Search

Methods for direct imputation of summary statistics, e.g. our group’s DISTMIX tool, were shown to be practically as accurate as genotype imputation method, while incurring orders of magnitude lower computational burden. Given that such imputation needs a precise estimation of linkage disequilibrium (LD) for mixed ethnicity (cosmopolitan) cohorts, there is a great need i) for much larger and diverse panels and ii) to estimate the ethnic composition of the cohort, e.g. the weights for subpopulations in the diverse panel. Unfortunately, DISTMIX and its main competitors are largely using a very small reference panel of ~2,500 subject coming from the 1000 Genome (1KG) Project. DISTMIX computed the ethnic weights of a cohort based on in-cohort allele frequency (AF) estimates. Unfortunately, due to privacy issues, most genome wide association studies (GWAS) largely stopped providing cohort AFs. Thus, to accurately estimate the LD needed for an exhaustive analysis of cosmopolitan cohorts, we propose DISTMIX2. When compared to DISTMIX and its competitors, the proposed method adds a i) much larger and diverse reference panel and ii) novel estimation for weights of ethnic mixture based solely on Z -scores (when AFs not available). To build a larger and more diverse reference panel, we use the publicly and privately available data to obtain a 33,000 (33K) panel which includes ~11K Han Chinese. The proposed method of estimating ethnic weights adequately controls the Type I error rates, especially when the subpopulations in the study are well represented in the reference panel. However, using naive pre-estimated weights incurs a much higher false positive rate. We apply DISTMIX2 to the GWAS summary statistics from the Psychiatric Genetic Consortium (PGC). Our method which uncover signals in numerous new regions, with most of these findings coming from the rarer variants. Author summary By predicting summary statistics at unmeasured genetic variants, direct imputation is a promising method for enhancing the resolution of genetic studies. However, for a better prediction of statistics at unmeasured variants, there is a need to address two urgent issues. First, there is a need for very large and diverse reference panels that greatly improve on the mostly European ones having ~2,500 (2.5K) subjects. We address this shortcoming by building a large and diverse reference panel (33K subjects, ~20K Europeans and ~11K Asians). Second, there is a need to estimate the ethnic composition of the study cohort, even when they do not report in-cohort allele frequency for genetic variants. We solve this issue by using a novel method that uses only Z -scores, which are easily computed from reported summary statistics. Our method that implements the two above solutions i) adequately controls the false positive rate and ii) provides much improved resolution when compared to methods based on older reference panels. Practical application to reported summary statistics from studies of psychiatric disorders greatly increase the number of regions harboring signals. Most of these findings are associated with rarer variants that could not be robustly assessed using smaller panels.

Background: Minimal phenotyping refers to the reliance on self-reported responses to one or two questions for disease case identification instead of full diagnostic criteria. This strategy has been applied in genome-wide association studies (GWAS) on major depressive disorder (MDD), leading to lowering of phenotyping costs, increasing sample sizes, and more GWAS hits. It assumes that any increase in diagnostic noise, and its impact on the nature of GWAS loci thus identified, can be mitigated by the large increase in sample size. Methods: We assess the impact of using different definitions of MDD in 337,198 White-British, unrelated samples in the UKBiobank with GWAS, analyses of heritability and genetic correlation, and comparison with previously published statistics on MDD and other psychiatric conditions. Definitions of MDD include seeking medical help for anxiety or depression, reporting MDD or its cardinal symptoms, and meeting full DSM criteria for MDD. Findings: Heritability of depression defined by minimal phenotyping (

Genetic signal detection in genome-wide association studies (GWAS) can be enhanced by pooling small signals from multiple Single Nucleotide Polymorphism (SNP), e.g. across genes and pathways. Because genes are believed to influence traits via gene expression, it is of interest to combine information from expression Quantitative Trait Loci (eQTLs) in a gene or genes in the same pathway. Such methods, widely referred as transcriptomic, already exists for gene analysis, e.g. our group’s J oint E ffect on P henotype of e QTLs associated with a G ene in Mix ed cohorts (JEPEGMIX/JEPEGMIX2). However, due to the its quadratic (in the number of SNPs) computational burden for computing linkage disequilibrium (LD) across large regions, transcriptomic methods are not yet available for arbitrarily large pathways/gene sets. To overcome this obstacle, we propose JEPEGMIX2-pathways (JEPEGMIX-P), which implements a novel transcriptomic pathway method having a desirable linear computational burden. It 1) automatically estimates the ethnic composition (weights) of the cohort using a very large and diverse reference panel (33K subjects, including ∼11K Han Chinese), 2) uses these weights and the reference panel to estimate the LD between gene transcriptomic statistics and 3) uses the estimated LD values along with GWAS summary statistics to rapidly test for the association between trait and the expression of genes even in the largest pathways. To underline its potential for increasing the power to uncover genetic signals over the state-of-the-art and commonly used non-transcriptomics (agnostic) methods, e.g. MAGMA, we applied JEPEGMIX2-P to summary statistics of several large meta-analyses from Psychiatric Genetics Consortium (PGC). Surprisingly, most of these significant pathways do not seem to be directly involved in the activity of the central nervous system. While our work is just the first step on the road toward the end goal of clinical translation, PGC anorexia results suggest possible avenues for (personalized) treatment. Author summary By using summary statistics from genetic studies to infer the association between the biologically relevant measure of gene expression and traits, transcriptomics methods are a promising avenue for uncovering risk genes and pathway of genes for complex human diseases. While numerous such transcriptomic methods were used to uncover a large number of gene level signals, due to the extreme computational burden, none of these methods was successfully extended for detecting signals at the, probably even more biologically relevant, pathway of genes level. In this paper we propose a novel transcriptomic pathway method that has a close to minimally attainable computation burden and is applicable “as-is” to ethnically diverse studies. The proposed method adequately controls the false positive rates. Its application to psychiatric disorder studies unveils numerous new signals that were not detected by state-of-the art non-transcriptomic (agnostic) methods.

Rapid technological advances, such as the development of next-generation sequencing, are driving the need for comprehensive computational tools to analyse the wealth of generated genomic data. Systematic benchmarking has been successful in many research disciplines to help non-computational researchers evaluate the accuracy and applicability of new computational tools. Adopting a standardized benchmarking practice and following established principles for the design of new benchmarking studies could help researchers using omics data to better leverage technological innovation. In this Review, we propose step-by-step instructions for increasing reusability, transparency, and reproducibility of benchmarking studies by using containerization, common data representation, open data, and systematic parameter description.

SummaryControl over the number of mtDNA molecules per cell appears to be tightly regulated, but the mechanisms involved are largely unknown. Reversible alterations in the amount of mtDNA occur in response to stress suggesting that control over the amount of mtDNA is involved in stress-related diseases including major depressive disorder (MDD). Using low-coverage sequence data from 10,442 Chinese women to compute the normalized numbers of reads mapping to the mitochondrial genome as a proxy for the amount of mtDNA, we identified two loci that contribute to mtDNA levels: one within the TFAM gene on chromosome 10 (rs11006126, p value = 8.73 × 10−28, variance explained = 1.90%) and one over the CDK6 gene on chromosome 7 (rs445, p value = 6.03 × 10−16, variance explained = 0.50%). Both loci replicated in an independent cohort. CDK6 is thus a new molecule involved in the control of mtDNA. We identify increased rates of heteroplasmy in women with MDD, and show from an experimental paradigm using mice that the increase is likely due to stress. Furthermore, at least one heteroplasmic variant is significantly associated with changes in the amount of mtDNA (position 513, p value = 3.27 × 10−9, variance explained = 0.48%) suggesting site-specific heteroplasmy as a possible link between stress and increase in amount of mtDNA. These findings indicate the involvement of mitochondrial genome copy number and sequence in an organism’s response to stress.

The China, Oxford and Virginia Commonwealth University Experimental Research on Genetic Epidemiology (CONVERGE) project on Major Depressive Disorder (MDD) sequenced 11,670 female Han Chinese at low-coverage (1.7X), providing the first large-scale whole genome sequencing resource representative of the largest ethnic group in the world. Samples are collected from 58 hospitals from 23 provinces around China. We are able to call 22 million high quality single nucleotide polymorphisms (SNP) from the nuclear genome, representing the largest SNP call set from an East Asian population to date. We use these variants for imputation of genotypes across all samples, and this has allowed us to perform a successful genome wide association study (GWAS) on MDD. The utility of these data can be extended to studies of genetic ancestry in the Han Chinese and evolutionary genetics when integrated with data from other populations. Molecular phenotypes, such as copy number variations and structural variations can be detected, quantified and analysed in similar ways.