Genetic association test for multiple traits at gene level

A copula-based set-variant association test for bivariate continuous, binary or mixed phenotypes

In genome wide association studies (GWAS), researchers are often dealing with dichotomous and non-normally distributed traits, or a mixture of discrete-continuous traits. However, most of the current region-based methods rely on multivariate linear mixed models (mvLMMs) and assume a multivariate normal distribution for the phenotypes of interest. Hence, these methods are not applicable to disease or non-normally distributed traits. Therefore, there is a need to develop unified and flexible methods to study association between a set of (possibly rare) genetic variants and non-normal multivariate phenotypes. Copulas are multivariate distribution functions with uniform margins on the [0, 1] interval and they provide suitable models to deal with non-normality of errors in multivariate association studies. We propose a novel unified and flexible copula-based multivariate association test (CBMAT) for discovering association between a genetic region and a bivariate continuous, binary or mixed phenotype. We also derive a data-driven analytic p-value procedure of the proposed region-based score-type test. Through simulation studies, we demonstrate that CBMAT has well controlled type I error rates and higher power to detect associations compared with other existing methods, for discrete and non-normally distributed traits. At last, we apply CBMAT to detect the association between two genes located on chromosome 11 and several lipid levels measured on 1477 subjects from the ASLPAC study.

Genetic and Environmental Predictors of Adolescent PTSD Symptom Trajectories Following a Natural Disaster

: Genes, environmental factors, and their interplay affect posttrauma symptoms. Although environmental predictors of the longitudinal course of posttraumatic stress disorder (PTSD) symptoms are documented, there remains a need to incorporate genetic risk into these models, especially in youth who are underrepresented in genetic studies. In an epidemiologic sample tornado-exposed adolescents (n = 707, 51% female, Mage = 14.54 years), trajectories of PTSD symptoms were examined at baseline and at 4-months and 12-months following baseline. This study aimed to determine if rare genetic variation in genes previously found in the sample to be related to PTSD diagnosis at baseline (MPHOSPH9, LGALS13, SLC2A2), environmental factors (disaster severity, social support), or their interplay were associated with symptom trajectories. A series of mixed effects models were conducted. Symptoms decreased over the three time points. Elevated tornado severity was associated with elevated baseline symptoms. Elevated recreational support was associated with lower baseline symptoms and attenuated improvement over time. Greater LGLAS13 variants attenuated symptom improvement over time. An interaction between MPHOSPH9 variants and tornado severity was associated with elevated baseline symptoms, but not change over time. Findings suggest the importance of rare genetic variation and environmental factors on the longitudinal course of PTSD symptoms following natural disaster trauma exposure.

A preliminary investigation of rare variants associated with genetic risk for PTSD in a natural disaster-exposed adolescent sample

Background: Posttraumatic stress disorder (PTSD) involves a complex interaction of biological, psychological, and social factors. Numerous studies have demonstrated genetic variation associated with the development of PTSD, primarily in adults. However, the contribution of low frequency and rare genetic variants to PTSD is unknown to date. Moreover, there is limited work on genetic risk for PTSD in child and adolescent populations. Objective: This preliminary study aimed to identify the low frequency and rare genetic variation that contributes to PTSD using an exome array. Method: This post-disaster, adolescent sample (n = 707, 51% females, M _age = 14.54) was assessed for PTSD diagnosis and symptom count following tornado exposure. Results: Gene-based models, covarying for ancestry principal components, age, sex, tornado severity, and previous trauma identified variants in four genes associated with diagnosis and 276 genes associated with symptom count (at p _adj < .001). Functional class analyses suggested an association with variants in the nonsense class (nonsynonymous variant that results in truncation of, and usually non-functional, protein) with both outcomes. An exploratory gene network pathway analysis showed a great number of significant genes involved in brain and immune function, illustrating the usefulness of downstream examination of gene-based findings that may point to relevant biological processes. Conclusions: While further investigation in larger samples is warranted, findings align with extant PTSD literature that has identified variants associated with biological conditions such as immune function.

Powerful and robust cross-phenotype association test for case-parent trios

There has been increasing interest in identifying genes within the human genome that influence multiple diverse phenotypes. In the presence of pleiotropy, joint testing of these phenotypes is not only biologically meaningful but also statistically more powerful than univariate analysis of each separate phenotype accounting for multiple testing. Although many cross-phenotype association tests exist, the majority of such methods assume samples composed of unrelated subjects and therefore are not applicable to family-based designs, including the valuable case-parent trio design. In this paper, we describe a robust gene-based association test of multiple phenotypes collected in a case-parent trio study. Our method is based on the kernel distance covariance (KDC) method, where we first construct a similarity matrix for multiple phenotypes and a similarity matrix for genetic variants in a gene; we then test the dependency between the two similarity matrices. The method is applicable to either common variants or rare variants in a gene, and resulting tests from the method are by design robust to confounding due to population stratification. We evaluated our method through simulation studies and observed that the method is substantially more powerful than standard univariate testing of each separate phenotype. We also applied our method to phenotypic and genotypic data collected in case-parent trios as part of the Genetics of Kidneys in Diabetes (GoKinD) study and identified a genome-wide significant gene demonstrating cross-phenotype effects that was not identified using standard univariate approaches.

A functional U-statistic method for association analysis of sequencing data

Although sequencing studies hold great promise for uncovering novel variants predisposing to human diseases, the high dimensionality of the sequencing data brings tremendous challenges to data analysis. Moreover, for many complex diseases (e.g., psychiatric disorders) multiple related phenotypes are collected. These phenotypes can be different measurements of an underlying disease, or measurements characterizing multiple related diseases for studying common genetic mechanism. Although jointly analyzing these phenotypes could potentially increase the power of identifying disease-associated genes, the different types of phenotypes pose challenges for association analysis. To address these challenges, we propose a nonparametric method, functional U-statistic method (FU), for multivariate analysis of sequencing data. It first constructs smooth functions from individuals' sequencing data, and then tests the association of these functions with multiple phenotypes by using a U-statistic. The method provides a general framework for analyzing various types of phenotypes (e.g., binary and continuous phenotypes) with unknown distributions. Fitting the genetic variants within a gene using a smoothing function also allows us to capture complexities of gene structure (e.g., linkage disequilibrium, LD), which could potentially increase the power of association analysis. Through simulations, we compared our method to the multivariate outcome score test (MOST), and found that our test attained better performance than MOST. In a real data application, we apply our method to the sequencing data from Minnesota Twin Study (MTS) and found potential associations of several nicotine receptor subunit (CHRN) genes, including CHRNB3, associated with nicotine dependence and/or alcohol dependence.

Resistance to RHD virus in wild Australian rabbits: Comparison of susceptible and resistant individuals using a genomewide approach

Deciphering the genes involved in disease resistance is essential if we are to understand host-pathogen coevolutionary processes. The rabbit haemorrhagic disease virus (RHDV) was imported into Australia in 1995 as a biocontrol agent to manage one of the most successful and devastating invasive species, the European rabbit (Oryctolagus cuniculus). During the first outbreaks of the disease, RHDV caused mortality rates of up to 97%. Recently, however, increased genetic resistance to RHDV has been reported. Here, we have aimed to identify genomic differences between rabbits that survived a natural infection with RHDV and those that died in the field using a genomewide next-generation sequencing (NGS) approach. We detected 72 SNPs corresponding to 133 genes associated with survival of a RHD infection. Most of the identified genes have known functions in virus infections and replication, immune responses or apoptosis, or have previously been found to be regulated during RHD. Some of the genes identified in experimental studies, however, did not seem to play a role under natural selection regimes, highlighting the importance of field studies to complement the genomic background of wildlife diseases. Our study provides a set of candidate markers as a tool for the future scanning of wild rabbits for their resistance to RHDV. This is important both for wild rabbit populations in southern Europe where RHD is regarded as a serious problem decimating the prey of endangered predator species and for assessing the success of currently planned RHDV variant biocontrol releases in Australia.

Genome-wide association test of multiple continuous traits using imputed SNPs

More and more large cohort studies have conducted or are conducting genome-wide association studies (GWAS) to reveal the genetic components of many complex human diseases. These large cohort studies often collected a broad array of correlated phenotypes that reflect common physiological processes. By jointly analyzing these correlated traits, we can gain more power by aggregating multiple weak effects and shed light on the mechanisms underlying complex human diseases. The majority of existing multi-trait association test methods are based on jointly modeling the multivariate traits conditional on the genotype as covariate, and can readily accommodate the imputed SNPs by using their imputed dosage as a covariate. An alternative class of multi-trait association tests is based on the inverted regression, which models the distribution of genotypes conditional on the covariate and multivariate traits, and has been shown to have competitive performance. To our knowledge, all existing inverted regression approaches have implicitly used the "best-guess" genotypes, which is not efficient and known to lead to dramatic power loss, and there have not been any proposed methods of incorporating imputation uncertainty into inverted regressions. In this work, we propose a general and efficient framework that can account for the imputation uncertainty to further improve the association test power of inverted regression models for imputed SNPs. We demonstrate through extensive numerical studies that the proposed method has competitive performance. We further illustrate its usefulness by application to association test of diabetes-related glycemic traits in the Atherosclerosis Risk in Communities (ARIC) Study.

Meta-analysis of quantitative pleiotropic traits for next-generation sequencing with multivariate functional linear models

To analyze next-generation sequencing data, multivariate functional linear models are developed for a meta-analysis of multiple studies to connect genetic variant data to multiple quantitative traits adjusting for covariates. The goal is to take the advantage of both meta-analysis and pleiotropic analysis in order to improve power and to carry out a unified association analysis of multiple studies and multiple traits of complex disorders. Three types of approximate F -distributions based on Pillai-Bartlett trace, Hotelling-Lawley trace, and Wilks's Lambda are introduced to test for association between multiple quantitative traits and multiple genetic variants. Simulation analysis is performed to evaluate false-positive rates and power of the proposed tests. The proposed methods are applied to analyze lipid traits in eight European cohorts. It is shown that it is more advantageous to perform multivariate analysis than univariate analysis in general, and it is more advantageous to perform meta-analysis of multiple studies instead of analyzing the individual studies separately. The proposed models require individual observations. The value of the current paper can be seen at least for two reasons: (a) the proposed methods can be applied to studies that have individual genotype data; (b) the proposed methods can be used as a criterion for future work that uses summary statistics to build test statistics to meta-analyze the data.

Joint analysis of multiple blood pressure phenotypes in GAW19 data by using a multivariate rare-variant association test

INTRODUCTION

Large-scale sequencing studies often measure many related phenotypes in addition to the genetic variants. Joint analysis of multiple phenotypes in genetic association studies may increase power to detect disease-associated loci.

METHODS

We apply a recently developed multivariate rare-variant association test to the Genetic Analysis Workshop 19 data in order to test associations between genetic variants and multiple blood pressure phenotypes simultaneously. We also compare this multivariate test with a widely used univariate test that analyzes phenotypes separately.

RESULTS

The multivariate test identified 2 genetic variants that have been previously reported as associated with hypertension or coronary artery disease. In addition, our region-based analyses also show that the multivariate test tends to give smaller p values than the univariate test.

CONCLUSIONS

Hence, the multivariate test has potential to improve test power, especially when multiple phenotypes are correlated.

Quality Control for the Illumina HumanExome BeadChip

The Illumina HumanExome BeadChip and other exome-based genotyping arrays offer inexpensive genotyping of some 240,000 mostly nonsynonymous coding variants across the human genome. The HumanExome chip, with its highly non-uniform distribution of markers and emphasis on rare coding variants, presents some unique challenges for quality control (QC) and data cleaning. Here, we describe QC procedures for HumanExome data, with examples of challenges specific to exome arrays from our experience cleaning a data set of ∼7,500 samples from the NEIGHBORHOOD Consortium. We focus on standard procedures for QC of genome-wide array data including genotype calling, sex verification, sample identity verification, relationship checking, and population structure that are complicated by the HumanExome panel's enrichment in rare, exonic variation. © 2016 by John Wiley & Sons, Inc.

Functional Characterization of Schizophrenia-Associated Variation in CACNA1C

Calcium channel subunits, including CACNA1C, have been associated with multiple psychiatric disorders. Specifically, genome wide association studies (GWAS) have repeatedly identified the single nucleotide polymorphism (SNP) rs1006737 in intron 3 of CACNA1C to be strongly associated with schizophrenia and bipolar disorder. Here, we show that rs1006737 marks a quantitative trait locus for CACNA1C transcript levels. We test 16 SNPs in high linkage disequilibrium with rs1007637 and find one, rs4765905, consistently showing allele-dependent regulatory function in reporter assays. We find allele-specific protein binding for 13 SNPs including rs4765905. Using protein microarrays, we identify several proteins binding ≥3 SNPs, but not control sequences, suggesting possible functional interactions and combinatorial haplotype effects. Finally, using circular chromatin conformation capture, we show interaction of the disease-associated region including the 16 SNPs with the CACNA1C promoter and other potential regulatory regions. Our results elucidate the pathogenic relevance of one of the best-supported risk loci for schizophrenia and bipolar disorder.

Powerful and Adaptive Testing for Multi-trait and Multi-SNP Associations with GWAS and Sequencing Data

Testing for genetic association with multiple traits has become increasingly important, not only because of its potential to boost statistical power, but also for its direct relevance to applications. For example, there is accumulating evidence showing that some complex neurodegenerative and psychiatric diseases like Alzheimer's disease are due to disrupted brain networks, for which it would be natural to identify genetic variants associated with a disrupted brain network, represented as a set of multiple traits, one for each of multiple brain regions of interest. In spite of its promise, testing for multivariate trait associations is challenging: if not appropriately used, its power can be much lower than testing on each univariate trait separately (with a proper control for multiple testing). Furthermore, differing from most existing methods for single-SNP-multiple-trait associations, we consider SNP set-based association testing to decipher complicated joint effects of multiple SNPs on multiple traits. Because the power of a test critically depends on several unknown factors such as the proportions of associated SNPs and of traits, we propose a highly adaptive test at both the SNP and trait levels, giving higher weights to those likely associated SNPs and traits, to yield high power across a wide spectrum of situations. We illuminate relationships among the proposed and some existing tests, showing that the proposed test covers several existing tests as special cases. We compare the performance of the new test with that of several existing tests, using both simulated and real data. The methods were applied to structural magnetic resonance imaging data drawn from the Alzheimer's Disease Neuroimaging Initiative to identify genes associated with gray matter atrophy in the human brain default mode network (DMN). For genome-wide association studies (GWAS), genes AMOTL1 on chromosome 11 and APOE on chromosome 19 were discovered by the new test to be significantly associated with the DMN. Notably, gene AMOTL1 was not detected by single SNP-based analyses. To our knowledge, AMOTL1 has not been highlighted in other Alzheimer's disease studies before, although it was indicated to be related to cognitive impairment. The proposed method is also applicable to rare variants in sequencing data and can be extended to pathway analysis.

A method for analyzing multiple continuous phenotypes in rare variant association studies allowing for flexible correlations in variant effects

For region-based sequencing data, power to detect genetic associations can be improved through analysis of multiple related phenotypes. With this motivation, we propose a novel test to detect association simultaneously between a set of rare variants, such as those obtained by sequencing in a small genomic region, and multiple continuous phenotypes. We allow arbitrary correlations among the phenotypes and build on a linear mixed model by assuming the effects of the variants follow a multivariate normal distribution with a zero mean and a specific covariance matrix structure. In order to account for the unknown correlation parameter in the covariance matrix of the variant effects, a data-adaptive variance component test based on score-type statistics is derived. As our approach can calculate the P-value analytically, the proposed test procedure is computationally efficient. Broad simulations and an application to the UK10K project show that our proposed multivariate test is generally more powerful than univariate tests, especially when there are pleiotropic effects or highly correlated phenotypes.

Association Tests of Multiple Phenotypes: ATeMP

Joint analysis of multiple phenotypes has gained growing attention in genome-wide association studies (GWASs), especially for the analysis of multiple intermediate phenotypes which measure the same underlying complex human disorder. One of the multivariate methods, MultiPhen (O’ Reilly et al. 2012), employs the proportional odds model to regress a genotype on multiple phenotypes, hence ignoring the phenotypic distributions. Despite the flexibilities of MultiPhen, the properties and performance of MultiPhen are not well understood, especially when the phenotypic distributions are non-normal. In fact, it is well known in the statistical literature that the estimation is attenuated when the explanatory variables contain measurement errors. In this study, we first established an equivalence relationship between MultiPhen and the generalized Kendall tau association test, shedding light on why MultiPhen can perform well for joint association analysis of multiple phenotypes. Through the equivalence, we show that MultiPhen may lose power when the phenotypes are non-normal. To maintain the power, we propose two solutions (ATeMP-rn and ATeMP-or) to improve MultiPhen, and demonstrate their effectiveness through extensive simulation studies and a real case study from the Guangzhou Twin Eye Study.

Pleiotropy analysis of quantitative traits at gene level by multivariate functional linear models

In genetics, pleiotropy describes the genetic effect of a single gene on multiple phenotypic traits. A common approach is to analyze the phenotypic traits separately using univariate analyses and combine the test results through multiple comparisons. This approach may lead to low power. Multivariate functional linear models are developed to connect genetic variant data to multiple quantitative traits adjusting for covariates for a unified analysis. Three types of approximate F-distribution tests based on Pillai-Bartlett trace, Hotelling-Lawley trace, and Wilks's Lambda are introduced to test for association between multiple quantitative traits and multiple genetic variants in one genetic region. The approximate F-distribution tests provide much more significant results than those of F-tests of univariate analysis and optimal sequence kernel association test (SKAT-O). Extensive simulations were performed to evaluate the false positive rates and power performance of the proposed models and tests. We show that the approximate F-distribution tests control the type I error rates very well. Overall, simultaneous analysis of multiple traits can increase power performance compared to an individual test of each trait. The proposed methods were applied to analyze (1) four lipid traits in eight European cohorts, and (2) three biochemical traits in the Trinity Students Study. The approximate F-distribution tests provide much more significant results than those of F-tests of univariate analysis and SKAT-O for the three biochemical traits. The approximate F-distribution tests of the proposed functional linear models are more sensitive than those of the traditional multivariate linear models that in turn are more sensitive than SKAT-O in the univariate case. The analysis of the four lipid traits and the three biochemical traits detects more association than SKAT-O in the univariate case.

A genome-wide association study of early spontaneous preterm delivery

Preterm birth is the leading cause of infant morbidity and mortality. Despite extensive research, the genetic contributions to spontaneous preterm birth (SPTB) are not well understood. Term controls were matched with cases by race/ethnicity, maternal age, and parity prior to recruitment. Genotyping was performed using Affymetrix SNP Array 6.0 assays. Statistical analyses utilized PLINK to compare allele occurrence rates between case and control groups, and incorporated quality control and multiple-testing adjustments. We analyzed DNA samples from mother-infant pairs from early SPTB cases (20(0/7)-33(6/7) weeks, 959 women and 979 neonates) and term delivery controls (39(0/7)-41(6/7) weeks, 960 women and 985 neonates). For validation purposes, we included an independent validation cohort consisting of early SPTB cases (293 mothers and 243 infants) and term controls (200 mothers and 149 infants). Clustering analysis revealed no population stratification. Multiple maternal SNPs were identified with association P-values between 10×10(-5) and 10×10(-6). The most significant maternal SNP was rs17053026 on chromosome 3 with an odds ratio (OR) 0.44 with a P-value of 1.0×10(-6). Two neonatal SNPs reached the genome-wide significance threshold, including rs17527054 on chromosome 6p22 with a P-value of 2.7×10(-12) and rs3777722 on chromosome 6q27 with a P-value of 1.4×10(-10). However, we could not replicate these findings after adjusting for multiple comparisons in a validation cohort. This is the first report of a genome-wide case-control study to identify single nucleotide polymorphisms (SNPs) that correlate with SPTB.

A genetic deconstruction of neurocognitive traits in schizophrenia and bipolar disorder

Background

Impairments in cognitive functions are common in patients suffering from psychiatric disorders, such as schizophrenia and bipolar disorder. Cognitive traits have been proposed as useful for understanding the biological and genetic mechanisms implicated in cognitive function in healthy individuals and in the dysfunction observed in psychiatric disorders.

Methods

Sets of genes associated with a range of cognitive functions often impaired in schizophrenia and bipolar disorder were generated from a genome-wide association study (GWAS) on a sample comprising 670 healthy Norwegian adults who were phenotyped for a broad battery of cognitive tests. These gene sets were then tested for enrichment of association in GWASs of schizophrenia and bipolar disorder. The GWAS data was derived from three independent single-centre schizophrenia samples, three independent single-centre bipolar disorder samples, and the multi-centre schizophrenia and bipolar disorder samples from the Psychiatric Genomics Consortium.

Results

The strongest enrichments were observed for visuospatial attention and verbal abilities sets in bipolar disorder. Delayed verbal memory was also enriched in one sample of bipolar disorder. For schizophrenia, the strongest evidence of enrichment was observed for the sets of genes associated with performance in a colour-word interference test and for sets associated with memory learning slope.

Conclusions

Our results are consistent with the increasing evidence that cognitive functions share genetic factors with schizophrenia and bipolar disorder. Our data provides evidence that genetic studies using polygenic and pleiotropic models can be used to link specific cognitive functions with psychiatric disorders.

NCK2 is significantly associated with opiates addiction in African-origin men

Substance dependence is a complex environmental and genetic disorder with significant social and medical concerns. Understanding the etiology of substance dependence is imperative to the development of effective treatment and prevention strategies. To this end, substantial effort has been made to identify genes underlying substance dependence, and in recent years, genome-wide association studies (GWASs) have led to discoveries of numerous genetic variants for complex diseases including substance dependence. Most of the GWAS discoveries were only based on single nucleotide polymorphisms (SNPs) and a single dichotomized outcome. By employing both SNP- and gene-based methods of analysis, we identified a strong (odds ratio = 13.87) and significant (P value = 1.33E - 11) association of an SNP in the NCK2 gene on chromosome 2 with opiates addiction in African-origin men. Codependence analysis also identified a genome-wide significant association between NCK2 and comorbidity of substance dependence (P value = 3.65E - 08) in African-origin men. Furthermore, we observed that the association between the NCK2 gene (P value = 3.12E - 10) and opiates addiction reached the gene-based genome-wide significant level. In summary, our findings provided the first evidence for the involvement of NCK2 in the susceptibility to opiates addiction and further revealed the racial and gender specificities of its impact.

Large scale association analysis for drug addiction: results from SNP to gene

Many genetic association studies used single nucleotide polymorphisms (SNPs) data to identify genetic variants for complex diseases. Although SNP-based associations are most common in genome-wide association studies (GWAS), gene-based association analysis has received increasing attention in understanding genetic etiologies for complex diseases. While both methods have been used to analyze the same data, few genome-wide association studies compare the results or observe the connection between them. We performed a comprehensive analysis of the data from the Study of Addiction: Genetics and Environment (SAGE) and compared the results from the SNP-based and gene-based analyses. Our results suggest that the gene-based method complements the individual SNP-based analysis, and conceptually they are closely related. In terms of gene findings, our results validate many genes that were either reported from the analysis of the same dataset or based on animal studies for substance dependence.