Fig. 2

Exploring average profiles and heterogeneity in region sets. A genome-wide region set of protein-coding genes was analyzed based on published ChIP-seq and Hi-C data, centered on the TSS. Data show averages of genes localized on the forward strand, with the gene body facing towards the right from the TSS (center coordinate). a Average ChIP-seq reads of CTCF and SMC3 (i), average Hi-C contact maps of unperturbed wildtype cells (ii), and average Hi-C contact maps of NIPBL-depleted cells (iii). b Heterogeneity of histone posttranslational modifications within the protein-coding gene region set, visualized by stacked line profiles sorted by the read density of H3Kac (i) and displayed for H3K27me3 ChIP-seq read density (ii), with the sorting order coupled to i, and regional heterogeneity analysis based on 10 histone posttranslational modifications by embedding and clustering (iii). c, d Regional subsets created by clustering as shown in b (iii) were analyzed for 10 different histone posttranslational modifications. Red indicates enrichment, blue indicates depletion. c Cluster 1 contains TSS regions of protein-coding genes enriched in marks for actively transcribed chromatin. d Cluster 2 contains TSS regions enriched in marks for transcriptionally repressed chromatin. e, f Chromosome conformation analysis around TSS region subsets as shown in c, d. e HiC average contact maps of cluster 1 from unperturbed wildtype cells (i) and cells depleted of NIPBL (ii). f HiC average contact maps as in e for cluster 2