Transcriptomic Profiling of Glioblastoma

The transcriptomic data of GBM patients were obtained from two independent data portals (CGGA and TCGA). As part of the CGGA project, Zhao et al.^{46 (link)} provided the RNA-seq transcriptomic profiles of 325 gliomas samples. The raw RNA-seq data were downloaded from the NCBI Sequence Read Archive (SRA, https://www.ncbi.nlm.nih.gov/sra) using the accession numbers SRP027383 and SRP091303. The raw fastq files were converted from SRA using Fastq-dump app (version: 2.8.1), and the sam and read count data were obtained by the hisat2 (version 2.1.0) and htseq software (version 0.10.0), respectively. A total of 137 GBM samples with clinical data were selected from the CGGA cohort. Data of an independent cohort of 158 patients was obtained from the TCGA database. The transcriptomic profiles (level 3 data) and corresponding clinical data of these patients were downloaded from the Data Coordinating Center. All RNA-seq libraries were sequenced using the Illumina HiSeq2000 Systems. The raw count data were normalized by the trimmed mean of M-values (TMM) method^{47 (link)}, and genes with extremely low total abundance (count per million <1) were filtered out. The data preprocessing was performed using edgeR package^{48 (link)} (version 3.20.9) in Bioconductor.