Figure 5

Overview of the workflow inside Mod_Parallel. Relying on four inputs (miR2gene, TF2gene, TF2miRNA, and a miR/mRNA parallel expression dataset), Mod_Parallel decides significant miRNA-gene and TF-gene regulations from mRNA expression dataset and significant TF-miRNA regulations from miRNA expression dataset through two multi-variate linear regression models. The reserved regulations of three different types are combined to form a comprehensive TF-and-miRNA-involving combinatorial gene regulatory network. Topological analyses of the combinatorial network are further carried out which yield single regulators, regulator pairs, and triangle regulation motifs of remakable notice. In the formula, E _g and E _m stand for the mRNA level of a target protein-coding gene g and a miRNA m respectively; E _tfg , E _mg , and E _tfm are vector-form notations of the transcript profiles of g's regulating TFs, g's regulating miRNAs, and m's regulating TFs; E _tfg , E _mg , and E _tfm quantify the to-be-estimated regulating coefficients (in vector forms) of the same three sets of regulators.