Fig. 2

Prediction of UMs. A Relationship between the number of predicted motifs in a dataset and the size (the number of binding peaks in the dataset). The datasets are sorted in the ascending order of their sizes. B Distribution of cooccurrence scores (\({S}_{c}\)) of motif pairs found in each dataset. The dotted vertical line indicates the cutoff value of \({S}_{c}\) for predicting cooccurring pairs (CPs). C Number of putative binding sites in each of the UMs sorted in the ascending order. D Distribution of the lengths of the UMs and known motifs in the HOCOMOCO and JASPAR databases. E The motif similarity graph (upper panel) and the logo (bottom panel) of UM41 containing 11,799 member motifs. In the graph, the nodes (colored in blue, each representing a member motif) are arranged on the rim of the ova, and two member motifs are connected by an edge (colored in green) with SPIC score > 0.8. F Logos of six examples of highly similar member motifs of UM41. G UM41 matches known motifs of five TFs of the JUN-related family. H Heatmap of the cooccurrence/interaction networks of the 238 UMs, names of most UMs are omitted for clarity. I A blowup view of the indicated cluster in H, formed by 14 UMs, of which UM116, UM14, UM26, UM28, UM29, UM32, UM45, UM53, UM55, and UM57 match known motifs (see main text)