A versatile gene-based test for genome-wide association studies

DOT: Gene-set analysis by combining decorrelated association statistics

Historically, the majority of statistical association methods have been designed assuming availability of SNP-level information. However, modern genetic and sequencing data present new challenges to access and sharing of genotype-phenotype datasets, including cost of management, difficulties in consolidation of records across research groups, etc. These issues make methods based on SNP-level summary statistics particularly appealing. The most common form of combining statistics is a sum of SNP-level squared scores, possibly weighted, as in burden tests for rare variants. The overall significance of the resulting statistic is evaluated using its distribution under the null hypothesis. Here, we demonstrate that this basic approach can be substantially improved by decorrelating scores prior to their addition, resulting in remarkable power gains in situations that are most commonly encountered in practice; namely, under heterogeneity of effect sizes and diversity between pairwise LD. In these situations, the power of the traditional test, based on the added squared scores, quickly reaches a ceiling, as the number of variants increases. Thus, the traditional approach does not benefit from information potentially contained in any additional SNPs, while our decorrelation by orthogonal transformation (DOT) method yields steady gain in power. We present theoretical and computational analyses of both approaches, and reveal causes behind sometimes dramatic difference in their respective powers. We showcase DOT by analyzing breast cancer and cleft lip data, in which our method strengthened levels of previously reported associations and implied the possibility of multiple new alleles that jointly confer disease risk.

Genome-wide association study of cognitive performance in U.S. veterans with schizophrenia or bipolar disorder

Cognitive impairment is a frequent and serious problem in patients with various forms of severe mental illnesses (SMI), including schizophrenia (SZ) and bipolar disorder (BP). Recent research suggests genetic links to several cognitive phenotypes in both SMI and in the general population. Our goal in this study was to identify potential genomic signatures of cognitive functioning in veterans with severe mental illness and compare them to previous findings for cognition across different populations. Veterans Affairs (VA) Cooperative Studies Program (CSP) Study #572 evaluated cognitive and functional capacity measures among SZ and BP patients. In conjunction with the VA Million Veteran Program, 3,959 European American (1,095 SZ, 2,864 BP) and 2,601 African American (1,095 SZ, 2,864 BP) patients were genotyped using a custom Affymetrix Axiom Biobank array. We performed a genome-wide association study of global cognitive functioning, constructed polygenic scores for SZ and cognition in the general population, and examined genetic correlations with 2,626 UK Biobank traits. Although no single locus attained genome-wide significance, observed allelic effects were strongly consistent with previous studies. We observed robust associations between global cognitive functioning and polygenic scores for cognitive performance, intelligence, and SZ risk. We also identified significant genetic correlations with several cognition-related traits in UK Biobank. In a diverse cohort of U.S. veterans with SZ or BP, we demonstrate broad overlap of common genetic effects on cognition in the general population, and find that greater polygenic loading for SZ risk is associated with poorer cognitive performance.

Genome-wide association meta-analysis of corneal curvature identifies novel loci and shared genetic influences across axial length and refractive error

Corneal curvature, a highly heritable trait, is a key clinical endophenotype for myopia - a major cause of visual impairment and blindness in the world. Here we present a trans-ethnic meta-analysis of corneal curvature GWAS in 44,042 individuals of Caucasian and Asian with replication in 88,218 UK Biobank data. We identified 47 loci (of which 26 are novel), with population-specific signals as well as shared signals across ethnicities. Some identified variants showed precise scaling in corneal curvature and eye elongation (i.e. axial length) to maintain eyes in emmetropia (i.e. HDAC11/FBLN2 rs2630445, RBP3 rs11204213); others exhibited association with myopia with little pleiotropic effects on eye elongation. Implicated genes are involved in extracellular matrix organization, developmental process for body and eye, connective tissue cartilage and glycosylation protein activities. Our study provides insights into population-specific novel genes for corneal curvature, and their pleiotropic effect in regulating eye size or conferring susceptibility to myopia.

Germline Features Associated with Immune Infiltration in Solid Tumors

The immune composition of the tumor microenvironment influences response and resistance to immunotherapies. While numerous studies have identified somatic correlates of immune infiltration, germline features that associate with immune infiltrates in cancers remain incompletely characterized. We analyze seven million autosomal germline variants in the TCGA cohort and test for association with established immune-related phenotypes that describe the tumor immune microenvironment. We identify one SNP associated with the amount of infiltrating follicular helper T cells; 23 candidate genes, some of which are involved in cytokine-mediated signaling and others containing cancer-risk SNPs; and networks with genes that are part of the DNA repair and transcription elongation pathways. In addition, we find a positive association between polygenic risk for rheumatoid arthritis and amount of infiltrating CD8+ T cells. Overall, we identify multiple germline genetic features associated with tumor-immune phenotypes and develop a framework for probing inherited features that contribute to differences in immune infiltration.

Phenotypic and Genomic Local Adaptation across Latitude and Altitude in Populus trichocarpa

Local adaptation to climate allows plants to cope with temporally and spatially heterogeneous environments, and parallel phenotypic clines provide a natural experiment to uncover the genomic architecture of adaptation. Though extensive effort has been made to investigate the genomic basis of local adaptation to climate across the latitudinal range of tree species, less is known for altitudinal clines. We used exome capture to genotype 451 Populus trichocarpa genotypes across altitudinal and latitudinal gradients spanning the natural species range, and phenotyped these trees for a variety of adaptive traits in two common gardens. We observed clinal variation in phenotypic traits across the two transects, which indicates climate-driven selection, and coupled gene-based genotype–phenotype and genotype–environment association scans to identify imprints of climatic adaptation on the genome. Although many of the phenotype- and climate-associated genes were unique to one transect, we found evidence of parallelism between latitude and altitude, as well as significant convergence when we compared our outlier genes with those putatively involved in climatic adaptation in two gymnosperm species. These results suggest that not only genomic constraint during adaptation to similar environmental gradients in poplar but also different environmental contexts, spatial scale, and perhaps redundant function among potentially adaptive genes and polymorphisms lead to divergent adaptive architectures.

Shared Molecular Genetic Mechanisms Underlie Endometriosis and Migraine Comorbidity

Observational epidemiological studies indicate that endometriosis and migraine co-occur within individuals more than expected by chance. However, the aetiology and biological mechanisms underlying their comorbidity remain unknown. Here we examined the relationship between endometriosis and migraine using genome-wide association study (GWAS) data. Single nucleotide polymorphism (SNP) effect concordance analysis found a significant concordance of SNP risk effects across endometriosis and migraine GWAS. Linkage disequilibrium score regression analysis found a positive and highly significant genetic correlation (rG = 0.38, P = 2.30 × 10-25) between endometriosis and migraine. A meta-analysis of endometriosis and migraine GWAS data did not reveal novel genome-wide significant SNPs, and Mendelian randomisation analysis found no evidence for a causal relationship between the two traits. However, gene-based analyses identified two novel loci for migraine. Also, we found significant enrichment of genes nominally associated (Pgene < 0.05) with both traits (Pbinomial-test = 9.83 × 10-6). Combining gene-based p-values across endometriosis and migraine, three genes, two (TRIM32 and SLC35G6) of which are at novel loci, were genome-wide significant. Genes having Pgene < 0.1 for both endometriosis and migraine (Pbinomial-test = 1.85 ×10-°3) were significantly enriched for biological pathways, including interleukin-1 receptor binding, focal adhesion-PI3K-Akt-mTOR-signaling, MAPK and TNF-α signalling. Our findings further confirm the comorbidity of endometriosis and migraine and indicate a non-causal relationship between the two traits, with shared genetically-controlled biological mechanisms underlying the co-occurrence of the two disorders.

Potential functional variants in SMC2 and TP53 in the AURORA pathway genes and risk of pancreatic cancer

The AURORA pathway participates in mitosis and cell division, and alterations in mitosis and cell division can lead to carcinogenesis. Therefore, genetic variants in the AURORA pathway genes may be associated with susceptibility to pancreatic cancer. To test this hypothesis, we used three large publically available pancreatic cancer genome-wide association study (GWAS) datasets (PanScan I, II/III and PanC4) to assess the associations of 7168 single nucleotide polymorphisms (SNPs) in a set of 62 genes of this pathway with pancreatic cancer risk in 8477 cases and 6946 controls of European ancestry. We identify 15 significant pancreatic cancer risk-associated SNPs in three genes (SMC2, ARHGEF7 and TP53) after correction for multiple comparisons by a false discovery rate < 0.20. Through further linkage disequilibrium analysis, SNP functional prediction and stepwise logistic regression analysis, we focused on three SNPs: rs3818626 in SMC2, rs79447092 in ARHGEF7 and rs9895829 in TP53. We found that these three SNPs were associated with pancreatic cancer risk [odds ratio (OR) = 1.12, 95% confidence interval (CI) = 1.07-1.17 and P = 2.20E-06 for the rs3818626 C allele; OR = 0.76, CI = 0.66-0.88 and P = 1.46E-04 for the rs79447092 A allele and OR = 0.82, CI = 0.74-0.91 and P = 1.51E-04 for the rs9895829 G allele]. Their joint effect as the number of protective genotypes also showed a significant association with pancreatic cancer risk (trend test P ≤ 0.001). Finally, we performed an expression quantitative trait loci analysis and found that rs3818626 and rs9895829 were significantly associated with SMC2 and TP53 messenger RNA expression levels in 373 lymphoblastoid cell lines, respectively. In conclusion, these three representative SNPs may be potentially susceptibility loci for pancreatic cancer and warrant additional validation.

Powerful and efficient SNP-set association tests across multiple phenotypes using GWAS summary data

MOTIVATION

Many GWAS conducted in the past decade have identified tens of thousands of disease related variants, which in total explained only part of the heritability for most traits. There remain many more genetics variants with small effect sizes to be discovered. This has motivated the development of sequencing studies with larger sample sizes and increased resolution of genotyped variants, e.g., the ongoing NHLBI Trans-Omics for Precision Medicine (TOPMed) whole genome sequencing project. An alternative approach is the development of novel and more powerful statistical methods. The current dominating approach in the field of GWAS analysis is the "single trait single variant" association test, despite the fact that most GWAS are conducted in deeply-phenotyped cohorts with many correlated traits measured. In this paper, we aim to develop rigorous methods that integrate multiple correlated traits and multiple variants to improve the power to detect novel variants. In recognition of the difficulty of accessing raw genotype and phenotype data due to privacy and logistic concerns, we develop methods that are applicable to publicly available GWAS summary data.

RESULTS

We build rigorous statistical models for GWAS summary statistics to motivate novel multi-trait SNP-set association tests, including variance component test, burden test and their adaptive test, and develop efficient numerical algorithms to quickly compute their analytical P-values. We implement the proposed methods in an open source R package. We conduct thorough simulation studies to verify the proposed methods rigorously control type I errors at the genome-wide significance level, and further demonstrate their utility via comprehensive analysis of GWAS summary data for multiple lipids traits and glycemic traits. We identified many novel loci that were not detected by the individual trait based GWAS analysis.

AVAILABILITY AND IMPLEMENTATION

We have implemented the proposed methods in an R package freely available at http://www.github.com/baolinwu/MSKAT.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Comparison of brain connectomes by MRI and genomics and its implication in Alzheimer's disease

BACKGROUND

The human brain is complex and interconnected structurally. Brain connectome change is associated with Alzheimer's disease (AD) and other neurodegenerative diseases. Genetics and genomics studies have identified molecular changes in AD; however, the results are often limited to isolated brain regions and are difficult to interpret its findings in respect to brain connectome. The mechanisms of how one brain region impacts the molecular pathways in other regions have not been systematically studied. And how the brain regions susceptible to AD pathology interact with each other at the transcriptome level and how these interactions relate to brain connectome change are unclear.

METHODS

Here, we compared structural brain connectomes defined by probabilistic tracts using diffusion magnetic resonance imaging data in Alzheimer's Disease Neuroimaging Initiative database and a brain transcriptome dataset covering 17 brain regions.

RESULTS

We observed that the changes in diffusion measures associated with AD diagnosis status and the associations were replicated in an independent cohort. The result suggests that disease associated white matter changes are focal. Analysis of the brain connectome by genomic data, tissue-tissue transcriptional synchronization between 17 brain regions, indicates that the regions connected by AD-associated tracts were likely connected at the transcriptome level with high number of tissue-to-tissue correlated (TTC) gene pairs (P = 0.03). And genes involved in TTC gene pairs between white matter tract connected brain regions were enriched in signaling pathways (P = 6.08 × 10-9). Further pathway interaction analysis identified ionotropic glutamate receptor pathway and Toll receptor signaling pathways to be important for tissue-tissue synchronization at the transcriptome level. Transcript profile entailing Toll receptor signaling in the blood was significantly associated with diffusion properties of white matter tracts, notable association between fractional anisotropy and bilateral cingulum angular bundles (Ppermutation = 1.0 × 10-2 and 4.9 × 10-4 for left and right respectively).

CONCLUSIONS

In summary, our study suggests that brain connectomes defined by MRI and transcriptome data overlap with each other.

Novel Genetic Variants of ALG6 and GALNTL4 of the Glycosylation Pathway Predict Cutaneous Melanoma-Specific Survival

Because aberrant glycosylation is known to play a role in the progression of melanoma, we hypothesize that genetic variants of glycosylation pathway genes are associated with the survival of cutaneous melanoma (CM) patients. To test this hypothesis, we used a Cox proportional hazards regression model in a single-locus analysis to evaluate associations between 34,096 genetic variants of 227 glycosylation pathway genes and CM disease-specific survival (CMSS) using genotyping data from two previously published genome-wide association studies. The discovery dataset included 858 CM patients with 95 deaths from The University of Texas MD Anderson Cancer Center, and the replication dataset included 409 CM patients with 48 deaths from Harvard University nurse/physician cohorts. In the multivariable Cox regression analysis, we found that two novel single-nucleotide polymorphisms (SNPs) (ALG6 rs10889417 G>A and GALNTL4 rs12270446 G>C) predicted CMSS, with an adjusted hazards ratios of 0.60 (95% confidence interval = 0.44-0.83 and p = 0.002) and 0.66 (0.52-0.84 and 0.004), respectively. Subsequent expression quantitative trait loci (eQTL) analysis revealed that ALG6 rs10889417 was associated with mRNA expression levels in the cultured skin fibroblasts and whole blood cells and that GALNTL4 rs12270446 was associated with mRNA expression levels in the skin tissues (all p < 0.05). Our findings suggest that, once validated by other large patient cohorts, these two novel SNPs in the glycosylation pathway genes may be useful prognostic biomarkers for CMSS, likely through modulating their gene expression.

Genetic architecture study of rheumatoid arthritis and juvenile idiopathic arthritis

BACKGROUND

Rheumatoid arthritis and juvenile idiopathic arthritis are two types of autoimmune diseases with inflammation at the joints, occurring to adults and children respectively. There are phenotypic overlaps between these two types of diseases, despite the age difference in patient groups.

METHODS

To systematically compare the genetic architecture of them, we conducted analyses at gene and pathway levels and constructed protein-protein-interaction network based on summary statistics of genome-wide association studies of these two diseases. We examined their difference and similarity at each level.

RESULTS

We observed extensive overlap in significant SNPs and genes at the human leukocyte antigen region. In addition, several SNPs in other regions of the human genome were also significantly associated with both diseases. We found significantly associated genes enriched in 32 pathways shared by both diseases. Excluding genes in the human leukocyte antigen region, significant enrichment is present for pathways like interleukin-27 pathway and NO2-dependent interleukin-12 pathway in natural killer cells.

DISCUSSION

The identification of commonly associated genes and pathways may help in finding population at risk for both diseases, as well as shed light on repositioning and designing drugs for both diseases.

Region-based interaction detection in genome-wide case-control studies

Background

In genome-wide association study (GWAS), conventional interaction detection methods such as BOOST are mostly based on SNP-SNP interactions. Although single nucleotides are the building blocks of human genome, single nucleotide polymorphisms (SNPs) are not necessarily the smallest functional unit for complex phenotypes. Region-based strategies have been proved to be successful in studies aiming at marginal effects.

Methods

We propose a novel region-region interaction detection method named RRIntCC (region-region interaction detection for case-control studies). RRIntCC uses the correlations between individual SNP-SNP interactions based on linkage disequilibrium (LD) contrast test.

Results

Simulation experiments showed that our method can achieve a higher power than conventional SNP-based methods with similar type-I-error rates. When applied to two real datasets, RRIntCC was able to find several significant regions, while BOOST failed to identify any significant results. The source code and the sample data of RRIntCC are available at http://bioinformatics.ust.hk/RRIntCC.html.

Conclusion

In this paper, a new region-based interaction detection method with better performance than SNP-based interaction detection methods has been proposed.

Genome-wide analysis identifies molecular systems and 149 genetic loci associated with income

Socioeconomic position (SEP) is a multi-dimensional construct reflecting (and influencing) multiple socio-cultural, physical, and environmental factors. In a sample of 286,301 participants from UK Biobank, we identify 30 (29 previously unreported) independent-loci associated with income. Using a method to meta-analyze data from genetically-correlated traits, we identify an additional 120 income-associated loci. These loci show clear evidence of functionality, with transcriptional differences identified across multiple cortical tissues, and links to GABAergic and serotonergic neurotransmission. By combining our genome wide association study on income with data from eQTL studies and chromatin interactions, 24 genes are prioritized for follow up, 18 of which were previously associated with intelligence. We identify intelligence as one of the likely causal, partly-heritable phenotypes that might bridge the gap between molecular genetic inheritance and phenotypic consequence in terms of income differences. These results indicate that, in modern era Great Britain, genetic effects contribute towards some of the observed socioeconomic inequalities.

Genome-wide analysis identifies molecular systems and 149 genetic loci associated with income

Socioeconomic position (SEP) is a multi-dimensional construct reflecting (and influencing) multiple socio-cultural, physical, and environmental factors. In a sample of 286,301 participants from UK Biobank, we identify 30 (29 previously unreported) independent-loci associated with income. Using a method to meta-analyze data from genetically-correlated traits, we identify an additional 120 income-associated loci. These loci show clear evidence of functionality, with transcriptional differences identified across multiple cortical tissues, and links to GABAergic and serotonergic neurotransmission. By combining our genome wide association study on income with data from eQTL studies and chromatin interactions, 24 genes are prioritized for follow up, 18 of which were previously associated with intelligence. We identify intelligence as one of the likely causal, partly-heritable phenotypes that might bridge the gap between molecular genetic inheritance and phenotypic consequence in terms of income differences. These results indicate that, in modern era Great Britain, genetic effects contribute towards some of the observed socioeconomic inequalities.

Benefits and limitations of genome-wide association studies

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

Improving the odds of drug development success through human genomics: modelling study

Lack of efficacy in the intended disease indication is the major cause of clinical phase drug development failure. Explanations could include the poor external validity of pre-clinical (cell, tissue, and animal) models of human disease and the high false discovery rate (FDR) in preclinical science. FDR is related to the proportion of true relationships available for discovery (γ), and the type 1 (false-positive) and type 2 (false negative) error rates of the experiments designed to uncover them. We estimated the FDR in preclinical science, its effect on drug development success rates, and improvements expected from use of human genomics rather than preclinical studies as the primary source of evidence for drug target identification. Calculations were based on a sample space defined by all human diseases - the 'disease-ome' - represented as columns; and all protein coding genes - 'the protein-coding genome'- represented as rows, producing a matrix of unique gene- (or protein-) disease pairings. We parameterised the space based on 10,000 diseases, 20,000 protein-coding genes, 100 causal genes per disease and 4000 genes encoding druggable targets, examining the effect of varying the parameters and a range of underlying assumptions, on the inferences drawn. We estimated γ, defined mathematical relationships between preclinical FDR and drug development success rates, and estimated improvements in success rates based on human genomics (rather than orthodox preclinical studies). Around one in every 200 protein-disease pairings was estimated to be causal (γ = 0.005) giving an FDR in preclinical research of 92.6%, which likely makes a major contribution to the reported drug development failure rate of 96%. Observed success rate was only slightly greater than expected for a random pick from the sample space. Values for γ back-calculated from reported preclinical and clinical drug development success rates were also close to the a priori estimates. Substituting genome wide (or druggable genome wide) association studies for preclinical studies as the major information source for drug target identification was estimated to reverse the probability of late stage failure because of the more stringent type 1 error rate employed and the ability to interrogate every potential druggable target in the same experiment. Genetic studies conducted at much larger scale, with greater resolution of disease end-points, e.g. by connecting genomics and electronic health record data within healthcare systems has the potential to produce radical improvement in drug development success rate.

Network-Based Functional Prediction Augments Genetic Association To Predict Candidate Genes for Histamine Hypersensitivity in Mice

Genetic mapping is a primary tool of genetics in model organisms; however, many quantitative trait loci (QTL) contain tens or hundreds of positional candidate genes. Prioritizing these genes for validation is often ad hoc and biased by previous findings. Here we present a technique for prioritizing positional candidates based on computationally inferred gene function. Our method uses machine learning with functional genomic networks, whose links encode functional associations among genes, to identify network-based signatures of functional association to a trait of interest. We demonstrate the method by functionally ranking positional candidates in a large locus on mouse Chr 6 (45.9 Mb to 127.8 Mb) associated with histamine hypersensitivity (Histh). Histh is characterized by systemic vascular leakage and edema in response to histamine challenge, which can lead to multiple organ failure and death. Although Histh risk is strongly influenced by genetics, little is known about its underlying molecular or genetic causes, due to genetic and physiological complexity of the trait. To dissect this complexity, we ranked genes in the Histh locus by predicting functional association with multiple Histh-related processes. We integrated these predictions with new single nucleotide polymorphism (SNP) association data derived from a survey of 23 inbred mouse strains and congenic mapping data. The top-ranked genes included Cxcl12, Ret, Cacna1c, and Cntn3, all of which had strong functional associations and were proximal to SNPs segregating with Histh. These results demonstrate the power of network-based computational methods to nominate highly plausible quantitative trait genes even in challenging cases involving large QTL and extreme trait complexity.

Heritability of human visual contour integration-an integrated genomic study

Contour integration, a key visual function to deal with occlusion and discontinuity in natural scenes, is essential to human survival. However, individuals are not equally well equipped with this ability. In particular, contour integration deficiencies are commonly detected in patients with mental disorders, especially schizophrenia. To understand the underlying sources of these individual differences, the current study investigated the genetic basis of contour integration in humans. A total of 2619 normal participants were tested on their ability to detect continuous contours embedded in a cluttered background. Quantitative genomic analysis was performed, involving heritability estimation based on single nucleotide polymorphisms (SNPs) and association testing at SNP, gene, and pathway levels. Heritability estimation showed that common SNPs contributed 49.5% (standard error of the mean = 15.6%) of overall phenotypic variation, indicating moderate heritability of contour integration. Two-stage genome-wide association analysis (GWAS) detected four SNPs reaching genome-wide significance in the discovery test (N = 1931) but not passing the replication test (N = 688). Gene-level analysis further revealed a significant genome-wide association of a microRNA-encoding gene MIR1178 in both the discovery and replication cohorts. Another gene poly(A)-binding protein nuclear 1 like, cytoplasmic (PABPN1L) showed suggestive significance in the discovery cohort (p < 1 × 10-4) and was replicated in the replication cohort (p = 0.009). The pathway analysis did not detect any significant pathway. Taken together, this study identified significant gene associations with contour integration and provided support for a genetic transmission of the ability to perceive continuous contours in the environment.

Translating GWAS findings into therapies for depression and anxiety disorders: gene-set analyses reveal enrichment of psychiatric drug classes and implications for drug repositioning

BACKGROUND

Depression and anxiety disorders (AD) are the first and sixth leading causes of disability worldwide. Despite their high prevalence and significant disability resulted, there are limited advances in new drug development. Recently, genome-wide association studies (GWAS) have greatly advanced our understanding of the genetic basis underlying psychiatric disorders.

METHODS

Here we employed gene-set analyses of GWAS summary statistics for drug repositioning. We explored five related GWAS datasets, including two on major depressive disorder (MDD2018 and MDD-CONVERGE, with the latter focusing on severe melancholic depression), one on AD, and two on depressive symptoms and neuroticism in the population. We extracted gene-sets associated with each drug from DSigDB and examined their association with each GWAS phenotype. We also performed repositioning analyses on meta-analyzed GWAS data, integrating evidence from all related phenotypes.

RESULTS

Importantly, we showed that the repositioning hits are generally enriched for known psychiatric medications or those considered in clinical trials. Enrichment was seen for antidepressants and anxiolytics but also for antipsychotics. We also revealed new candidates or drug classes for repositioning, some of which were supported by experimental or clinical studies. For example, the top repositioning hit using meta-analyzed p values was fendiline, which was shown to produce antidepressant-like effects in mouse models by inhibition of acid sphingomyelinase.

CONCLUSION

Taken together, our findings suggest that human genomic data such as GWAS are useful in guiding drug discoveries for depression and AD.

Detecting Weak Signals by Combining Small P-Values in Genetic Association Studies

We approach the problem of combining top-ranking association statistics or P-values from a new perspective which leads to a remarkably simple and powerful method. Statistical methods, such as the rank truncated product (RTP), have been developed for combining top-ranking associations, and this general strategy proved to be useful in applications for detecting combined effects of multiple disease components. To increase power, these methods aggregate signals across top ranking single nucleotide polymorphisms (SNPs), while adjusting for their total number assessed in a study. Analytic expressions for combined top statistics or P-values tend to be unwieldy, which complicates interpretation and practical implementation and hinders further developments. Here, we propose the augmented rank truncation (ART) method that retains main characteristics of the RTP but is substantially simpler to implement. ART leads to an efficient form of the adaptive algorithm, an approach where the number of top ranking SNPs is varied to optimize power. We illustrate our methods by strengthening previously reported associations of μ-opioid receptor variants with sensitivity to pain.

A flexible and nearly optimal sequential testing approach to randomized testing: QUICK-STOP

In the analysis of current life science datasets, we often encounter scenarios in which the application of asymptotic theory to hypothesis testing can be problematic. Besides improved asymptotic results, permutation/simulation-based tests are a general approach to address this issue. However, these randomized tests can impose a massive computational burden, for example, in scenarios in which large numbers of statistical tests are computed, and the specified significance level is very small. Stopping rules aim to assess significance with the smallest possible number of draws while controlling the probabilities of errors due to statistical uncertainty. In this communication, we derive a general stopping rule, QUICK-STOP, based on the sequential testing theory that is easy to implement, controls the error probabilities rigorously, and is nearly optimal in terms of expected draws. In a simulation study, we show that our approach outperforms current stopping approaches for general randomized tests by factor 10 and does not impose an additional computational burden. We illustrate our approach by applying our stopping rule to a single-variant analysis of a whole-genome sequencing study for lung function.

A powerful conditional gene-based association approach implicated functionally important genes for schizophrenia

MOTIVATION

It remains challenging to unravel new susceptibility genes of complex diseases and the mechanisms in genome-wide association studies. There are at least two difficulties, isolation of the genuine susceptibility genes from many indirectly associated genes and functional validation of these genes.

RESULTS

We first proposed a novel conditional gene-based association test which can use only summary statistics to isolate independently associated genes of a disease. Applying this method, we detected 185 genes of independent association with schizophrenia. We then designed an in-silico experiment based on expression/co-expression to systematically validate pathogenic potential of these genes. We found that genes of independent association with schizophrenia formed more co-expression pairs in normal post-natal but not pre-natal human brain regions than expected. Interestingly, no co-expression enrichment was found in the brain regions of schizophrenia patients. The genes with independent association also had more significant P-values for differential expression between schizophrenia patients and controls in the brain regions. In contrast, indirectly associated genes or associated genes by other widely-used gene-based tests had no such differential expression and co-expression patterns. In summary, this conditional gene-based association test is effective for isolating directly associated genes from indirectly associated genes, and the results insightfully suggest that common variants might contribute to schizophrenia largely by distorting expression and co-expression in post-natal brains.

AVAILABILITY AND IMPLEMENTATION

The conditional gene-based association test has been implemented in a platform 'KGG' in Java and is publicly available at http://grass.cgs.hku.hk/limx/kgg/.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Sex-dependent and sex-independent regulatory systems of size variation in natural populations

Size of organs/organisms is a polygenic trait. Many of the growth-regulatory genes constitute conserved growth signaling pathways. However, how these multiple genes are orchestrated at the systems level to attain the natural variation in size including sexual size dimorphism is mostly unknown. Here we take a multi-layered systems omics approach to study size variation in the Drosophila wing. We show that expression levels of many critical growth regulators such as Wnt and TGFβ pathway components significantly differ between sexes but not between lines exhibiting size differences within each sex, suggesting a primary role of these regulators in sexual size dimorphism. Only a few growth genes including a receptor of steroid hormone ecdysone exhibit association with between-line size differences. In contrast, we find that between-line size variation is largely regulated by genes with a diverse range of cellular functions, most of which have never been implicated in growth. In addition, we show that expression quantitative trait loci (eQTLs) linked to these novel growth regulators accurately predict population-wide, between-line wing size variation. In summary, our study unveils differential gene regulatory systems that control wing size variation between and within sexes.

Gene-based analysis of ADHD using PASCAL: a biological insight into the novel associated genes

BACKGROUND

Attention-Deficit Hyperactivity Disorder (ADHD) is a complex neurodevelopmental disorder (NDD) which may significantly impact on the affected individual's life. ADHD is acknowledged to have a high heritability component (70-80%). Recently, a meta-analysis of GWAS (Genome Wide Association Studies) has demonstrated the association of several independent loci. Our main aim here, is to apply PASCAL (pathway scoring algorithm), a new gene-based analysis (GBA) method, to the summary statistics obtained in this meta-analysis. PASCAL will take into account the linkage disequilibrium (LD) across genomic regions in a different way than the most commonly employed GBA methods (MAGMA or VEGAS (Versatile Gene-based Association Study)). In addition to PASCAL analysis a gene network and an enrichment analysis for KEGG and GO terms were carried out. Moreover, GENE2FUNC tool was employed to create gene expression heatmaps and to carry out a (DEG) (Differentially Expressed Gene) analysis using GTEX v7 and BrainSpan data.

RESULTS

PASCAL results have revealed the association of new loci with ADHD and it has also highlighted other genes previously reported by MAGMA analysis. PASCAL was able to discover new associations at a gene level for ADHD: FEZF1 (p-value: 2.2 × 10- 7) and FEZF1-AS1 (p-value: 4.58 × 10- 7). In addition, PASCAL has been able to highlight association of other genes that share the same LD block with some previously reported ADHD susceptibility genes. Gene network analysis has revealed several interactors with the associated ADHD genes and different GO and KEGG terms have been associated. In addition, GENE2FUNC has demonstrated the existence of several up and down regulated expression clusters when the associated genes and their interactors were considered.

CONCLUSIONS

PASCAL has been revealed as an efficient tool to extract additional information from previous GWAS using their summary statistics. This study has identified novel ADHD associated genes that were not previously reported when other GBA methods were employed. Moreover, a biological insight into the biological function of the ADHD associated genes across brain regions and neurodevelopmental stages is provided.

Genetic and Expression Analysis of COPI Genes and Alzheimer's Disease Susceptibility

Alzheimer's disease (AD) is the most common neurodegenerative disease in the elderly and the leading cause of dementia in humans. Evidence shows that cellular trafficking and recycling machineries are associated with AD risk. A recent study found that the coat protein complex I (COPI)-dependent trafficking in vivo could significantly reduce amyloid plaques in the cortex and hippocampus of neurological in the AD mouse models and identified 12 single-nucleotide polymorphisms in COPI genes to be significantly associated with increased AD risk using 6,795 samples. Here, we used a large-scale GWAS dataset to investigate the potential association between the COPI genes and AD susceptibility by both SNP and gene-based tests. The results showed that only rs9898218 was associated with AD risk with P = 0.017. We further conducted an expression quantitative trait loci (eQTLs) analysis and found that rs9898218 G allele was associated with increased COPZ2 expression in cerebellar cortex with P = 0.0184. Importantly, the eQTLs analysis in whole blood further indicated that 11 of these 12 genetic variants could significantly regulate the expression of COPI genes. Hence, these findings may contribute to understand the association between COPI genes and AD susceptibility.

Genetic Susceptibility Loci for Cardiovascular Disease and Their Impact on Atherosclerotic Plaques

Supplemental Digital Content is available in the text.

Background:

Atherosclerosis is a chronic inflammatory disease in part caused by lipid uptake in the vascular wall, but the exact underlying mechanisms leading to acute myocardial infarction and stroke remain poorly understood. Large consortia identified genetic susceptibility loci that associate with large artery ischemic stroke and coronary artery disease. However, deciphering their underlying mechanisms are challenging. Histological studies identified destabilizing characteristics in human atherosclerotic plaques that associate with clinical outcome. To what extent established susceptibility loci for large artery ischemic stroke and coronary artery disease relate to plaque characteristics is thus far unknown but may point to novel mechanisms.

Methods:

We studied the associations of 61 established cardiovascular risk loci with 7 histological plaque characteristics assessed in 1443 carotid plaque specimens from the Athero-Express Biobank Study. We also assessed if the genotyped cardiovascular risk loci impact the tissue-specific gene expression in 2 independent biobanks, Biobank of Karolinska Endarterectomy and Stockholm Atherosclerosis Gene Expression.

Results:

A total of 21 established risk variants (out of 61) nominally associated to a plaque characteristic. One variant (rs12539895, risk allele A) at 7q22 associated to a reduction of intraplaque fat, P=5.09×10⁻⁶ after correction for multiple testing. We further characterized this 7q22 Locus and show tissue-specific effects of rs12539895 on HBP1 expression in plaques and COG5 expression in whole blood and provide data from public resources showing an association with decreased LDL (low-density lipoprotein) and increase HDL (high-density lipoprotein) in the blood.

Conclusions:

Our study supports the view that cardiovascular susceptibility loci may exert their effect by influencing the atherosclerotic plaque characteristics.

Genetic variations associated with response to dutasteride in the treatment of male subjects with androgenetic alopecia

Dutasteride, a dual inhibitor of both type I and II 5α-reductases, is used to treat male pattern hair loss (MPHL). However, patient response to dutasteride varies in each individual, the cause of which is yet to be identified. To identify genetic variants associated with response to dutasteride treatment for MPHL, a total of 42 men with moderate MPHL who had been treated with dutasteride for 6 months were genotyped and analysed by quantitative linear regression, case-control association tests, and Fisher’s exact test. The synonymous single nucleotide polymorphism (SNP) rs72623193 in DHRS9 was most significantly associated with response to dutasteride, followed by the non-synonymous SNP rs2241057 in CYP26B1. Additionally, variants in ESR1, SRD5A1, CYP19A1, and RXRG are suggested to be associated with response to dutasteride. Cumulative effect and interaction among these SNPs were presented in both additive and non-additive models.

Integrating Genome-Wide Association Studies With Pathway Analysis and Gene Expression Analysis Highlights Novel Osteoarthritis Risk Pathways and Genes

Osteoarthritis (OA) is the most common degenerative joint disorder worldwide. To identify more genetic signals, genome-wide association study (GWAS) has been widely used and elucidated some OA susceptibility genes. However, these susceptibility genes could only explain only a small part of heritability of OA. It is suggested that the identification of disease-related pathways may contribute to understand the genomic etiology of OA. Here, we integrated the GWAS into pathway analysis to identify novel OA risk pathways. In this study, we first selected 187 independent genetic variants identified by GWAS (P < 1.00E−05) and found that most of these genetic variants are noncoding mutations. We then conducted an expression quantitative trait loci analysis and found that 165 of these 187 genetic variants could significantly regulate the expression of nearby genes. Third, we identified OA susceptibility genes corresponding to these genetic variants, conducted a pathway analysis, and identified novel OA-related KEGG pathways, GO biological processes, GO molecular functions, and GO cellular components. In KEGG database, transforming growth factor β signaling pathway is the most significant signal (P = 5.98E−05) and is the only pathway after the BH multiple-test adjustment with false discovery rate (FDR) = 0.02. In GO database, we identified 24 statistically significant GO biological processes, one statistically significant GO molecular function, and five statistically significant GO cellular components (FDR < 0.05). These signals are related with chondrocyte differentiation and development, which are all known biological pathways associated with OA. Finally, we conducted an OA case–control gene expression analysis to evaluate the differential expression of these OA risk genes. Using an OA case–control gene expression analysis, we showed that 44 risk genes were suggestively differentially expressed in OA cases compared with controls (P < 0.05). Three genes, WWP2, COG5, and MAPT, were statistically differentially expressed in OA cases compared with controls (P < 0.05/122 = 4.10E−04). Hence, our findings may contribute to understanding the genomic etiology of OA.

Novel Gene-Based Analysis of ASD GWAS: Insight Into the Biological Role of Associated Genes

Background: Autism spectrum disorder (ASD) is a neurodevelopmental disorder characterized by its significant social impact and high heritability. The latest meta-analysis of ASD GWAS (genome-wide association studies) has revealed the association of several SNPs that were replicated in additional sets of independent samples. However, summary statistics from GWAS can be used to perform a gene-based analysis (GBA). GBA allows to combine all genetic information across the gene to create a single statistic (p-value for each gene). Thus, PASCAL (Pathway scoring algorithm), a novel GBA tool, has been applied to the summary statistics from the latest meta-analysis of ASD. GBA approach (testing the gene as a unit) provides an advantage to perform an accurate insight into the biological ASD mechanisms. Therefore, a gene-network analysis and an enrichment analysis for KEGG and GO terms were carried out. GENE2FUNC was used to create gene expression heatmaps and to carry out differential expression analysis (DEA) across GTEx v7 tissues and Brainspan data. dbMDEGA was employed to perform a DEG analysis between ASD and brain control samples for the associated genes and interactors.

Results: PASCAL has identified the following loci associated with ASD: XRN2, NKX2-4, PLK1S1, KCNN2, NKX2-2, CRHR1-IT1, C8orf74 and LOC644172. While some of these genes were previously reported by MAGMA (XRN2, PLK1S1, and KCNN2), PASCAL has been useful to highlight additional genes. The biological characterization of the ASD-associated genes and their interactors have demonstrated the association of several GO and KEGG terms. Moreover, DEA analysis has revealed several up- and down-regulated clusters. In addition, many of the ASD-associated genes and their interactors have shown association with ASD expression datasets.

Conclusions: This study identifies several associations at a gene level in ASD. Most of them were previously reported by MAGMA. This fact proves that PASCAL is an efficient GBA tool to extract additional information from previous GWAS. In addition, this study has characterized for the first time the biological role of the ASD-associated genes across brain regions, neurodevelopmental stages, and ASD gene-expression datasets.

Efficient pathway enrichment and network analysis of GWAS summary data using GSA-SNP2

Pathway-based analysis in genome-wide association study (GWAS) is being widely used to uncover novel multi-genic functional associations. Many of these pathway-based methods have been used to test the enrichment of the associated genes in the pathways, but exhibited low powers and were highly affected by free parameters. We present the novel method and software GSA-SNP2 for pathway enrichment analysis of GWAS P-value data. GSA-SNP2 provides high power, decent type I error control and fast computation by incorporating the random set model and SNP-count adjusted gene score. In a comparative study using simulated and real GWAS data, GSA-SNP2 exhibited high power and best prioritized gold standard positive pathways compared with six existing enrichment-based methods and two self-contained methods (alternative pathway analysis approach). Based on these results, the difference between pathway analysis approaches was investigated and the effects of the gene correlation structures on the pathway enrichment analysis were also discussed. In addition, GSA-SNP2 is able to visualize protein interaction networks within and across the significant pathways so that the user can prioritize the core subnetworks for further studies. GSA-SNP2 is freely available at https://sourceforge.net/projects/gsasnp2.

The GWAS Risk Genes for Depression May Be Actively Involved in Alzheimer's Disease

Depression is one of the most frequent psychiatric symptoms observed in people during the development of Alzheimer's disease (AD). We hypothesized that genetic factors conferring risk of depression might affect AD development. In this study, we screened 31 genes, which were located in 19 risk loci for major depressive disorder (MDD) identified by two recent large genome-wide association studies (GWAS), in AD patients at the genomic and transcriptomic levels. Association analysis of common variants was performed by using summary statistics of the International Genomics of Alzheimer's Project (IGAP), and association analysis of rare variants was conducted by sequencing the entire coding region of the 31 MDD risk genes in 107 Han Chinese patients with early-onset and/or familial AD. We also quantified the mRNA expression alterations of these MDD risk genes in brain tissues of AD patients and AD mouse models, followed by protein-protein interaction network prediction to show their potential effects in AD pathways. We found that common and rare variants of L3MBTL2 were significantly associated with AD. mRNA expression levels of 18 MDD risk genes, in particular SORCS3 and OAT, were differentially expressed in AD brain tissues. 13 MDD risk genes were predicted to physically interact with core AD genes. The involvement of HACE1, NEGR1, and SLC6A15 in AD was supported by convergent lines of evidence. Taken together, our results showed that MDD risk genes might play an active role in AD pathology and supported the notion that depression might be the "common cold" of psychiatry.

Genetic variation across RNA metabolism and cell death gene networks is implicated in the semantic variant of primary progressive aphasia

The semantic variant of primary progressive aphasia (svPPA) is a clinical syndrome characterized by neurodegeneration and progressive loss of semantic knowledge. Unlike many other forms of frontotemporal lobar degeneration (FTLD), svPPA has a highly consistent underlying pathology composed of TDP-43 (a regulator of RNA and DNA transcription metabolism). Previous genetic studies of svPPA are limited by small sample sizes and a paucity of common risk variants. Despite this, svPPA's relatively homogenous clinicopathologic phenotype makes it an ideal investigative model to examine genetic processes that may drive neurodegenerative disease. In this study, we used GWAS metadata, tissue samples from pathologically confirmed frontotemporal lobar degeneration, and in silico techniques to identify and characterize protein interaction networks associated with svPPA risk. We identified 64 svPPA risk genes that interact at the protein level. The protein pathways represented in this svPPA gene network are critical regulators of RNA metabolism and cell death, such as SMAD proteins and NOTCH1. Many of the genes in this network are involved in TDP-43 metabolism. Contrary to the conventional notion that svPPA is a clinical syndrome with few genetic risk factors, our analyses show that svPPA risk is complex and polygenic in nature. Risk for svPPA is likely driven by multiple common variants in genes interacting with TDP-43, along with cell death,x` working in combination to promote neurodegeneration.

KMgene: a unified R package for gene-based association analysis for complex traits

Summary

In this report, we introduce an R package KMgene for performing gene-based association tests for familial, multivariate or longitudinal traits using kernel machine (KM) regression under a generalized linear mixed model framework. Extensive simulations were performed to evaluate the validity of the approaches implemented in KMgene.

Availability and implementation

http://cran.r-project.org/web/packages/KMgene.

Supplementary information

Supplementary data are available at Bioinformatics online.

Mendelian randomization integrating GWAS and eQTL data reveals genetic determinants of complex and clinical traits

Genome-wide association studies (GWAS) have identified thousands of variants associated with complex traits, but their biological interpretation often remains unclear. Most of these variants overlap with expression QTLs, indicating their potential involvement in regulation of gene expression. Here, we propose a transcriptome-wide summary statistics-based Mendelian Randomization approach (TWMR) that uses multiple SNPs as instruments and multiple gene expression traits as exposures, simultaneously. Applied to 43 human phenotypes, it uncovers 3,913 putatively causal gene-trait associations, 36% of which have no genome-wide significant SNP nearby in previous GWAS. Using independent association summary statistics, we find that the majority of these loci were missed by GWAS due to power issues. Noteworthy among these links is educational attainment-associated BSCL2, known to carry mutations leading to a Mendelian form of encephalopathy. We also find pleiotropic causal effects suggestive of mechanistic connections. TWMR better accounts for pleiotropy and has the potential to identify biological mechanisms underlying complex traits.

Evidence that 6q25.1 variant rs6931104 confers susceptibility to chronic myeloid leukemia through RMND1 regulation

Chronic myeloid leukemia (CML) is a clonal myeloproliferative disorder. Our previous study reported novel loci as genetic markers associated with increased susceptibility to CML. The present study conducted an expression quantitative trait loci (eQTL) analysis to confirm that the single nucleotide polymorphisms (SNPs) at these loci affect the expression of candidate CML-susceptible genes. We identified that three SNPs (rs963193, rs6931104, and rs9371517) were related to the gene expression pattern of RMND1 (Required For Meiotic Nuclear Division 1 Homolog) in both granulocytes and mononuclear cells from 83 healthy donors. Furthermore, reduced expression of RMND1 expression was noted in CML patients compared with that in healthy individuals. We used the eQTL browsing tool to assess the regulatory information on the three associated significant SNPs, out of which rs6931104 showed strong evidence of regulatory effects. Chromatin immunoprecipitation (ChIP) assays demonstrated that A alleles of rs6931104 could significantly change the binding affinity of transcription factor (TF) RFX3 compared to the G alleles. Then, we performed in vitro experiments on BCR-ABL1-positive (BCR-ABL1+) cell lines. We found that expression of the CML-susceptible gene RMND1 is affected by the binding affinity of TF RFX3, suggesting that RFX3 plays a role in RMND1 expression. Our findings suggest potential target genes for associations of genetic susceptibility risk loci and provide further insights into the pathogenesis and mechanism of CML.

Mutation and association analyses of dementia-causal genes in Han Chinese patients with early-onset and familial Alzheimer's disease

Alzheimer's disease (AD) is the most common cause of dementia in the elderly. It shares clinical and pathological features with other types of dementia, such as vascular dementia (VaD), Lewy body dementia (LBD), and frontotemporal dementia (FTD). We have hypothesized that there might be an overlapping molecular mechanism and genetic basis to the different types of dementia. In this study, we analyzed the mutation pattern of dementia-causal genes in 169 Han Chinese patients with familial and early-onset AD by using whole exome sequencing or targeted resequencing. We identified 9 potentially pathogenic mutations in the AD-causal genes APP, PSEN1, PSEN2, and 6 mutations in a group of non-AD dementia-causal genes including the FTD-causal gene GRN and the VaD-causal gene NOTCH3. A common splice-site variant rs514492 in the FTD-causal gene VCP showed a positive association with AD risk (P = 0.0003, OR = 1.618), whereas the rare missense variant rs33949390 (p. R 1628P) in the LBD-causal gene LRRK2 showed a protective effect on AD risk (P = 0.0004, OR = 0.170). The presence of putative pathogenic mutations and risk variants in these causal genes for different types of dementia in clinically diagnosed familial and early-onset AD patients suggests a need to screen for mutations of the dementia-causal genes in cases of AD to avoid misdiagnosis. These mutations also support the idea that there are overlapping pathomechanisms between AD and other forms of dementia.

PolyGEE: a generalized estimating equation approach to the efficient and robust estimation of polygenic effects in large-scale association studies

To quantify polygenic effects, i.e. undetected genetic effects, in large-scale association studies, we propose a generalized estimating equation (GEE) based estimation framework. We develop a marginal model for single-variant association test statistics of complex diseases that generalizes existing approaches such as LD Score regression and that is applicable to population-based designs, to family-based designs or to arbitrary combinations of both. We extend the standard GEE approach so that the parameters of the proposed marginal model can be estimated based on working-correlation/linkage-disequilibrium (LD) matrices from external reference panels. Our method achieves substantial efficiency gains over standard approaches, while it is robust against misspecification of the LD structure, i.e. the LD structure of the reference panel can differ substantially from the true LD structure in the study population. In simulation studies and in applications to population-based and family-based studies, we illustrate the features of the proposed GEE framework. Our results suggest that our approach can be up to 100% more efficient than existing methodology.

Genes in immune pathways associated with abnormal white matter integrity in first-episode and treatment-naïve patients with schizophrenia

BACKGROUND

Previous studies have inferred a strong genetic component in schizophrenia. However, the genetic variants involved in the susceptibility to schizophrenia remain unclear.AimsTo detect potential gene pathways and networks associated with schizophrenia, and to explore the relationship between common and rare variants in these pathways and abnormal white matter integrity in schizophrenia.

METHOD

The analysis included 100 first-episode treatment-naïve patients with schizophrenia and 140 healthy controls. A network-based analysis was carried out on the data collected from the Psychiatric Genomics Consortium Phase I (PGC-I). Based on our genome-wide association study and whole-exome sequencing data-sets, we performed a gene-set analysis to detect associations between the combining effects of common and rare genetic variants and abnormal white matter integrity in schizophrenia.

RESULTS

Patients had significantly reduced functional anisotropy in the left and right anterior cingulate cortex, left and right precuneus and extra-nuclear (t = 4.61-5.10, PFDR &lt; 0.01), compared with controls. Generated from co-expression network analysis of the PGC-1 summary statistics of schizophrenia, a subnetwork of 207 genes associated with schizophrenia was identified (P &lt; 0.01), and 176 genes were co-expressed in four gene modules. Functional enrichment analysis for genes in each module revealed that the yellow module was enriched with highly co-expressed, innate immune response genes. Furthermore, rare variants of enriched genes in the yellow module were associated with reduced functional anisotropy in the left anterior cingulate cortex (P = 0.006; Padjusted = 0.024) in patients only.

CONCLUSIONS

The pathogenesis of schizophrenia may be substantially influenced by genes involved in the immune system, via both pathway and network.Declaration of interestsNone.

Shared genes between Alzheimer's disease and ischemic stroke

Aims

Although converging evidence from experimental and epidemiological studies indicates Alzheimer's disease (AD) and ischemic stroke (IS) are related, the genetic basis underlying their links is less well characterized. Traditional SNP‐based genome‐wide association studies (GWAS) have failed to uncover shared susceptibility variants of AD and IS. Therefore, this study was designed to investigate whether pleiotropic genes existed between AD and IS to account for their phenotypic association, although this was not reported in previous studies.

Methods

Taking advantage of large‐scale GWAS summary statistics of AD (17,008 AD cases and 37,154 controls) and IS (10,307 IS cases and 19,326 controls), we performed gene‐based analysis implemented in VEGAS2 and Fisher's meta‐analysis of the set of overlapped genes of nominal significance in both diseases. Subsequently, gene expression analysis in AD‐ or IS‐associated expression datasets was conducted to explore the transcriptional alterations of pleiotropic genes identified.

Results

16 AD‐IS pleiotropic genes surpassed the cutoff for Bonferroni‐corrected significance. Notably, MS4A4A and TREM2, two established AD‐susceptibility genes showed remarkable alterations in the spleens and brains afflicted by IS, respectively. Among the prioritized genes identified by virtue of literature‐based knowledge, most are immune‐relevant genes (EPHA1, MS4A4A, UBE2L3 and TREM2), implicating crucial roles of the immune system in the pathogenesis of AD and IS.

Conclusions

The observation that AD and IS had shared disease‐associated genes offered mechanistic insights into their common pathogenesis, predominantly involving the immune system. More importantly, our findings have important implications for future research directions, which are encouraged to verify the involvement of these candidates in AD and IS and interpret the exact molecular mechanisms of action.

Effects of the interaction between genetic factors and maltreatment on child and adolescent psychiatric disorders

We evaluated the effects of the interaction between child maltreatment (CM) and single nucleotide polymorphisms (SNPs) on development of mental disorders (MD) and psychopathology. We genotyped 720 individuals from a Brazilian community school-based prospective study, focusing on SNPs in 21 genes known to be associated with mental disorders. CM was assessed via a multi-informant-measure, which was previously validated. To test G × CM, we used linear or logistic models depending on variable evaluated (MD or dimensional psychopathology). After Bonferroni multiple comparison correction, we did not find any statistically significant association of G × CM with either MD or psychopathology.

Powerful gene set analysis in GWAS with the Generalized Berk-Jones statistic

A common complementary strategy in Genome-Wide Association Studies (GWAS) is to perform Gene Set Analysis (GSA), which tests for the association between one phenotype of interest and an entire set of Single Nucleotide Polymorphisms (SNPs) residing in selected genes. While there exist many tools for performing GSA, popular methods often include a number of ad-hoc steps that are difficult to justify statistically, provide complicated interpretations based on permutation inference, and demonstrate poor operating characteristics. Additionally, the lack of gold standard gene set lists can produce misleading results and create difficulties in comparing analyses even across the same phenotype. We introduce the Generalized Berk-Jones (GBJ) statistic for GSA, a permutation-free parametric framework that offers asymptotic power guarantees in certain set-based testing settings. To adjust for confounding introduced by different gene set lists, we further develop a GBJ step-down inference technique that can discriminate between gene sets driven to significance by single genes and those demonstrating group-level effects. We compare GBJ to popular alternatives through simulation and re-analysis of summary statistics from a large breast cancer GWAS, and we show how GBJ can increase power by incorporating information from multiple signals in the same gene. In addition, we illustrate how breast cancer pathway analysis can be confounded by the frequency of FGFR2 in pathway lists. Our approach is further validated on two other datasets of summary statistics generated from GWAS of height and schizophrenia.

Beyond genome-wide significance: integrative approaches to the interpretation and extension of GWAS findings for alcohol use disorder

Alcohol use disorder (AUD) is a heritable complex behavior. Due to the highly polygenic nature of AUD, identifying genetic variants that comprise this heritable variation has proved to be challenging. With the exception of functional variants in alcohol metabolizing genes (e.g. ADH1B and ALDH2), few other candidate loci have been confidently linked to AUD. Genome-wide association studies (GWAS) of AUD and other alcohol-related phenotypes have either produced few hits with genome-wide significance or have failed to replicate on further study. These issues reinforce the complex nature of the genetic underpinnings for AUD and suggest that both GWAS studies with larger samples and additional analysis approaches that better harness the nominally significant loci in existing GWAS are needed. Here, we review approaches of interest in the post-GWAS era, including in silico functional analyses; functional partitioning of single nucleotide polymorphism heritability; aggregation of signal into genes and gene networks; and validation of identified loci, genes and gene networks in postmortem brain tissue and across species. These integrative approaches hold promise to illuminate our understanding of the biological basis of AUD; however, we recognize that the main challenge continues to be the extremely polygenic nature of AUD, which necessitates large samples to identify multiple loci associated with AUD liability.

Candidate gene analyses for acute pain and morphine analgesia after pediatric day surgery: African American versus European Caucasian ancestry and dose prediction limits

Acute pain and opioid analgesia demonstrate inter-individual variability and polygenic influence. In 241 children of African American and 277 of European Caucasian ancestry, we sought to replicate select candidate gene associations with morphine dose and postoperative pain and then to estimate dose prediction limits. Twenty-seven single-nucleotide polymorphisms (SNPs) from nine genes (ABCB1, ARRB2, COMT, DRD2, KCNJ6, MC1R, OPRD1, OPRM1, and UGT2B7) met selection criteria and were analyzed along with TAOK3. Few associations replicated: morphine dose (mcg/kg) in African American children and ABCB1 rs1045642 (A allele, β = -9.30, 95% CI: -17.25 to -1.35, p = 0.02) and OPRM1 rs1799971 (G allele, β = 23.19, 95% CI: 3.27-43.11, p = 0.02); KCNJ6 rs2211843 and high pain in African American subjects (T allele, OR 2.08, 95% CI: 1.17-3.71, p = 0.01) and in congruent European Caucasian pain phenotypes; and COMT rs740603 for high pain in European Caucasian subjects (A allele, OR: 0.69, 95% CI: 0.48-0.99, p = 0.046). With age, body mass index, and physical status as covariates, simple top SNP candidate gene models could explain theoretical maximums of 24.2% (European Caucasian) and 14.6% (African American) of morphine dose variances.

Eleven loci with new reproducible genetic associations with allergic disease risk

BACKGROUND

A recent genome-wide association study (GWAS) identified 99 loci that contain genetic risk variants shared between asthma, hay fever, and eczema. Many more risk loci shared between these common allergic diseases remain to be discovered, which could point to new therapeutic opportunities.

OBJECTIVE

We sought to identify novel risk loci shared between asthma, hay fever, and eczema by applying a gene-based test of association to results from a published GWAS that included data from 360,838 subjects.

METHODS

We used approximate conditional analysis to adjust the results from the published GWAS for the effects of the top risk variants identified in that study. We then analyzed the adjusted GWAS results with the EUGENE gene-based approach, which combines evidence for association with disease risk across regulatory variants identified in different tissues. Novel gene-based associations were followed up in an independent sample of 233,898 subjects from the UK Biobank study.

RESULTS

Of the 19,432 genes tested, 30 had a significant gene-based association at a Bonferroni-corrected P value of 2.5 × 10-6. Of these, 20 were also significantly associated (P < .05/30 = .0016) with disease risk in the replication sample, including 19 that were located in 11 loci not reported to contain allergy risk variants in previous GWASs. Among these were 9 genes with a known function that is directly relevant to allergic disease: FOSL2, VPRBP, IPCEF1, PRR5L, NCF4, APOBR, IL27, ATXN2L, and LAT. For 4 genes (eg, ATXN2L), a genetically determined decrease in gene expression was associated with decreased allergy risk, and therefore drugs that inhibit gene expression or function are predicted to ameliorate disease symptoms. The opposite directional effect was observed for 14 genes, including IL27, a cytokine known to suppress TH2 responses.

CONCLUSION

Using a gene-based approach, we identified 11 risk loci for allergic disease that were not reported in previous GWASs. Functional studies that investigate the contribution of the 19 associated genes to the pathophysiology of allergic disease and assess their therapeutic potential are warranted.

Prioritizing Crohn's disease genes by integrating association signals with gene expression implicates monocyte subsets

Genome-wide association studies have identified ~170 loci associated with Crohn's disease (CD) and defining which genes drive these association signals is a major challenge. The primary aim of this study was to define which CD locus genes are most likely to be disease related. We developed a gene prioritization regression model (GPRM) by integrating complementary mRNA expression datasets, including bulk RNA-Seq from the terminal ileum of 302 newly diagnosed, untreated CD patients and controls, and in stimulated monocytes. Transcriptome-wide association and co-expression network analyses were performed on the ileal RNA-Seq datasets, identifying 40 genome-wide significant genes. Co-expression network analysis identified a single gene module, which was substantially enriched for CD locus genes and most highly expressed in monocytes. By including expression-based and epigenetic information, we refined likely CD genes to 2.5 prioritized genes per locus from an average of 7.8 total genes. We validated our model structure using cross-validation and our prioritization results by protein-association network analyses, which demonstrated significantly higher CD gene interactions for prioritized compared with non-prioritized genes. Although individual datasets cannot convey all of the information relevant to a disease, combining data from multiple relevant expression-based datasets improves prediction of disease genes and helps to further understanding of disease pathogenesis.

Comparison of methods for multivariate gene-based association tests for complex diseases using common variants

Complex diseases are usually associated with multiple correlated phenotypes, and the analysis of composite scores or disease status may not fully capture the complexity (or multidimensionality). Joint analysis of multiple disease-related phenotypes in genetic tests could potentially increase power to detect association of a disease with common SNPs (or genes). Gene-based tests are designed to identify genes containing multiple risk variants that individually are weakly associated with a univariate trait. We combined three multivariate association tests (O'Brien method, TATES, and MultiPhen) with two gene-based association tests (GATES and VEGAS) and compared performance (type I error and power) of six multivariate gene-based methods using simulated data. Data (n = 2000) for genetic sequence and correlated phenotypes were simulated by varying causal variant proportions and phenotype correlations for various scenarios. These simulations showed that two multivariate association tests (TATES and MultiPhen, but not O'Brien) paired with VEGAS have inflated type I error in all scenarios, while the three multivariate association tests paired with GATES have correct type I error. MultiPhen paired with GATES has higher power than competing methods if the correlations among phenotypes are low (r < 0.57). We applied these gene-based association methods to a GWAS dataset from the Alzheimer's Disease Genetics Consortium containing three neuropathological traits related to Alzheimer disease (neuritic plaque, neurofibrillary tangles, and cerebral amyloid angiopathy) measured in 3500 autopsied brains. Gene-level significant evidence (P < 2.7 × 10^-6) was identified in a region containing three contiguous genes (TRAPPC12, TRAPPC12-AS1, ADI1) using O'Brien and VEGAS. Gene-wide significant associations were not observed in univariate gene-based tests.

Genome-Wide Gene-Environment Interaction Analysis Using Set-Based Association Tests

The identification of gene-environment interactions (G × E) may eventually guide health-related choices and medical interventions for complex diseases. More powerful methods must be developed to identify G × E. The “adaptive combination of Bayes factors method” (ADABF) has been proposed as a powerful genome-wide polygenic approach to detect G × E. In this work, we evaluate its performance when serving as a gene-based G × E test. We compare ADABF with six tests including the “Set-Based gene-EnviRonment InterAction test” (SBERIA), “gene-environment set association test” (GESAT), etc. With extensive simulations, SBERIA and ADABF are found to be more powerful than other G × E tests. However, SBERIA suffers from a power loss when 50% SNP main effects are in the same direction with the SNP × E interaction effects while 50% are in the opposite direction. We further applied these seven G × E methods to the Taiwan Biobank data to explore gene× alcohol interactions on blood pressure levels. The ADAMTS7P1 gene at chromosome 15q25.2 was detected to interact with alcohol consumption on diastolic blood pressure (p = 9.5 × 10⁻⁷, according to the GESAT test). At this gene, the P-values provided by other six tests all reached the suggestive significance level (p < 5 × 10⁻⁵). Regarding the computation time required for a genome-wide G × E analysis, SBERIA is the fastest method, followed by ADABF. Considering the validity, power performance, robustness, and computation time, ADABF is recommended for genome-wide G × E analyses.

GWAS and network analysis of co-occurring nicotine and alcohol dependence identifies significantly associated alleles and network

Alcohol dependence (AD) and nicotine dependence (ND) co-occur frequently (AD+ND). We integrated SNP-based, gene-based, and protein-protein interaction network analyses to identify shared risk genes or gene subnetworks for AD+ND in African Americans (AAs, N = 2,094) and European Americans (EAs, N = 1,207). The DSM-IV criterion counts for AD and ND were modeled as two dependent variables in a multivariate linear mixed model, and analyzed separately for the two populations. The most significant SNP was rs6579845 in EAs (p < 1.29 × 10^-8 ) in GM2A, which encodes GM2 ganglioside activator, and is a cis-expression quantitative locus that affects GM2A expression in blood and brain tissues. However, this SNP was not replicated in our another small sample (N = 678). We identified a subnetwork of 24 genes that contributed to the AD+ND criterion counts. In the gene-set analysis for the subnetwork in an independent sample, the Study of Addiction: Genetics and Environment project (predominately EAs), these 24 genes as a set differed in AD+ND versus control subjects in EAs (p = .041). Functional enrichment analysis for this subnetwork revealed that the gene enrichment involved primarily nerve growth factor pathways, and cocaine and amphetamine addiction. In conclusion, we identified a genome-wide significant variant at GM2A and a gene subnetwork underlying the genetic trait of shared AD+ND. These results increase our understanding of the shared (pleiotropic) genetic risk that underlies AD+ND.

Sex-specific gene and pathway modeling of inherited glioma risk

Background

To date, genome-wide association studies (GWAS) have identified 25 risk variants for glioma, explaining 30% of heritable risk. Most histologies occur with significantly higher incidence in males, and this difference is not explained by currently known risk factors. A previous GWAS identified sex-specific glioma risk variants, and this analysis aims to further elucidate risk variation by sex using gene- and pathway-based approaches.

Methods

Results from the Glioma International Case-Control Study were used as a testing set, and results from 3 GWAS were combined via meta-analysis and used as a validation set. Using summary statistics for nominally significant autosomal SNPs (P < 0.01 in a previous meta-analysis) and nominally significant X-chromosome SNPs (P < 0.01), 3 algorithms (Pascal, BimBam, and GATES) were used to generate gene scores, and Pascal was used to generate pathway scores. Results were considered statistically significant in the discovery set when P < 3.3 × 10-6 and in the validation set when P < 0.001 in 2 of 3 algorithms.

Results

Twenty-five genes within 5 regions and 19 genes within 6 regions reached statistical significance in at least 2 of 3 algorithms in males and females, respectively. EGFR was significantly associated with all glioma and glioblastoma in males only and a female-specific association in TERT, all of which remained nominally significant after conditioning on known risk loci. There were nominal associations with the BioCarta telomeres pathway in both males and females.

Conclusions

These results provide additional evidence that there may be differences by sex in genetic risk for glioma. Additional analyses may further elucidate the biological processes through which this risk is conferred.