Validation of FBL and FFL motif predictions. 108 published TF→miRNA regulations (Human) were identified from Qiu and colleagues  for the 19 TFs in common with MIR@NT@N database. FBL (A) and FFL motifs B) were identified from the input list of miRNAs included in Qiu database, and by filtering actors with the 19 common TFs and with the 108 published TF→miRNA regulations. Non stringent criteria were used for TF regulation predictions (TFBS score ≥ 0.65, TFBS length ≥ 6, number of miRNA per TF ≥ 1, genes targeted by at least 1 TF), and to filter miRNA regulation predictions we used criteria corresponding to a unified score ≥ 0.8 (miRBase score ≥ 17 and p-value ≤ 0.01, and MicroRNA.org score ≥ 152), and 1 gene per miRNA. Squares represent TFs, diamonds represent miRNAs, and edges represent predicted regulations (blue scale color code used for the prediction score). For FBL sub-network, edges represent double regulations.
Le Béchec et al. BMC Bioinformatics 2011 12:67 doi:10.1186/1471-2105-12-67