Figure 6
From: GraphProt: modeling binding preferences of RNA-binding proteins
The difference in predictive power using RNA structure in comparison to sequence-only models. Full sequence-andstructure models (y-axis) and sequence-only (x-axis) models were trained on RNAcompete (A) and CLIP-seq data (B). The gray ribbons denote the standard deviation of the differences between full structure and sequence-only models. APR, average precision; AUROC, area under the receiver operating characteristic curve; CLIP-seq, cross-linking and immunoprecipitation sequencing.