CoMM-S

CoMM-S integrates GWAS summary statistics with eQTL data using a probabilistic collaborative mixed-model framework to infer how genetic variants influence complex traits through gene expression.

Key Features:

• Integration of GWAS Summary Statistics: Uses GWAS summary statistics instead of individual-level GWAS data to perform association analyses with eQTL information.
• Probabilistic Modeling Approach: Employs a probabilistic collaborative mixed-model that models the relationship between gene expression and genotypes using eQTL data and links phenotypic traits to predicted gene expressions from the expression model.
• Modeling of Gene Expression Regulation: Explicitly models expression quantitative trait loci (eQTL) effects to connect regulatory variation to downstream phenotypic variation.
• Statistical Efficiency: Demonstrates statistical efficiency and robustness comparable to the original CoMM based on simulation studies and real-data analyses.

Scientific Applications:

• Mechanistic dissection of genetic associations: Investigating mechanistic links between genetic variants and complex traits when individual-level GWAS data are unavailable.
• Role of gene expression regulation: Assessing the contribution of eQTL-mediated gene expression to phenotypic variation and complex diseases.

Methodology:

Combines eQTL data with GWAS summary statistics and applies a probabilistic collaborative mixed-model framework that (i) models gene expression–genotype relationships using eQTL data and (ii) links phenotypic traits to predicted gene expressions, with probabilistic inference to account for uncertainty in summary-level data.