforce research strategies for improving our understanding of mechanisms shaping genetic effects on these traits. Investigation of mechanisms will help determine conditions that activate specific genetic variants or profiles and explore to what extent these conditions that shape genetic effects are conserved across human lives and generations.

Maximizing the power of principal-component analysis of correlated phenotypes in genome-wide association studies

Many human traits are highly correlated. This correlation can be leveraged to improve the power of genetic association tests to identify markers associated with one or more of the traits. Principal component analysis (PCA) is a useful tool that has been widely used for the multivariate analysis of correlated variables. PCA is usually applied as a dimension reduction method: the few top principal components (PCs) explaining most of total trait variance are tested for association with a predictor of interest, and the remaining components are not analyzed. In this study we review the theoretical basis of PCA and describe the behavior of PCA when testing for association between a SNP and correlated traits. We then use simulation to compare the power of various PCA-based strategies when analyzing up to 100 correlated traits. We show that contrary to widespread practice, testing only the top PCs often has low power, whereas combining signal across all PCs can have greater power. This power gain is primarily due to increased power to detect genetic variants with opposite effects on positively correlated traits and variants that are exclusively associated with a single trait. Relative to other methods, the combined-PC approach has close to optimal power in all scenarios considered while offering more flexibility and more robustness to potential confounders. Finally, we apply the proposed PCA strategy to the genome-wide association study of five correlated coagulation traits where we identify two candidate SNPs that were not found by the standard approach.

A comparison of multivariate genome-wide association methods

Joint association analysis of multiple traits in a genome-wide association study (GWAS), i.e. a multivariate GWAS, offers several advantages over analyzing each trait in a separate GWAS. In this study we directly compared a number of multivariate GWAS methods using simulated data. We focused on six methods that are implemented in the software packages PLINK, SNPTEST, MultiPhen, BIMBAM, PCHAT and TATES, and also compared them to standard univariate GWAS, analysis of the first principal component of the traits, and meta-analysis of univariate results. We simulated data (N = 1000) for three quantitative traits and one bi-allelic quantitative trait locus (QTL), and varied the number of traits associated with the QTL (explained variance 0.1%), minor allele frequency of the QTL, residual correlation between the traits, and the sign of the correlation induced by the QTL relative to the residual correlation. We compared the power of the methods using empirically fixed significance thresholds (α = 0.05). Our results showed that the multivariate methods implemented in PLINK, SNPTEST, MultiPhen and BIMBAM performed best for the majority of the tested scenarios, with a notable increase in power for scenarios with an opposite sign of genetic and residual correlation. All multivariate analyses resulted in a higher power than univariate analyses, even when only one of the traits was associated with the QTL. Hence, use of multivariate GWAS methods can be recommended, even when genetic correlations between traits are weak.

Multivariate genetic analyses in heterogeneous populations

Martin and Eaves (Heredity 38(1):79-95, 1977) proposed a multivariate model for twin and family data in order to investigate potential differences in the genetic and environmental architecture of multivariate phenotypes. The general form of the model is the independent pathway model, which differentiates between genetic and environmental influences at the item level, and therefore permits the decomposition to differ across items. A restricted version is the common pathway model, where the decomposition takes place at the factor level. The paper has spurred numerous studies, and evidence for differences in genetic and environmental architecture has been established for personality and several other psychiatric phenotypes by showing a better fit of the independent pathway model compared to the common pathway model. We show that genome-wide association studies (GWAS) that use an aggregate score computed from multiple questionnaire items as a univariate phenotype implicitly assume a similar structure as the common pathway model. It has been shown that in case of a differential genetic and environmental architecture, multivariate GWAS methods can outperform the univariate GWAS approach. However, current multivariate methods rely on the assumptions of phenotypic and genetic homogeneity, that is, item responses are assumed to have the same means and covariances, and genetic effects are assumed to be the same for all subjects. We describe a distance-based regression technique that is designed to account for subgroups in the population, and that therefore can account for differential genetic effects. A first evaluation with simulated data shows a substantial increase of power compared to univariate GWAS.

A rapid gene-based genome-wide association test with multivariate traits

OBJECTIVES

A gene-based genome-wide association study (GWAS) provides a powerful alternative to the traditional single single nucleotide polymorphism (SNP) association analysis due to its substantial reduction in the multiple testing burden and possible gain in power due to modeling multiple SNPs within a gene. A gene-based association analysis on multivariate traits is often of interest, but it imposes substantial analytical as well as computational challenges to implement it at a genome-wide level.

METHODS

We propose a rapid implementation of the multivariate multiple linear regression (RMMLR) approach in unrelated individuals as well as in families. Our approach allows for covariates. Moreover, the asymptotic distribution of the test statistic is not heavily influenced by the linkage disequilibrium (LD) among the SNPs and hence can be used efficiently to perform a gene-based GWAS. We have developed a corresponding R package to implement such multivariate gene-based GWAS with this RMMLR approach.

RESULTS

Through extensive simulation, we compared several approaches for both single and multivariate traits. Our RMMLR approach maintained a correct type I error level even for sets of SNPs in strong LD. It also demonstrated a substantial gain in power to detect a gene when it is associated with a subset of the traits. We also studied performances of the approaches on the Minnesota Center for Twin Family Research dataset.

CONCLUSIONS

In our overall comparison, our RMMLR approach provides an efficient and powerful tool to perform a gene-based GWAS with single or multivariate traits and maintains the type I error appropriately.

A general framework for association tests with multivariate traits in large-scale genomics studies

Genetic association studies often collect data on multiple traits that are correlated. Discovery of genetic variants influencing multiple traits can lead to better understanding of the etiology of complex human diseases. Conventional univariate association tests may miss variants that have weak or moderate effects on individual traits. We propose several multivariate test statistics to complement univariate tests. Our framework covers both studies of unrelated individuals and family studies and allows any type/mixture of traits. We relate the marginal distributions of multivariate traits to genetic variants and covariates through generalized linear models without modeling the dependence among the traits or family members. We construct score-type statistics, which are computationally fast and numerically stable even in the presence of covariates and which can be combined efficiently across studies with different designs and arbitrary patterns of missing data. We compare the power of the test statistics both theoretically and empirically. We provide a strategy to determine genome-wide significance that properly accounts for the linkage disequilibrium (LD) of genetic variants. The application of the new methods to the meta-analysis of five major cardiovascular cohort studies identifies a new locus (HSCB) that is pleiotropic for the four traits analyzed.

Pleiotropy in complex traits: challenges and strategies

Genome-wide association studies have identified many variants that each affects multiple traits, particularly across autoimmune diseases, cancers and neuropsychiatric disorders, suggesting that pleiotropic effects on human complex traits may be widespread. However, systematic detection of such effects is challenging and requires new methodologies and frameworks for interpreting cross-phenotype results. In this Review, we discuss the evidence for pleiotropy in contemporary genetic mapping studies, new and established analytical approaches to identifying pleiotropic effects, sources of spurious cross-phenotype effects and study design considerations. We also outline the molecular and clinical implications of such findings and discuss future directions of research.