A versatile data repository for GWAS summary statistics-based downstream genomic analysis of human complex traits

Genome-wide association studies (GWAS) have identified thousands of loci associated with complex traits and diseases, yet the biological interpretation of these findings remains limited. We developed gact, an R package that integrates GWAS summary statistics with diverse genomic resources to facilitate the discovery of causal genes, pathways, and disease mechanisms. The package enables construction of a local database linking variants to genes, biological pathways, protein complexes, and drug-gene interactions, thereby supporting downstream analyses such as fine-mapping, polygenic scoring, and gene set enrichment. Applying gact to large-scale GWAS of coronary artery disease (CAD) and type 2 diabetes (T2D), we identified 142 and 577 significant genes, respectively, including canonical loci for T2D (PATJ, DEAF1) and CAD (OLIG1), as well as pleiotropic genes such as TCF7 and HNF1B. Bayesian gene set analyses revealed distinct biological signatures-lipid and vascular remodeling pathways in CAD versus beta-cell and islet biology in T2D-together with shared enrichment in extracellular matrix and immune signaling. Polygenic score (PGS) analyses demonstrated higher predictive accuracy for CAD than T2D, consistent with differences in common-variant heritability and GWAS power. Partitioned PGS further delineated T2D subgroups through archetypal clustering, separating individuals with predominantly inflammatory versus metabolic risk profiles. These results establish gact as a versatile platform for integrating genomic resources and advancing the biological interpretation of GWAS. By linking genetic associations to biological pathways and subtypes, gact enables a deeper understanding of disease heterogeneity and informs future precision medicine strategies.