2 Methods

TCGA-SARC dataset is retrieved using TCGAbiolinks package. Raw HTSeq counts are normalized using DESeq2 variance stabilizing transformation. TCGA barcodes have meaning explained here. Importantly, downloaded samples should be to select only tumors with values <10 in the 14 and 15 positions in the barcode.

Information will have prefix 'paper_'. For SARC subtype, they come from:doi:10.1016/j.cell.2017.10.014

Figure 2.1: All projects with retrievable data

Figure 2.2: Selected projects for this study, to match with Kaplan-Meier analyses done by Philippe Naquet and Richard Miallot