C-JAMP

C-JAMP performs copula-based joint rare-variant association testing of multiple phenotypes to model trait dependencies and improve power and control of type I error in genetic association studies.

Key Features:

• Joint Modeling: Simultaneously analyzes multiple phenotypes together with genetic markers and other covariates in a joint model.
• Copula-Based Approach: Uses copula functions to flexibly model the dependency structure between different phenotypic traits.
• Empirical Validation: Demonstrated by extensive simulation studies to produce unbiased genetic effect estimates, maintain controlled type I error, and benchmarked against univariate and multivariate single-marker and multi-marker rare-variant tests.
• Adaptability to Trait Dependence: Shows consistently high power for strongly dependent traits, with relative performance under weak or moderate dependence depending on variant effect sizes.
• Robustness to Model Misspecification: Retains high power in several scenarios even when the copula function is misspecified.

Scientific Applications:

• GWAS of adiponectin: Applied to a genome-wide association study of adiponectin plasma concentrations, identifying 20 rare variants with p-values smaller than 10^-5.
• Complex trait marker discovery: Used to detect candidate rare-variant associations in complex genetic trait studies involving multiple correlated phenotypes.

Methodology:

C-JAMP implements a copula-based joint model integrating multiple phenotypic measurements with genetic markers and other covariates, adjusts test statistics to ensure valid type I error, and has been validated through simulation studies and benchmarking; the method is provided as an R package and reported to be computationally efficient.