Fig. 1

Analysis pipeline of single sample data from TCGA-LGG and TCGA-GBM, and a multi-sample JPN-LGG dataset. A read count matrix obtained from RNA-sequencing experiments for LGG (631 patients) and GBM (169 patients) was used to generate a list of differentially expressed genes (tumour vs. normal samples). The resulting list was intersected with the list of differentially expressed genes retrieved from GEPIA2 [17]. The final list of differentially expressed genes in LGG and GBM was analysed for enriched KEGG pathways, and the analysis of expression patterns of genes of interest. A list of genes affected by SSMs or CNAs in LGG and GBM was obtained. From the list, driver genes with annotated mutations, losses or gains were selected and used for the reconstruction of progression patterns in LGG and GBM. Phylogenetic trees were reconstructed based on the mutations and CNAs from multi-sample data in JPN-LGG cohort. The temporal ordering of alterations in genes of interest was obtained from the phylogenetic trees in JPN-LGG and progression patterns in the TCGA-LGG and TCGA-GBM