Figure 1

Statistics of four integrated disease databases (i.e., OMIM, DO, HPO, and GAD) and the overall schema of three representative quantities to identify disease comorbidity.

(A) Disease overlap among four disease databases. The number in parentheses represents the total number of genes in each database. (B) Disease gene coverage of the integrated disease database in comparison with STRING network. The x-axis represents the proportion of overlap between associated genes of a disease and all genes in the STRING network. The y-axis indicates the fraction of diseases. The fraction of diseases (more than 80% of disease genes are covered by STRING) is over 95%. (C) Two different strategies to represent the degree of comorbidity between diseases A and B. “Direct gene overlap” and “Function network structure” are used to consider overlap between associated genes of the two diseases and the number of direct as well as indirect interactions between associated genes of the two diseases in a function network, respectively. The “Function network structure” strategy to explain the disease comorbidity utilizes disease-associated genes as well as the neighborhood genes which are connected to the disease-associated genes. In our study, the STRING interaction database has been used to identify the functional interactions.