Figure 4.

Quagga, Stripenn, and Zebra were compared for their ability to reasonably call CTCF-anchored stripe based on stripe length, CTCF/RAD21 enrichment, and their proportion of off-target stripes found. (A) Distribution of CTCF-stripe lengths from stripes detected by Quagga, Zebra, and Stripenn. (B) Histogram depicting the frequency of stripe occurrence depending on the length of the found stripe for CTCF-anchored stripes. (C) Breakdown of our analysis for stripes belonging to CTCF-stripes, to stripes intersecting enhancer–promoter interacting annotated regions but not CTCF-occupied motifs (EP-stripes), and stripes that do not intersect with CTCF motifs or EP annotated regions (indeterminate-stripes). Indeterminate stripes may be false positives. Quagga best minimizes indeterminate stripes.

159f04