Skip to main content
Figure 1 | BMC Genomics

Figure 1

From: Characterising ChIP-seq binding patterns by model-based peak shape deconvolution

Figure 1

MeDiChISeq performance evaluated in the context of several ChIP-seq datasets and relevant Peak calling algorithms. (A) MeDiChISeq and three other peak callers (MACS, BayesPeak and PeakRanger) were used to identify binding events in ChIP-seq datasets for three TFs (SRF, MAX, NRSF), the sequence-specific insulator protein CTCF and two histone modification marks (H3K9Ac, H3K27Ac, H3K4me3). The default confidence threshold parameters described for each peak caller were applied to assess the number of peaks per dataset. Note that for each ChIP-seq two biological replicates were processed. (B) Peaks commonly identified by two of the indicated peak callers for two replicates of CTCF (top panel) and H3K4me3 (bottom panel) are displayed as percentages of the total sites found by a given method (indicated at the right). (C) Representative genome browser screenshots illustrating the ability to deconvolve binding/modification patterns of peak callers. Note that most of the peak callers identify a similar number of “sharp” binding events for CTCF, while MeDiChISeq has the highest potential of deconvolution for the H3K4me3 pattern.

Back to article page