TCGA.zip

Gene expression data set, originally acquired in the scope of the TCGA-BRCA program.We downloaded raw expression counts of primary tumor tissue and corresponding metadata information from the NCI Genomic Data Commons public database.From the downloaded data, we removed all samples without PAM50 subtype assignment in the metadata.We further filtered genes that had less than 60 read counts in more than 30 percent of a class' samples and applied TMM normalization with subsequent log2-transformation.The final data set contains 1,090 samples and 20,950 genes.