Translational Genomics Research Institute (complete)

Glioblastoma Multiforme Orthotopic Xenograft Transcriptome

This dataset contains the mRNA transcriptional profiles of 39 human primary glioblastoma orthotopic xenografts.

Experimental Approaches

The expression study was performed on the Agilent-014580 Whole Human Genome Microarray 4 X44K G4112F (GPL6480). This multi-pack (4X44K) formatted microarray represents a compiled view of the human genome as it is understood today. The sequence information used to design this product was derived from a broad survey of well-known sources such as RefSeq, Goldenpath, Ensembl, Unigene, and others. The resulting view of the human genome covers 41,000 unique genes and transcripts which have been verified and optimized by alignment to the human genome assembly and by Agilent's Empirical Validation process. The full raw data files are deposited in GEO (series GSE38814). The .gct matrix file contains log2 transformed, lowess normalized, and probe level expression values for each xenograft.

Identification of Pathways Enriched with Condition-specific Statistical Dependencies Across Four Subtypes of Glioblastoma Multiforme (GBM)

Evaluation of Differential DependencY (EDDY) is a statistical test for the differential dependency relationship of a set of genes between two given conditions. For each condition, possible dependency network structures are enumerated and their likelihoods are computed to represent a probability distribution of dependency networks. The difference between the probability distributions of dependency networks is computed between conditions, and its statistical significance is evaluated with random permutations of condition labels on the samples.  

For this project, the CTD² Center at the Translational Genomics Research Institute applied EDDY to the gene expression data of glioblastoma multiforme (GBM) from The Cancer Genome Atlas (TCGA) to reveal the functional difference between the four subtypes of GBM - Proneural (PN), Neural (N), Mesenchymal (MES) and Classical (CL).  The results show that the proposed method can identify novel gene sets that could not be found with Gene Set Enrichment Analysis, which is considered a representative method of considering only differential expressions, while providing many results specific to the subtypes of GBM.

Experimental Approaches

Gene expression data of GBM (AgilentG4502A_07; Level 3) obtained from TCGA were used for these analyses.

Quantified Cancer Cell Line Encyclopedia RNA-seq Data

Many applications analyze quantified transcript-level abundances to make inferences.  Having completed this computation across the large sample set, the CTD² Center at the Translational Genomics Research Institute presents the quantified data in a straightforward, consolidated form for these types of analyses.

Experimental Approaches

After downloading RNA-seq data for 935 cell lines from the Cancer Cell Line Encyclopedia (CCLE), transcript-level abundance was quantified using Salmon¹. All data were aligned using Salmon 0.4.2 using Homo Sapiens GRCh37.74 for reference. Raw BAM files used to generate this data is avaliable at GDC. The resulting 935 quantification files, named by sample ID, have 4 columns for ensemble gene ID, length, number of reads, and transcripts per million (TPM). Other Salmon arguments were "--libType IU" (inward, unstranded).