Genome-wide association studies (GWAS) have greatly expanded our knowledge of common diseases by discovering many susceptibility common variants. So far, analysis of genetic signal has been typically based on the single marker/gene approach. Gene set based analysis can be an alternative and effective approach to detecting the combined effect of multiple variants within a gene set from GWAS dataset(s).

dmGWAS is designed to identify significant protein-protein interaction (PPI) modules and, from which, the candidate genes for complex diseases by an integrative analysis of GWAS dataset(s) and PPI network. It implements a dense module searching method previously developed for gene expression data analysis (Ideker et al., 2002). We adapted the method specifically for GWAS datasets, including data preparation, integration, searching, and validation in GWAS permutation data. Specifically, we proposed two strategies to select modules for single GWAS and multiple GWAS datasets. In the later case, additional GWAS dataset(s) can be used for evaluation/validation of the modules identified by the primary (discovery) GWAS dataset.