Fig. 1.

Sample-to-sample correlation of mRNA expression of 100 most differently expressed genes in a human glioblastoma cell line panel does not match with inherent therapy resistance a Tabular presentation of characteristics of the human glioblastoma cell lines as obtained from the Cellosaurus database (https://web.expasy.org/cellosaurus/). b Unsupervised hierarchical clustering and principal component analysis (PCA) of mRNA expression levels of top 20 genes differently expressed between glioblastoma and low-grade glioma (LGG) patient samples (data from the TCGA database (https://tcga-data.nci.nih.gov/docs/publications/lgggbm_2015/), clustering method ward.D and euclidean distance measure) in 560 glioblastoma and 463 LGG patient samples, and in glioblastoma cell lines. c Unsupervised hierarchical clustering and PCA of G-CIMP signatures for hypermethylation phenotypes in 410 glioblastoma and 516 LGG patient samples (data from the TCGA database), and in glioblastoma cell lines. d Sample-to-sample correlation analysis of mRNA expression of 100 genes with highest intra-panel variation in expression in human glioblastoma cell lines. Expression values were determined by global gene expression microarray analysis. PCA-derived scores of inherent resistance (PC1s as described in [7, 8]) to single-shot radiotherapy (RTX), fractionated RTX, TMZ, and TMZ + single-shot RTX are depicted on top by unsupervised hierarchical clustering