Co-expression-wide association studies implicate protein–protein interactions in complex disease risk DOI Creative Commons
Mykhaylo M. Malakhov, Wei Pan

medRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 4, 2024

Abstract Transcriptome- and proteome-wide association studies (TWAS/PWAS) have proven successful in prioritizing genes proteins whose genetically regulated expression modulates disease risk, but they ignore potential co-expression interaction effects. To address this limitation, we introduce the co-expression-wide study (COWAS) method, which can identify pairs of or is associated with complex traits. COWAS first trains models to predict conditional on genetic variation, then tests for between imputed trait interest while also accounting direct effects from each exposure. We applied our method plasma proteomic concentrations UK Biobank, identifying dozens interacting protein cholesterol levels, Alzheimer’s disease, Parkinson’s disease. Notably, results demonstrate that may affect traits even if neither detected influence when considered its own. show how help disentangle effects, providing a richer picture molecular networks mediate outcomes.

Language: Английский

A Multi-omic study integrating plasma protein, multiple tissues, and single-cell identifies RNASET2 as a key gene for lung cancer DOI Creative Commons
Jiaxin Shi, Linyou Zhang

Discover Oncology, Journal Year: 2025, Volume and Issue: 16(1)

Published: Feb. 10, 2025

Lung cancer (LC) has the highest cancer-related mortality rate. Even though genome-wide association studies (GWAS) have identified numerous loci linked to LC risk, underlying causal genes and biological processes are still mostly unknown. The GWAS summary data comprised 29,863 cases 55,586 controls of European ancestry. weight file related files plasma protein, multi-tissue, single-cell were obtained from Zhang's study, Mancuso lab, Thompson's respectively. We conducted transcriptome (TWAS) employing functional Summary-based Imputation (FUSION) two levels, which multiple tissues single cell. proteome-wide (PWAS) protein. Conditional joint (COJO) analysis multi-marker genomic annotation (MAGMA) used further screen PWAS/TWAS results. Summary-data-based Mendelian randomization (SMR) colocalization utilized explain between variables A total 13, 251, 16 calculated three dimensions, tissues, cell, RNASET2 IREB2 through intersecting these sets genes. COJO MAGMA replicated successfully. Then, was in both eQTL-SMR mQTL-SMR following analysis. In summary, we a multi-omic studies, integrated levels investigate novel targets for LC. Through series verifications, as key gene current research.

Language: Английский

Citations

0
Enhancing nonlinear transcriptome- and proteome-wide association studies via trait imputation with applications to Alzheimer’s disease DOI Creative Commons
Ruoyu He,

Jingchen Ren,

Mykhaylo M. Malakhov

et al.

PLoS Genetics, Journal Year: 2025, Volume and Issue: 21(4), P. e1011659 - e1011659

Published: April 10, 2025

Genome-wide association studies (GWAS) performed on large cohort and biobank datasets have identified many genetic loci associated with Alzheimer’s disease (AD). However, the younger demographic of participants relative to typical age late-onset AD has resulted in an insufficient number cases, limiting statistical power GWAS any downstream analyses. To mitigate this limitation, several trait imputation methods been proposed impute expected future status individuals who may not yet developed disease. This paper explores use imputed nonlinear transcriptome/proteome-wide (TWAS/PWAS) identify genes proteins whose genetically regulated expression is risk. In particular, we considered TWAS/PWAS method DeLIVR, which utilizes deep learning model effects We trained transcriptome proteome models for DeLIVR data from Genotype-Tissue Expression (GTEx) Project UK Biobank (UKB), respectively, UKB as outcome. Next, hypothesis testing using clinically diagnosed cases Disease Sequencing (ADSP). Our results demonstrate that outcomes successfully identifies known putative risk proteins. Notably, found training can increase without inflating false positives, enabling discovery molecular exposures potentially neurodegeneration.

Language: Английский

Citations

0
Multimodal analysis of RNA sequencing data powers discovery of complex trait genetics DOI Creative Commons
Daniel Munro, Nava Ehsan, Seyed Mehdi Esmaeili-Fard

et al.

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: Nov. 29, 2024

Abstract RNA sequencing has the potential to reveal many modalities of transcriptional regulation, such as various splicing phenotypes, but studies on gene regulation are often limited expression due complexity extracting and analyzing multiple phenotypes. Here, we present Pantry, a framework efficiently generate diverse phenotypes from data perform downstream integrative analyses with genetic data. Pantry generates six (gene expression, isoform ratios, splice junction usage, alternative TSS/polyA stability) integrates them via QTL mapping, TWAS, colocalization testing. We apply Geuvadis GTEx data, finding that 4768 genes no identified eQTL in have at least one other modality, resulting 66% increase over mapping. further found exhibit modality-specific functional properties reinforced by joint analysis different modalities. also show generalizing TWAS approximately doubles discovery unique gene-trait associations, enhances identification regulatory mechanisms underlying GWAS signal 42% previously associated pairs.

Language: Английский

Citations

1
Multimodal analysis of RNA sequencing data powers discovery of complex trait genetics DOI Creative Commons
Daniel Munro, Nava Ehsan, Seyed Mehdi Esmaeili-Fard

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: May 15, 2024

Transcriptome data is commonly used to understand genome function via quantitative trait loci (QTL) mapping and identify the molecular mechanisms driving wide association study (GWAS) signals through colocalization analysis transcriptome-wide studies (TWAS). While RNA sequencing (RNA-seq) has potential reveal many modalities of transcriptional regulation, such as various splicing phenotypes, are often limited gene expression due complexity extracting analyzing multiple phenotypes. Here, we present Pantry (Pan-transcriptomic phenotyping), a framework efficiently generate diverse phenotypes from RNA-seq perform downstream integrative analyses with genetic data. currently generates six regulation (gene expression, isoform ratios, splice junction usage, alternative TSS/polyA stability) integrates them QTL mapping, TWAS, testing. We applied Geuvadis GTEx data, found that 4,768 genes no identified in had QTLs at least one other modality, resulting 66% increase over mapping. further exhibit modality-specific functional properties reinforced by joint different modalities. also show generalizing TWAS (xTWAS) approximately doubles discovery unique gene-trait associations, enhances identification regulatory underlying GWAS signal 42% previously associated pairs. provide code, all samples, xQTL xTWAS results on web.

Language: Английский

Citations

0
Co-expression-wide association studies implicate protein–protein interactions in complex disease risk DOI Creative Commons
Mykhaylo M. Malakhov, Wei Pan

medRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 4, 2024

Abstract Transcriptome- and proteome-wide association studies (TWAS/PWAS) have proven successful in prioritizing genes proteins whose genetically regulated expression modulates disease risk, but they ignore potential co-expression interaction effects. To address this limitation, we introduce the co-expression-wide study (COWAS) method, which can identify pairs of or is associated with complex traits. COWAS first trains models to predict conditional on genetic variation, then tests for between imputed trait interest while also accounting direct effects from each exposure. We applied our method plasma proteomic concentrations UK Biobank, identifying dozens interacting protein cholesterol levels, Alzheimer’s disease, Parkinson’s disease. Notably, results demonstrate that may affect traits even if neither detected influence when considered its own. show how help disentangle effects, providing a richer picture molecular networks mediate outcomes.

Language: Английский

Citations

0