Skip to main content
. 2021 Jul 12;37(Suppl 1):i308–i316. doi: 10.1093/bioinformatics/btab300

Fig. 4.

Fig. 4.

Bipartite motif models learned on in vitro data match their in vivo counterparts. Bipartite motifs are shown for those RBPs in Figure 3 whose best replicate has a tool-averaged AUROC of at least 0.75. The models learned in vitro and in vivo match not only in the sequence preference but also the relative positioning of the two motif cores, with the exception of KHDRBS1, which shows a bipartite motif only in the HTR-SELEX data