Fig. 6: Early epigenetic activation of RAS/MAPK pathway is later reinforced by genetic activation.

a H3K27me3 ChIP-Seq density was profiled ± 15 kb around the transcription start site (TSS) of RAS target genes in parental and CRISPR-M27K BT245 pHGG cells⁴⁹. b Gene set enrichment analysis (GSEA) of RAS target genes in CRISPR-M27K versus parental BT245 pHGG cells⁴⁹. NES: Normalised Enrichment Score. FDR: false discovery rate. c H3K27me3 ChIP-Seq density was profiled ± 15 kb around the TSS of RAS target genes in parental, H3.3K27R or H3.3K27M G477 cells⁴⁹. d H3K27me3 ChIP-Seq density was profiled ± 15 kb around the TSS of RAS target genes in H3K27M (red) and H3WT (blue) pHGG cells⁴⁹. e Western blot of control and H3.3K27M mouse tumour tissue showing increased phospho-ERK in tumours versus their respective normal counterparts (Blots were run twice with similar results). #: RAS pathway WT, ^: RAS pathway mutant. f Proportion of H3.3K27M DIPG or mouse tumours with RAS pathway or MYC mutations. g RAS target gene ssGSEA ES from RAS pathway WT (n = 8)/mutant (n = 18) H3.3K27M DIPG and normal brain (n = 20). ES: enrichment score. h Clonal evolution of single nucleotide (SNV) and copy number (CNV) variants in KP29 high-grade glioma (HGG) and lymphoma, showing known tumour drivers. The circle area of the clones and germline (normal) is proportional to the percentage of the sample comprised of each clone. Note that clone 2 evolved from clone 1, and thus contains all the alterations in clone 1. Driving events are shown. Source data are provided as a Source Data file. Statistical tests: ANOVA (g). NS: not significant.