Figure 5.

Workflow of data collection, miRNA-TF synergistic regulatory network construction, and motif identification. A: Lung cancer-related gene collection. Lung cancer genes were obtained from five databases, differential expressed genes in two microarray datasets were calculated, and the overlapped set were used in the present study. B, C, and D showed the data source or algorithms of regulator-target relations. E: Workflow of synergistic regulatory network construction and motif identification. Elements with a star mark '*’ are 'lung cancer related-’. First, by Gene^* from step A and miRNA target relations from step B, we obtained miRNA^* using a hypergeometric test. After a similar procedure for getting TF^*, we combined any two of Gene^*, miRNA^*, and TF^*with their regulatory relations to obtain four types of regulatory relation^*. Then, we merged them to construct the miRNA-TF synergistic regulatory network^*. Last, 10 motif types were identified.