Figure 5 : Characterization of interaction hotspots in promoter–promoter interactions.

From: Constructing 3D interaction maps from 1D epigenomes

Figure 5

(a) Comparison of hotspots and non-hotspots in terms of chromatin accessibility, histone modifications, gene expression, TF binding and motif enrichment. Peaks of DNaseI-seq, histone modification and TF ChIP-seq data in around 100 cell types (see Supplementary Tables 1–3 for a complete list) were called and the occurrence frequency of peaks were counted for each hotspot and non-hotspot promoter. RPKM values from RNA-seq data in 57 cell types (see Supplementary Table 1 for a complete list) were used to classify promoters into expressed and non-expressed ones (RPKM cut-off=5) and occurrence frequency of expressed promoters were counted for each hotspot and non-hotpot promoter. Motif enrichment values were used to classify each of the 292 TFs as being present or absent in each hotspot and non-hotspot promoter (motif enrichment cut-off=0.6, see ‘Methods’ section). Occurrence frequency of present motifs were counted for each hotspot and non-hotspot promoter. P values calculated using Wilcoxon test are denoted as **P≤0.001; P>0.05. (b) GO term analysis of hotspot promoters.