Fig. 1: Highly expressed genes prognostic for overall survival of GBM patients fall into the categories unfolded protein response, glycolysis, hypoxia and mTORC1 signaling.

A Schematic workflow of the GBM transcriptome analysis for candidate identification based on TCGA GBM patient datasets. B Depiction of differentially expressed genes (DEG; GBM versus normal brain) found in the TCGA GBM patient cohorts. Significantly up- and downregulated genes are shown in red and blue, respectively, and determined by ANOVA with a cut-off |Log₂ (fold change GBM versus normal brain)| > 0.3, p <  0.05, using GEPIA web tool. C Alignment of differential gene expression levels from TCGA, presented as Log₂ (fold), and the gene expression levels from the RNA-seq data of GBM cell models (Log₂ CPM; n = 4) shown as Z-score. Heatmap was created with R and data were hierarchically clustered in rows based on DEG and in columns based on gene expression. D Waterfall plot of overexpressed genes from TCGA GBM patient cohorts ranked by high gene expression in GBM models displayed as -Log₁₀ (p-value; t-test; p < 0.05). The top-200 highly expressed genes were used for the selection of gene candidates associated with shorter overall survival of GBM patients (indicated). E Comparative mRNA expression in tumor and normal brain tissue. Data were obtained from TCGA GBM Affymetrix HT HG U133A dataset (Oncomine.com) and analyzed by one-way ANOVA (***p < 0.005). F Overall survival analyses based on gene expression (low versus high) from the TCGA GBM patient cohorts. Kaplan-Meier curves display the confidence intervals, the log-rank test p-values and the number of patients at risk. Data were downloaded from betastasis.com. G Identified hallmark biological processes for the 12 gene candidates from ‘E’ and ‘F’ using the hallmark gene set collection in gsea-msigdb.org.