Workflow for causal gene prediction. The schema illustrates the workflow to predict causal genes for a disease of interest. In step 1, similar diseases were gathered on the basis of the disease similarity threshold (similarity P-value and overlap threshold). For example, if obesity were the disease of interest and a P-value of 0.01 as well as a minimal overlap of two causal genes were applied as similarity cut-off, this step would select the vicinal disorders of obesity described in Table 1. Step 2 only considered genes of similar diseases which satisfied the similarity threshold with known causal genes of the disease of interest. These genes were eventually predicted as causal genes of the disease of interest (Step 3).