Fig. 2. Developmental origins and oncogenic pathways regulated by the myeloma enhanceome.

a Heatmap showing the ATAC-seq signal for all peaks found to be differentially open and within 1 Mb of a significantly differentially regulated gene. Data are row scaled. Samples are clustered using Pearson’s correlation distance and the bar above the columns shows the subtype of the sample coded as in Fig. 1a, i.e., MAF: orange, CCND1: light blue, MMSET: green, HD: blue, others: pink and ND PC: black/grey. Vertical bars to the right highlight regions where signal is >2-fold different to ND PC. b Example region around HGF showing: average ATAC-seq signal in each subtype; H3K27ac signal in a MAF-translocated cell line (orange) or CCND1-translocated cell line (blue); FP density: density of footprints in each of the ATAC-seq signals. c, d Enrichment analysis for upregulated genes with a peak of increased accessibility within 1 Mb using gene sets from the “Oncogenic signatures”, “Hallmarks” or “Curated gene sets” subsets of the MSigDB database. e Chromatin state of differentially open pan-MM peaks across a developmental range of B-cell types as determined by the Blueprint Epigenomics consortium using ChromHMM (nB naive B cells, GCB germinal center B cell, mB memory B cell, tPC tonsil PC, MM myeloma PC). Enhancers with de novo formed peaks in myeloma (254) are indicated. Chromatin states with strong combined H3K27ac and H3K4me1 signals were considered as active enhancers and indicated by an asterisk. f Radial plot of Homer motif analysis displaying the top 50 over-represented TF motifs in de novo myeloma enhancers.