Gene size matters

Genetic Pathways and Functional Subnetworks for the Complex Nature of Bipolar Disorder in Genome-Wide Association Study

Bipolar disorder is a complex psychiatric trait that is also recognized as a high substantial heritability from a worldwide distribution. The success in identifying susceptibility loci for bipolar disorder (BPD) has been limited due to its complex genetic architecture. Growing evidence from association studies including genome-wide association (GWA) studies points to the need of improved analytic strategies to pinpoint the missing heritability for BPD. More importantly, many studies indicate that BPD has a strong association with dementia. We conducted advanced pathway analytics strategies to investigate synergistic effects of multilocus within biologically functional pathways, and further demonstrated functional effects among proteins in subnetworks to examine mechanisms underlying the complex nature of bipolarity using a GWA dataset for BPD. We allowed bipolar susceptible loci to play a role that takes larger weights in pathway-based analytic approaches. Having significantly informative genes identified from enriched pathways, we further built function-specific subnetworks of protein interactions using MetaCore. The gene-wise scores (i.e., minimum p-value) were corrected for the gene-length, and the results were corrected for multiple tests using Benjamini and Hochberg’s method. We found 87 enriched pathways that are significant for BPD; of which 36 pathways were reported. Most of them are involved with several metabolic processes, neural systems, immune system, molecular transport, cellular communication, and signal transduction. Three significant and function-related subnetworks with multiple hotspots were reported to link with several Gene Ontology processes for BPD. Our comprehensive pathway-network frameworks demonstrated that the use of prior knowledge is promising to facilitate our understanding between complex psychiatric disorders (e.g., BPD) and dementia for the access to the connection and clinical implications, along with the development and progression of dementia.

Gene Size Matters: An Analysis of Gene Length in the Human Genome

While it is expected for gene length to be associated with factors such as intron number and evolutionary conservation, we are yet to understand the connections between gene length and function in the human genome. In this study, we show that, as expected, there is a strong positive correlation between gene length, transcript length, and protein size as well as a correlation with the number of genetic variants and introns. Among tissue-specific genes, we find that the longest transcripts tend to be expressed in the blood vessels, nerves, thyroid, cervix uteri, and the brain, while the smallest transcripts tend to be expressed in the pancreas, skin, stomach, vagina, and testis. We report, as shown previously, that natural selection suppresses changes for genes with longer transcripts and promotes changes for genes with smaller transcripts. We also observe that genes with longer transcripts tend to have a higher number of co-expressed genes and protein-protein interactions, as well as more associated publications. In the functional analysis, we show that bigger transcripts are often associated with neuronal development, while smaller transcripts tend to play roles in skin development and in the immune system. Furthermore, pathways related to cancer, neurons, and heart diseases tend to have genes with longer transcripts, with smaller transcripts being present in pathways related to immune responses and neurodegenerative diseases. Based on our results, we hypothesize that longer genes tend to be associated with functions that are important in the early development stages, while smaller genes tend to play a role in functions that are important throughout the whole life, like the immune system, which requires fast responses.

Using prior information from humans to prioritize genes and gene-associated variants for complex traits in livestock

Genome-Wide Association Studies (GWAS) in large human cohorts have identified thousands of loci associated with complex traits and diseases. For identifying the genes and gene-associated variants that underlie complex traits in livestock, especially where sample sizes are limiting, it may help to integrate the results of GWAS for equivalent traits in humans as prior information. In this study, we sought to investigate the usefulness of results from a GWAS on human height as prior information for identifying the genes and gene-associated variants that affect stature in cattle, using GWAS summary data on samples sizes of 700,000 and 58,265 for humans and cattle, respectively. Using Fisher's exact test, we observed a significant proportion of cattle stature-associated genes (30/77) that are also associated with human height (odds ratio = 5.1, p = 3.1e-10). Result of randomized sampling tests showed that cattle orthologs of human height-associated genes, hereafter referred to as candidate genes (C-genes), were more enriched for cattle stature GWAS signals than random samples of genes in the cattle genome (p = 0.01). Randomly sampled SNPs within the C-genes also tend to explain more genetic variance for cattle stature (up to 13.2%) than randomly sampled SNPs within random cattle genes (p = 0.09). The most significant SNPs from a cattle GWAS for stature within the C-genes did not explain more genetic variance for cattle stature than the most significant SNPs within random cattle genes (p = 0.87). Altogether, our findings support previous studies that suggest a similarity in the genetic regulation of height across mammalian species. However, with the availability of a powerful GWAS for stature that combined data from 8 cattle breeds, prior information from human-height GWAS does not seem to provide any additional benefit with respect to the identification of genes and gene-associated variants that affect stature in cattle.

The metabolome as a link in the genotype-phenotype map for peroxide resistance in the fruit fly, Drosophila melanogaster

BACKGROUND

Genetic association studies that seek to explain the inheritance of complex traits typically fail to explain a majority of the heritability of the trait under study. Thus, we are left with a gap in the map from genotype to phenotype. Several approaches have been used to fill this gap, including those that attempt to map endophenotype such as the transcriptome, proteome or metabolome, that underlie complex traits. Here we used metabolomics to explore the nature of genetic variation for hydrogen peroxide (H2O2) resistance in the sequenced inbred Drosophila Genetic Reference Panel (DGRP).

RESULTS

We first studied genetic variation for H2O2 resistance in 179 DGRP lines and along with identifying the insulin signaling modulator u-shaped and several regulators of feeding behavior, we estimate that a substantial amount of phenotypic variation can be explained by a polygenic model of genetic variation. We then profiled a portion of the aqueous metabolome in subsets of eight 'high resistance' lines and eight 'low resistance' lines. We used these lines to represent collections of genotypes that were either resistant or sensitive to the stressor, effectively modeling a discrete trait. Across the range of genotypes in both populations, flies exhibited surprising consistency in their metabolomic signature of resistance. Importantly, the resistance phenotype of these flies was more easily distinguished by their metabolome profiles than by their genotypes. Furthermore, we found a metabolic response to H2O2 in sensitive, but not in resistant genotypes. Metabolomic data further implicated at least two pathways, glycogen and folate metabolism, as determinants of sensitivity to H2O2. We also discovered a confounding effect of feeding behavior on assays involving supplemented food.

CONCLUSIONS

This work suggests that the metabolome can be a point of convergence for genetic variation influencing complex traits, and can efficiently elucidate mechanisms underlying trait variation.

Polygenic Patterns of Adaptive Introgression in Modern Humans Are Mainly Shaped by Response to Pathogens

Anatomically modern humans carry many introgressed variants from other hominins in their genomes. Some of them affect their phenotype and can thus be negatively or positively selected. Several individual genes have been proposed to be the subject of adaptive introgression, but the possibility of polygenic adaptive introgression has not been extensively investigated yet. In this study, we analyze archaic introgression maps with refined functional enrichment methods to find signals of polygenic adaptation of introgressed variants. We first apply a method to detect sets of connected genes (subnetworks) within biological pathways that present higher-than-expected levels of archaic introgression. We then introduce and apply a new statistical test to distinguish between epistatic and independent selection in gene sets of present-day humans. We identify several known targets of adaptive introgression, and we show that they belong to larger networks of introgressed genes. After correction for genetic linkage, we find that signals of polygenic adaptation are mostly explained by independent and potentially sequential selection episodes. However, we also find some gene sets where introgressed variants present significant signals of epistatic selection. Our results confirm that archaic introgression has facilitated local adaptation, especially in immunity related and metabolic functions and highlight its involvement in a coordinated response to pathogens out of Africa.

Exploratory analysis of genetic variants influencing molecular traits in cerebral cortex of suicide completers

Genetic factors have been implicated in suicidal behavior. It has been suggested that one of the roles of genetic factors in suicide could be represented by the effect of genetic variants on gene expression regulation. Alteration in the expression of genes participating in multiple biological systems in the suicidal brain has been demonstrated, so it is imperative to identify genetic variants that could influence gene expression or its regulatory mechanisms. In this study, we integrated DNA methylation, gene expression, and genotype data from the prefrontal cortex of suicides to identify genetic variants that could be factors in the regulation of gene expression, generally called quantitative trait locus (xQTLs). We identify 6,224 methylation quantitative trait loci and 2,239 expression quantitative trait loci (eQTLs) in the prefrontal cortex of suicide completers. The xQTLs identified influence the expression of genes involved in neurodevelopment and cell organization. Two of the eQTLs identified (rs8065311 and rs1019238) were previously associated with cannabis dependence, highlighting a candidate genetic variant for the increased suicide risk in subjects with substance use disorders. Our findings suggest that genetic variants may regulate gene expression in the prefrontal cortex of suicides through the modulation of promoter and enhancer activity, and to a lesser extent, binding transcription factors.

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.

Regional enrichment analyses on genetic profiles for schizophrenia and bipolar disorder

Both schizophrenia (SZ) and bipolar disorder (BD) are highly heritable psychiatric disorders. The significant genomic risk loci are of great importance but with no guarantee of known functional impact and they cannot totally explain the genetic inheritance. In this study we present regional enrichment analyses across the genome, aiming to strike a balance between individual risk loci and integrated regional effects. Chromosomes were partitioned into 2 million base-pair regions (indicated by an underscore sign in the cytogenetic bands) on which enrichment tests are performed. In the discovery phase, we leverage the Psychiatric Genomics Consortium SZ and BD initial association test results for European Ancestry (EA) population and dbGAP SNP data for African Ancestry (AA) population. 78 and 48 regions show significantly enriched associations with SZ and BD respectively in the EA population, and nine are in common including MHC, 3p21.1, 7p22.3_2, 2q32.3_2, 8q24.3_4, and 19q13.33_1. The most unique SZ associated region is 1p21.3_3, while the most unique BD associated region is 6q25.2_1. For the AA population fewer regions are discovered with only 10% overlapping with that of EA population. A replication test using Wellcome Trust Case Control Consortium data for EA population verified 9% of the SZ enriched regions and 40% of the BD enriched regions. In summary, we showed that regional enrichment analyses produce reliable genetic association profiles using about one tenth of samples compared to single base-pair genome wide association approach. The identified association regions will be useful for further genetic functional studies.

Revealing new candidate genes for reproductive traits in pigs: combining Bayesian GWAS and functional pathways

Background

Reproductive traits such as number of stillborn piglets (SB) and number of teats (NT) have been evaluated in many genome-wide association studies (GWAS). Most of these GWAS were performed under the assumption that these traits were normally distributed. However, both SB and NT are discrete (e.g. count) variables. Therefore, it is necessary to test for better fit of other appropriate statistical models based on discrete distributions. In addition, although many GWAS have been performed, the biological meaning of the identified candidate genes, as well as their functional relationships still need to be better understood. Here, we performed and tested a Bayesian treatment of a GWAS model assuming a Poisson distribution for SB and NT in a commercial pig line. To explore the biological role of the genes that underlie SB and NT and identify the most likely candidate genes, we used the most significant single nucleotide polymorphisms (SNPs), to collect related genes and generated gene-transcription factor (TF) networks.

Results

Comparisons of the Poisson and Gaussian distributions showed that the Poisson model was appropriate for SB, while the Gaussian was appropriate for NT. The fitted GWAS models indicated 18 and 65 significant SNPs with one and nine quantitative trait locus (QTL) regions within which 18 and 57 related genes were identified for SB and NT, respectively. Based on the related TF, we selected the most representative TF for each trait and constructed a gene-TF network of gene-gene interactions and identified new candidate genes.

Conclusions

Our comparative analyses showed that the Poisson model presented the best fit for SB. Thus, to increase the accuracy of GWAS, counting models should be considered for this kind of trait. We identified multiple candidate genes (e.g. PTP4A2, NPHP1, and CYP24A1 for SB and YLPM1, SYNDIG1L, TGFB3, and VRTN for NT) and TF (e.g. NF-κB and KLF4 for SB and SOX9 and ELF5 for NT), which were consistent with known newborn survival traits (e.g. congenital heart disease in fetuses and kidney diseases and diabetes in the mother) and mammary gland biology (e.g. mammary gland development and body length).

Electronic supplementary material

The online version of this article (doi:10.1186/s12711-016-0189-x) contains supplementary material, which is available to authorized users.

Gene-based meta-analysis of genome-wide association studies implicates new loci involved in obesity

To date, genome-wide association studies (GWASs) have identified >100 loci with single variants associated with body mass index (BMI). This approach may miss loci with high allelic heterogeneity; therefore, the aim of the present study was to use gene-based meta-analysis to identify regions with high allelic heterogeneity to discover additional obesity susceptibility loci. We included GWAS data from 123 865 individuals of European descent from 46 cohorts in Stage 1 and Metabochip data from additional 103 046 individuals from 43 cohorts in Stage 2, all within the Genetic Investigation of ANthropometric Traits (GIANT) consortium. Each cohort was tested for association between ∼2.4 million (Stage 1) or ∼200 000 (Stage 2) imputed or genotyped single variants and BMI, and summary statistics were subsequently meta-analyzed in 17 941 genes. We used the 'VErsatile Gene-based Association Study' (VEGAS) approach to assign variants to genes and to calculate gene-based P-values based on simulations. The VEGAS method was applied to each cohort separately before a gene-based meta-analysis was performed. In Stage 1, two known (FTO and TMEM18) and six novel (PEX2, MTFR2, SSFA2, IARS2, CEP295 and TXNDC12) loci were associated with BMI (P < 2.8 × 10(-6) for 17 941 gene tests). We confirmed all loci, and six of them were gene-wide significant in Stage 2 alone. We provide biological support for the loci by pathway, expression and methylation analyses. Our results indicate that gene-based meta-analysis of GWAS provides a useful strategy to find loci of interest that were not identified in standard single-marker analyses due to high allelic heterogeneity.

An introductory review of parallel independent component analysis (p-ICA) and a guide to applying p-ICA to genetic data and imaging phenotypes to identify disease-associated biological pathways and systems in common complex disorders

Complex inherited phenotypes, including those for many common medical and psychiatric diseases, are most likely underpinned by multiple genes contributing to interlocking molecular biological processes, along with environmental factors (Owen et al., 2010). Despite this, genotyping strategies for complex, inherited, disease-related phenotypes mostly employ univariate analyses, e.g., genome wide association. Such procedures most often identify isolated risk-related SNPs or loci, not the underlying biological pathways necessary to help guide the development of novel treatment approaches. This article focuses on the multivariate analysis strategy of parallel (i.e., simultaneous combination of SNP and neuroimage information) independent component analysis (p-ICA), which typically yields large clusters of functionally related SNPs statistically correlated with phenotype components, whose overall molecular biologic relevance is inferred subsequently using annotation software suites. Because this is a novel approach, whose details are relatively new to the field we summarize its underlying principles and address conceptual questions regarding interpretation of resulting data and provide practical illustrations of the method.

Explore the Features of Brain-Derived Neurotrophic Factor in Mood Disorders

Objectives

Brain-derived neurotrophic factor (BDNF) plays important roles in neuronal survival and differentiation; however, the effects of BDNF on mood disorders remain unclear. We investigated BDNF from the perspective of various aspects of systems biology, including its molecular evolution, genomic studies, protein functions, and pathway analysis.

Methods

We conducted analyses examining sequences, multiple alignments, phylogenetic trees and positive selection across 12 species and several human populations. We summarized the results of previous genomic and functional studies of pro-BDNF and mature-BDNF (m-BDNF) found in a literature review. We identified proteins that interact with BDNF and performed pathway-based analysis using large genome-wide association (GWA) datasets obtained for mood disorders.

Results

BDNF is encoded by a highly conserved gene. The chordate BDNF genes exhibit an average of 75% identity with the human gene, while vertebrate orthologues are 85.9%-100% identical to human BDNF. No signs of recent positive selection were found. Associations between BDNF and mood disorders were not significant in most of the genomic studies (e.g., linkage, association, gene expression, GWA), while relationships between serum/plasma BDNF level and mood disorders were consistently reported. Pro-BDNF is important in the response to stress; the literature review suggests the necessity of studying both pro- and m-BDNF with regard to mood disorders. In addition to conventional pathway analysis, we further considered proteins that interact with BDNF (I-Genes) and identified several biological pathways involved with BDNF or I-Genes to be significantly associated with mood disorders.

Conclusions

Systematically examining the features and biological pathways of BDNF may provide opportunities to deepen our understanding of the mechanisms underlying mood disorders.

An application of MeSH enrichment analysis in livestock

An integral part of functional genomics studies is to assess the enrichment of specific biological terms in lists of genes found to be playing an important role in biological phenomena. Contrasting the observed frequency of annotated terms with those of the background is at the core of overrepresentation analysis (ORA). Gene Ontology (GO) is a means to consistently classify and annotate gene products and has become a mainstay in ORA. Alternatively, Medical Subject Headings (MeSH) offers a comprehensive life science vocabulary including additional categories that are not covered by GO. Although MeSH is applied predominantly in human and model organism research, its full potential in livestock genetics is yet to be explored. In this study, MeSH ORA was evaluated to discern biological properties of identified genes and contrast them with the results obtained from GO enrichment analysis. Three published datasets were employed for this purpose, representing a gene expression study in dairy cattle, the use of SNPs for genome‐wide prediction in swine and the identification of genomic regions targeted by selection in horses. We found that several overrepresented MeSH annotations linked to these gene sets share similar concepts with those of GO terms. Moreover, MeSH yielded unique annotations, which are not directly provided by GO terms, suggesting that MeSH has the potential to refine and enrich the representation of biological knowledge. We demonstrated that MeSH can be regarded as another choice of annotation to draw biological inferences from genes identified via experimental analyses. When used in combination with GO terms, our results indicate that MeSH can enhance our functional interpretations for specific biological conditions or the genetic basis of complex traits in livestock species.

Novel candidate genes putatively involved in stress fracture predisposition detected by whole-exome sequencing

While genetic factors in all likelihood contribute to stress fracture (SF) pathogenesis, a few studies focusing on candidate genes have previously been reported. The objective of this study is to gain better understanding on the genetic basis of SF in a gene-naive manner. Exome sequence capture followed by massive parallel sequencing of two pooled DNA samples from Israeli combat soldiers was employed: cases with high grade SF and ethnically matched healthy controls. The resulting sequence variants were individually verified using the Sequenom™ platform and the contribution of the genetic alterations was validated in a second cohort of cases and controls. In the discovery set that included DNA pool of cases (n = 34) and controls (n = 60), a total of 1174 variants with >600 reads/variant/DNA pool were identified, and 146 (in 127 genes) of these exhibited statistically significant (P < 0·05) different rates between SF cases and controls after multiple comparisons correction. Subsequent validation of these 146 sequence variants individually in a total of 136 SF cases and 127 controls using the Sequenom™ platform validated 20/146 variants. Of these, three missense mutations (rs7426114, rs4073918, rs3752135 in the NEB, SLC6A18 and SIGLEC12 genes, respectively) and three synonymous mutations (rs2071856, rs2515941, rs716745 in the ELFN2, GRK4, LRRC55 genes) displayed significant different rates in SF cases compared with controls. Exome sequencing seemingly unravelled novel candidate genes as involved in SF pathogenesis and predisposition.

Risk and information evaluation of prioritized genes for complex traits: application to bipolar disorder

Many susceptibility genes for complex traits were identified without conclusive findings. There is a strong need to integrate rapidly accumulated genomic data from multi-dimensional platforms, and to conduct risk evaluation for potential therapeutic and diagnostic usages. We set up an algorithm to computationally search for optimal weight-vector for various data sources, while minimized potential noises. Through gene-prioritization framework, combined scores for the resulting prioritized gene-set were calculated using a genome-wide association (GWA) dataset, following with evaluation using weighted genetic risk score and risk-attributed information using an independent GWA dataset. The significance of association of GWA data was corrected for gene length. Enriched functional pathways were identified for the prioritized gene-set using the Gene Ontology analysis. We illustrated our framework with bipolar disorder. 233 prioritized genes were identified from 10,830 candidates that curated from six platforms. The prioritized genes were significantly enriched (P(adjusted) < 1 × 10(-5)) in 18 biological functions and molecular mechanisms including membrane, synaptic transmission, transmission of nerve impulse, integral to membrane, and plasma membrane. Our risk evaluation demonstrated higher weighted genetic risk score in bipolar patients than controls (P-values ranged from 0.002 to 3.8 × 10(-6)). Substantial risk-information (71%) was extracted from prioritized genes for bipolar illness than other candidate-gene sets. Our evidence-based prioritized gene-set provides opportunity to explore the complex network and to conduct follow-up basic and clinical studies for complex traits.

A review of multivariate analyses in imaging genetics

Recent advances in neuroimaging technology and molecular genetics provide the unique opportunity to investigate genetic influence on the variation of brain attributes. Since the year 2000, when the initial publication on brain imaging and genetics was released, imaging genetics has been a rapidly growing research approach with increasing publications every year. Several reviews have been offered to the research community focusing on various study designs. In addition to study design, analytic tools and their proper implementation are also critical to the success of a study. In this review, we survey recent publications using data from neuroimaging and genetics, focusing on methods capturing multivariate effects accommodating the large number of variables from both imaging data and genetic data. We group the analyses of genetic or genomic data into either a priori driven or data driven approach, including gene-set enrichment analysis, multifactor dimensionality reduction, principal component analysis, independent component analysis (ICA), and clustering. For the analyses of imaging data, ICA and extensions of ICA are the most widely used multivariate methods. Given detailed reviews of multivariate analyses of imaging data available elsewhere, we provide a brief summary here that includes a recently proposed method known as independent vector analysis. Finally, we review methods focused on bridging the imaging and genetic data by establishing multivariate and multiple genotype-phenotype-associations, including sparse partial least squares, sparse canonical correlation analysis, sparse reduced rank regression and parallel ICA. These methods are designed to extract latent variables from both genetic and imaging data, which become new genotypes and phenotypes, and the links between the new genotype-phenotype pairs are maximized using different cost functions. The relationship between these methods along with their assumptions, advantages, and limitations are discussed.

Pathways systematically associated to Hirschsprung's disease

Despite it has been reported that several loci are involved in Hirschsprung's disease, the molecular basis of the disease remains yet essentially unknown. The study of collective properties of modules of functionally-related genes provides an efficient and sensitive statistical framework that can overcome sample size limitations in the study of rare diseases. Here, we present the extension of a previous study of a Spanish series of HSCR trios to an international cohort of 162 HSCR trios to validate the generality of the underlying functional basis of the Hirschsprung's disease mechanisms previously found. The Pathway-Based Analysis (PBA) confirms a strong association of gene ontology (GO) modules related to signal transduction and its regulation, enteric nervous system (ENS) formation and other processes related to the disease. In addition, network analysis recovers sub-networks significantly associated to the disease, which contain genes related to the same functionalities, thus providing an independent validation of these findings. The functional profiles of association obtained for patients populations from different countries were compared to each other. While gene associations were different at each series, the main functional associations were identical in all the five populations. These observations would also explain the reported low reproducibility of associations of individual disease genes across populations.