Fig. 3

a Global view of the transcription factors (TFs) and their target functions; 311 TFs and 1420 annotations with one or more significant associations at FDR 0.1 levels were retained. Red indicates positive associations, green indicates negative associations, white indicates FDR > 0.1. Intensity of the colors corresponds to the significance levels: FDR 0.1, 0.05, and 0.01. The TF and target function clustering showed on the left and top was performed based on the TF-target function association phi coefficient matrix. We used the literature rich gene universe for the association analysis except for the TF-GWAS phenotype association, for which the coding gene universe was used. b The network visualization [148] of TF–target function and TF–known function relationships. Edges are colored red or green the same way as in (a). A solid edge links a TF with a significant target function that is not a known function. A dashed edge links a TF with a known function. A dashed edge with color links a TF with a known function that is also a significant target function, while a grey dashed edge links a TF with a known function that is not a significant target function. Node colors and shapes correspond to function types – purple circles, TFs; grey rectangles, Reactome pathways; blue triangles, GO molecular functions; white diamonds, GO biological processes; red rhomboids, PharmGKB PK and PD pathways; yellow hexagon, Mendelian diseases; green octagons, GWAS phenotypes. c–g Local regions of TF-function networks selected from b