Skip to main content

Table 12 Null hypothesis probabilities for differences between Phase4 targets and all targets in disease association counts (abbreviated D.A.s) and related ‘hybrid’ graph-D.A.s parameters

From: Centrality of drug targets in protein networks

  1. Disease association counts were extrapolated from DisGeNET or Genetic Association Database. Additional parameters combining the percentile ranking of disease associations with network centrality metrics were evaluated: RNs score (reported in Ref. [50]); ‘R top’ and ‘R top-degree’ each combining the disease associations percentile (R) with one centrality parameter identified here as discriminating between ‘Phase4’ and ‘all targets’ protein sets: topological coefficient and log(degree*topological coefficient) (see Table 5), respectively. This analysis was performed on the String07 and Reactome networks, corrected for multiple testing across both centrality metric and target classes (Benjamini–Hochberg method). Cells are colored in increasingly darker shades of green according to statistical significance (N.S.: not significant)