Dataset Information

A statistical framework for cross-tissue transcriptome-wide association analysis.

ABSTRACT: Transcriptome-wide association analysis is a powerful approach to studying the genetic architecture of complex traits. A key component of this approach is to build a model to impute gene expression levels from genotypes by using samples with matched genotypes and gene expression data in a given tissue. However, it is challenging to develop robust and accurate imputation models with a limited sample size for any single tissue. Here, we first introduce a multi-task learning method to jointly impute gene expression in 44 human tissues. Compared with single-tissue methods, our approach achieved an average of 39% improvement in imputation accuracy and generated effective imputation models for an average of 120% more genes. We describe a summary-statistic-based testing framework that combines multiple single-tissue associations into a powerful metric to quantify the overall gene-trait association. We applied our method, called UTMOST (unified test for molecular signatures), to multiple genome-wide-association results and demonstrate its advantages over single-tissue strategies.

SUBMITTER: Hu Y

PROVIDER: S-EPMC6788740 | biostudies-literature | 2019 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A statistical framework for cross-tissue transcriptome-wide association analysis.

Hu Yiming Y Li Mo M Lu Qiongshi Q Weng Haoyi H Wang Jiawei J Zekavat Seyedeh M SM Yu Zhaolong Z Li Boyang B Gu Jianlei J Muchnik Sydney S Shi Yu Y Kunkle Brian W BW Mukherjee Shubhabrata S Natarajan Pradeep P Naj Adam A Kuzma Amanda A Zhao Yi Y Crane Paul K PK Lu Hui H Zhao Hongyu H

Nature genetics 20190225 3

Transcriptome-wide association analysis is a powerful approach to studying the genetic architecture of complex traits. A key component of this approach is to build a model to impute gene expression levels from genotypes by using samples with matched genotypes and gene expression data in a given tissue. However, it is challenging to develop robust and accurate imputation models with a limited sample size for any single tissue. Here, we first introduce a multi-task learning method to jointly imput ...[more]

PMID: 30804563

Similar Datasets

Project description:The methodology of transcriptome-wide association studies (TWAS) has become popular in integrating a reference expression quantitative trait (eQTL) data set with an independent main GWAS data set to identify (putatively) causal genes, shedding mechanistic insights to biological pathways from genetic variants to a GWAS trait mediated by gene expression. Statistically TWAS is a (two-sample) 2-stage least squares (2SLS) method in the framework of instrumental variables analysis for causal inference: in Stage 1 it uses the reference eQTL data to impute a genes expression for the main GWAS data, then in Stage 2 it tests for association between the imputed gene expression and the GWAS trait; if an association is detected in Stage 2, a (putatively) causal relationship between the gene and the GWAS trait is claimed. If a nonlinear model or a generalized linear model (GLM) is fitted in Stage 2 (e.g., for a binary GWAS trait), it is known that using only imputed gene expression, as in standard TWAS, in general does not lead to a consistent (i.e., asymptotically unbiased) estimate for the causal effect; accordingly, a variation of 2SLS, called two-stage residual inclusion (2SRI), has been proposed to yield better estimates (e.g., being consistent under suitable conditions). Our main goal is to investigate whether it is necessary or even better to apply 2SRI, instead of the standard 2SLS. In addition, due to the use of imputed gene expression (i.e., with measurement errors), it is known that in general some correction to the standard error estimate of the causal effect estimate has to be applied, while in the standard TWAS no correction is applied. Is this an issue? We also compare one-sample 2SLS with two-sample 2SLS (i.e., the standard TWAS). We used the Alzheimer's Disease Neuroimaging Initiative (ADNI) data and simulated data mimicking the ADNI data to address the above questions. At the end, we conclude that, in practice with the large sample sizes and small effect sizes of genetic variants, the standard TWAS performs well and is recommended.

Project description:Background: Although genome-wide association studies (GWAS) have identified 14 loci associated with frailty index (FI) susceptibility, the underlying causative genes and biological mechanisms remain elusive. Methods: A cross-tissue transcriptome-wide association study (TWAS) was conducted utilizing the Unified Test for Molecular Markers (UTMOST), which integrates GWAS summary statistics from 164,610 individuals of European ancestry and 10,616 Swedish participants, alongside gene expression matrices from the Genotype-Tissue Expression (GTEx) Project. Validation of the significant genes was performed through three distinct methods: FUSION, FOCUS, and Multiple Marker Analysis of Genome-wide Annotation (MAGMA). Exploration of tissue and functional enrichment for FI-associated SNPs was conducted using MAGMA. Conditional and joint analyses, along with fine mapping, were employed to enhance our understanding of FI's genetic architecture. Mendelian randomization was employed to ascertain causal relationships between significant genes and FI, and co-localization analysis was utilized to investigate shared SNPs between significant genes and FI. Results: In this study, two novel susceptibility genes associated with the risk of FI were identified through the application of four TWAS methods. Mendelian randomization demonstrated that HTT may elevate the risk of developing frailty, whereas LRPPRC could offer protection against the onset of frailty. Additionally, co-localization analysis identified a shared SNP between LRPPRC and FI. Tissue enrichment analyses revealed that genomic regions linked to SNPs associated with frailty were predominantly enriched in various brain regions, including the frontal cortex, cerebral cortex, and cerebellar hemispheres. Conditional, combined analyses, and fine mapping collectively identified two genetic regions associated with frailty: 2p21 and 4q16.3. Functional enrichment analyses revealed that the pathways associated with frailty were primarily related to the MHC complex, PD-1 signaling, cognition, inflammatory response to antigenic stimuli, and the production of second messenger molecules. Conclusion: This investigation uncovers two newly identified genes with forecasted expression levels associated with the risk of FI, offering new perspectives on the genetic architecture underlying FI.

Project description:Genome-wide association studies (GWASs) have identified 38 loci associated with ulcerative colitis (UC) susceptibility, but the risk genes and their biological mechanisms remained to be comprehensively elucidated. Multi-marker analysis of genomic annotation (MAGMA) software was used to annotate genes on GWAS summary statistics of UC from FinnGen database. Genetic analysis was performed to identify risk genes. Cross-tissue transcriptome-wide association study (TWAS) using the unified test for molecular signatures (UTMOST) was performed to compare GWAS summary statistics with gene expression matrix (from Genotype-Tissue Expression Project) for data integration. Subsequently, we used FUSION software to select key genes from the individual tissues. Additionally, conditional and joint analysis was conducted to improve our understanding on UC. Fine-mapping of causal gene sets (FOCUS) software was employed to accurately locate risk genes. The results of the four genetic analyses (MAGMA, UTMOST, FUSION and FOCUS) were combined to obtain a set of UC risk genes. Finally, Mendelian randomization (MR) analysis and Bayesian colocalization analysis were conducted to determine the causal relationship between the risk genes and UC. To test the robustness of our findings, the same approaches were taken to verify the GWAS data of UC on IEU. Multiple correction tests screened PIM3 as a risk gene for UC. The results of Bayesian colocalization analysis showed that the posterior probability of hypothesis 4 was 0.997 and 0.954 in the validation dataset. MR was conducted using the inverse variance weighting method and two single nucleotide polymorphisms (SNPs, rs28645887 and rs62231924) were included in the analysis (p < 0.001, 95%CI: 1.45-1.89). In the validation dataset, MR result was p < 0.001, 95%CI: 1.19-1.72, indicating a clear causal relationship between PIM3 and UC. Our study validated PIM3 as a key risk gene for UC and its expression level may be related to the risk of UC, providing a novel reference for further improving the current understanding on the genetic structure of UC.

Dataset Information

A statistical framework for cross-tissue transcriptome-wide association analysis.

Publications

A statistical framework for cross-tissue transcriptome-wide association analysis.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets