CMM

CMM performs joint genetic analysis of independently collected GWAS datasets with different phenotypes by using individual-level data and multivariate sparse mixed models to infer unknown phenotypes and control confounding factors.

Key Features:

• Joint analysis capability: Simultaneous analysis of two independently collected GWAS datasets, enabling inference of unknown phenotypes across datasets.
• Multivariate sparse mixed models: Implements multivariate sparse mixed models to model genetic effects jointly across datasets.
• Individual-level data analysis: Operates on individual-level GWAS data rather than relying on summary statistics.
• Confounding factor management: Accounts for population stratification, family structures, cryptic relatedness, and batch effects.
• Evaluation on simulated and real data: Performance validated via simulation experiments and applied to real GWAS datasets including Alzheimer's disease and substance use disorder.

Scientific Applications:

• Cross-phenotype association discovery: Identify shared or pleiotropic genetic associations across different phenotypes by joint analysis of GWAS datasets.
• Complex disorder genetics: Investigate common genetic associations in complex disorders, exemplified by analyses of Alzheimer's disease and substance use disorder.

Methodology:

Uses multivariate sparse mixed models to jointly analyze two GWAS datasets, infers unknown phenotypes, and incorporates controls for population stratification, family structure, cryptic relatedness, and batch effects; evaluated with simulation experiments and real-data analyses of Alzheimer's disease and substance use disorder.