graph-GPA

graph-GPA integrates multiple genome-wide association study (GWAS) datasets using a hidden Markov random field to leverage pleiotropy for improved detection of genetic variants associated with complex traits.

Key Features:

• Integration of Multiple GWAS Datasets: Implements a hidden Markov random field approach to integrate an extensive number of GWAS datasets across multiple phenotypes.
• Enhanced Statistical Power: Performs joint analyses across multiple datasets within a statistical framework that leverages pleiotropic relationships to increase power to identify risk variants.
• Phenotype Graph Generation: Generates phenotype graphs that represent shared genetic architecture among different traits.
• Application to Diverse Phenotypes: Has been applied to joint analyses of GWAS datasets for 12 distinct phenotypes.

Scientific Applications:

• Novel Biomarker Discovery: Identifies shared risk variants across traits that can inform discovery of biomarkers for diagnosis and personalized medicine.
• Therapeutic Target Identification: Elucidates pleiotropic effects to help pinpoint potential therapeutic targets.
• Genetic Mechanism Elucidation: Characterizes shared genetic architecture to advance understanding of genetic mechanisms underlying complex diseases.

Methodology:

Hidden Markov random field modeling of GWAS datasets; joint statistical analysis across multiple phenotypes leveraging pleiotropy; generation of phenotype graphs to represent shared genetic architecture.