LPM

LPM integrates summary statistics from multiple GWASs with functional annotations to elucidate the genetic architecture of complex traits and diseases.

Key Features:

• Integration of summary-level GWAS data and functional annotations: Combines summary statistics from multiple GWASs with genic category and cell-type specific functional annotations to link regulatory variants to phenotypes.
• Modeling pleiotropy and regulatory effects: Explicitly addresses pleiotropy and the contribution of regulatory variants to multiple phenotypes.
• Scalability and statistical robustness: Employs a computational framework capable of handling hundreds of annotations and phenotypes while maintaining statistical accuracy.
• Model parameter estimation and statistical inference: Provides procedures for estimating latent probit model parameters and conducting statistical inference for relationships among traits and variant prioritization.
• Simulation studies and comparative analysis: Uses extensive simulation studies and comparisons with related methods to evaluate performance and accuracy.
• Real-world dataset application: Applied to 44 GWASs incorporating 9 genic category annotations and 127 cell-type specific functional annotations.

Scientific Applications:

• Phenotype relationship characterization: Characterizes relationships among multiple phenotypes to reveal shared genetic architecture.
• Risk variant prioritization: Prioritizes risk variants by integrating regulatory information from functional annotations.
• Unified analysis of diverse genomic data: Provides a unified statistical approach to integrate and analyze diverse GWAS summary statistics and functional annotation resources.

Methodology:

Implements a latent probit model that integrates summary-level GWAS data with functional annotations, performs model parameter estimation and statistical inference, and evaluates performance via simulation studies and comparative analyses; the computational framework is scalable to hundreds of annotations and phenotypes.