veqtl-mapper: variance association mapping for molecular phenotypes.

MOTIVATION: Genetic loci associated with the variance of phenotypic traits have been of recent interest as they can be signatures of genetic interactions, gene by environment interactions, parent of origin effects and canalization. We present a fast efficient tool to map loci affecting variance of gene expression and other molecular phenotypes in cis. Results: Applied to the publicly available Geuvadis gene expression dataset, we identify 816 loci associated with variance of gene expression using an additive model, and 32 showing differences in variance between homozygous and heterozygous alleles, signatures of parent of origin effects. AVAILABILITY AND IMPLEMENTATION: Documentation and links to source code and binaries for linux can be found at https://funpopgen.github.io/veqm/ . CONTACT: andrew.brown@unige.ch. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Fast and efficient QTL mapper for thousands of molecular phenotypes.

MOTIVATION: In order to discover quantitative trait loci, multi-dimensional genomic datasets combining DNA-seq and ChiP-/RNA-seq require methods that rapidly correlate tens of thousands of molecular phenotypes with millions of genetic variants while appropriately controlling for multiple testing. RESULTS: We have developed FastQTL, a method that implements a popular cis-QTL mapping strategy in a user- and cluster-friendly tool. FastQTL also proposes an efficient permutation procedure to control for multiple testing. The outcome of permutations is modeled using beta distributions trained from a few permutations and from which adjusted P-values can be estimated at any level of significance with little computational cost. The Geuvadis & GTEx pilot datasets can be now easily analyzed an order of magnitude faster than previous approaches. AVAILABILITY AND IMPLEMENTATION: Source code, binaries and comprehensive documentation of FastQTL are freely available to download at http://fastqtl.sourceforge.net/ CONTACT: emmanouil.dermitzakis@unige.ch or olivier.delaneau@unige.ch SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Pathway analysis of genetic markers associated with a functional MRI faces paradigm implicates polymorphisms in calcium responsive pathways.

Several lines of evidence suggest that common polygenic variation influences brain function in humans. Combining high-density genetic markers with brain imaging techniques is constricted by the practicalities of collecting sufficiently large brain imaging samples. Pathway analysis promises to leverage knowledge on function of genes to detect recurring signals of moderate effect. We adapt this approach, exploiting the deep information collected on brain function by fMRI methods, to identify molecular pathways containing genetic variants which influence brain activation during a commonly applied experiment based on a face matching task (n=246) which was developed to study neural processing of faces displaying negative emotions. Genetic markers moderately associated (p<10(-4)) with whole brain activation phenotypes constructed by applying principal components to contrast maps, were tested for pathway enrichment using permutation based methods. The most significant pathways are related to post NMDA receptor activation events, driven by genetic variants in calcium/calmodulin-dependent protein kinase II (CAMK2G, CAMK2D) and a calcium-regulated nucleotide exchange factor (RASGRF2) in which all are activated by intracellular calcium/calmodulin. The most significant effect of the combined polygenic model were localized to the left inferior frontal gyrus (p=1.03 × 10(-9)), a region primarily involved in semantic processing but also involved in processing negative emotions. These findings suggest that pathway analysis of GWAS results derived from principal component analysis of fMRI data is a promising method, to our knowledge, not previously described.

Effect of DISC1 SNPs on brain structure in healthy controls and patients with a history of psychosis.

Disrupted-in-Schizophrenia-1 (DISC1) has been suggested as a susceptibility locus for a broad spectrum of psychiatric disorders. Risk variants have been associated with brain structural changes, which overlap alterations reported in schizophrenia and bipolar disorder patients. We used genome-wide genotyping data for a Norwegian sample of healthy controls (n = 171) and patients with a history of psychosis (n = 184), to investigate 61 SNPs in the DISC1 region for putative association with structural magnetic resonance imaging (sMRI) measures (hippocampal volume; mean cortical thickness; and total surface area, as well as cortical thickness and area divided into four lobar measures). SNP rs821589 was associated with mean temporal and total brain cortical thickness in controls (P(adjusted) = 0.009 and 0.02, respectively), but not in patients. SNPs rs11122319 and rs1417584 were associated with mean temporal cortical thickness in patients (P(adjusted) = 0.04 and 0.03, respectively), but not in controls, and both SNPs have previously been highly associated with DISC1 gene expression. There were significant genotype × case-control interactions. There was no significant association between SNPs and cortical area or hippocampal volume in controls, or with any of the structural measures in cases, after correction for multiple comparisons. In conclusion, DISC1 SNPs might impact brain structural variation, possibly differently in psychosis patients versus controls, but independent replication will be needed to confirm our findings.

Predicting causal variants affecting expression by using whole-genome sequencing and RNA-seq from multiple human tissues.

Genetic association mapping produces statistical links between phenotypes and genomic regions, but identifying causal variants remains difficult. Whole-genome sequencing (WGS) can help by providing complete knowledge of all genetic variants, but it is financially prohibitive for well-powered GWAS studies. We performed mapping of expression quantitative trait loci (eQTLs) with WGS and RNA-seq, and found that lead eQTL variants called with WGS were more likely to be causal. Through simulations, we derived properties of causal variants and used them to develop a method for identifying likely causal SNPs. We estimated that 25-70% of causal variants were located in open-chromatin regions, depending on the tissue and experiment. Finally, we identified a set of high-confidence causal variants and showed that these were more enriched in GWAS associations than other eQTLs. Of those, we found 65 associations with GWAS traits and provide examples in which genes implicated by expression are functionally validated as being relevant for complex traits.

Gene-gene and gene-environment interactions detected by transcriptome sequence analysis in twins.

Understanding the genetic architecture of gene expression is an intermediate step in understanding the genetic architecture of complex diseases. RNA sequencing technologies have improved the quantification of gene expression and allow measurement of allele-specific expression (ASE). ASE is hypothesized to result from the direct effect of cis regulatory variants, but a proper estimation of the causes of ASE has not been performed thus far. In this study, we take advantage of a sample of twins to measure the relative contributions of genetic and environmental effects to ASE, and we find substantial effects from gene × gene (G×G) and gene × environment (G×E) interactions. We propose a model where ASE requires genetic variability in cis, a difference in the sequence of both alleles, but where the magnitude of the ASE effect depends on trans genetic and environmental factors that interact with the cis genetic variants.

Genetic interactions affecting human gene expression identified by variance association mapping.

Non-additive interaction between genetic variants, or epistasis, is a possible explanation for the gap between heritability of complex traits and the variation explained by identified genetic loci. Interactions give rise to genotype dependent variance, and therefore the identification of variance quantitative trait loci can be an intermediate step to discover both epistasis and gene by environment effects (GxE). Using RNA-sequence data from lymphoblastoid cell lines (LCLs) from the TwinsUK cohort, we identify a candidate set of 508 variance associated SNPs. Exploiting the twin design we show that GxE plays a role in ∼70% of these associations. Further investigation of these loci reveals 57 epistatic interactions that replicated in a smaller dataset, explaining on average 4.3% of phenotypic variance. In 24 cases, more variance is explained by the interaction than their additive contributions. Using molecular phenotypes in this way may provide a route to uncovering genetic interactions underlying more complex traits.DOI: http://dx.doi.org/10.7554/eLife.01381.001.

CACNA1C risk variant and amygdala activity in bipolar disorder, schizophrenia and healthy controls.

OBJECTIVES: Several genetic studies have implicated the CACNA1C SNP rs1006737 in bipolar disorder (BD) and schizophrenia (SZ) pathology. This polymorphism was recently found associated with increased amygdala activity in healthy controls and patients with BD. We performed a functional Magnetic Resonance Imaging (fMRI) study in a sample of BD and SZ cases and healthy controls to test for altered amygdala activity in carriers of the rs1006737 risk allele (AA/AG), and to investigate if there were differences across the diagnostic groups. METHODS: Rs1006737 was genotyped in 250 individuals (N = 66 BD, 61 SZ and 123 healthy controls), all of Northern European origin, who underwent an fMRI negative faces matching task. Statistical tests were performed with a model correcting for sex, age, diagnostic category and medication status in the total sample, and then in each diagnostic group. RESULTS: In the total sample, carriers of the risk allele had increased activation in the left amygdala. Group-wise analyses showed that this effect was significant in the BD group, but not in the other diagnostic groups. However, there was no significant interaction effect for the risk allele between BD and the other groups. CONCLUSIONS: These results indicate that CACNA1C SNP rs1006737 affects amygdala activity during emotional processing across all diagnostic groups. The current findings add to the growing body of knowledge of the pleiotropic effect of this polymorphism, and further support that ion channel dysregulation is involved in the underlying mechanisms of BD and SZ.

Common sequence variants in the major histocompatibility complex region associate with cerebral ventricular size in schizophrenia.

BACKGROUND: Because of evidence from genetic linkage and genome-wide association studies, as well as suggested involvement of infection, the major histocompatibility complex (MHC) region on chromosome 6p21.3-22.1 has been implicated in the development of schizophrenia. METHODS: Here, we investigated how gene variants across the MHC region are associated with brain structure in a large ethnically homogenous sample (n = 420), including patients with schizophrenia spectrum disorders and other severe mental illness and healthy control subjects. RESULTS: We demonstrate highly significant associations between common gene markers in the MHC region and cerebral ventricular volume specifically in schizophrenia spectrum patients (uncorrected p values between 1.16 × 10⁻4 and 2.00 × 10⁻7). One single nucleotide polymorphism, rs2596532, survives Bonferroni correction for multiple testing across all single nucleotide polymorphisms and brain structure measures (adjusted p value 5.59 × 10⁻4). CONCLUSIONS: The results indicate that MHC variants are implicated in characteristic brain abnormalities of schizophrenia.

Candidate gene analysis of the human natural killer-1 carbohydrate pathway and perineuronal nets in schizophrenia: B3GAT2 is associated with disease risk and cortical surface area.

BACKGROUND: The Human Natural Killer-1 carbohydrate (HNK-1) is involved in neurodevelopment and synaptic plasticity. Extracellular matrix structures called perineuronal nets, condensed around subsets of neurons and proximal dendrites during brain maturation, regulate synaptic transmission and plasticity. METHODS: Ten genes of importance for HNK-1 biosynthesis (B3GAT1, B3GAT2, and CHST10) or for the formation of perineuronal nets (TNR, BCAN, NCAN, HAPLN1, HAPLN2, HAPLN3, and HAPLN4) were investigated for potential involvement in schizophrenia (SCZ) susceptibility, by genotyping 104 tagSNPs in the Scandinavian Collaboration on Psychiatric Etiology sample (849 cases; 1602 control subjects). Genome-wide association study imputation data from the European SGENE-plus sample (2663 cases; 13,498 control subjects) were used for comparison. The effect of SCZ risk alleles on brain structure was investigated in a Norwegian subset (98 cases; 177 control subjects) with structural magnetic resonance imaging data. RESULTS: Five single nucleotide polymorphisms (SNPs), located in two adjacent estimated linkage disequilibrium blocks in the first intron of ß-1,3-glucuronyltransferase 2 (B3GAT2), were nominally associated with SCZ (.004 ≤ P(empirical) ≤ .05). The rs2460691 was significantly associated in the comparison sample and in the meta-analysis after correction for all 121 SNP/haplotype tests (P(raw) = 1 × 10(-4); P(corrected) = .018). Increased dosage of the rs2460691 SCZ risk allele was associated with decreased cortical area (p = .002) but not thickness or hippocampal volume. A second SNP (r(2) = .24 with rs10945275), which conferred the highest SCZ risk effect in the Norwegian subset, was also associated with cortical area. CONCLUSIONS: The present results suggest that effects on biosynthesis of the neuronal epitope HNK-1, through common B3GAT2 variation, could increase the risk of SCZ, possibly by decreasing cortical area.