For a more detailed validation study using ToppGene, the readers are referred to our previous study (10). In the present study, to demonstrate the utility of ToppGene Suite, we focused on recently reported GWAS. The aim was to test whether ToppGene and ToppNet are capable of retrieving or prioritizing the GWAS-discovered novel disease genes in a training-test type of analysis. We used 20 gene–disease associations (including novel disease genes) representing five diseases (Bipolar Disorder, Cardiomyopathy, Celiac Disease, Crohns Disease and Obesity; Table 2). For each of these five disorders, we built a training set containing all the genes already known to play a role in that disorder according to the OMIM and NCBI's Entrez Gene records (limiting the search field to ‘Disease/Phenotype’ and organism ‘Homo sapiens’) (See ‘Supplementary’ section from ToppGene homepage). The test set consisted of the GWAS-reported disease gene plus 99 nearest neighboring genes based on their location on the same chromosome. ToppGene and ToppNet prioritization results are presented in Table 2. ToppGene ranked 19 of 20 (95%) candidate genes within the top 20%, while ToppNet ranked 12
