Comparison between CIRI, find_circ and segemehl based on HeLa cell transcriptome data. (A) Overlap of identified circRNAs among the three algorithms using the RNaseR- data. (B) Overlap of identified circRNAs between CIRI and find_circ using the RNaseR- data. find_circ-specific candidates can be removed by various filters in CIRI. (C) A schematic view of reads mapped to a circRNA region. RNaseR+ data were used to validate the circRNA candidates identified based on the RNaseR- data. Dashed rectangle indicates the flanking region adjacent to the junction, which was used to plot the read depth in subgraph E. (D) Two typical false positive types identified by the three algorithms. (E) Density plot of sequencing depth adjacent to the junction of circRNA candidates. The x-axis represents the relative coordinates of all circRNA candidates with the junction point set to zero; the y-axis represents normalized sequencing depth.