Fig. 5

Performance evaluation of RCRUNCH variants. “RCRUNCH standard” refers to peaks identified based on reads that map uniquely to the genome, and binding sites that are always centered on the most frequent crosslink position in each peak. Rows correspond to variants of the RCRUNCH workflow, while columns show different metrics used to compare each of the variants with the “standard” RCRUNCH: a–c Number of significant peaks in a sample (at FDR < 0.1). d–f Enrichment of the known motif of the RBP that was assayed in a particular sample. g–i Similarity of the motif identified de novo from the top peaks of a given sample and the known motif for the respective RBP. In the case of the RNA-seq background variant, some samples did not yield any significant peaks (FDR = 0.1). These samples are therefore not represented in the plots