Fig. 1

Meta-analysis of pancreatic ductal adenocarcinoma (PDAC) transcriptome data. a Unsupervised principal component analysis plots across the first three principle components (PC) on individual datasets included in the meta-analysis. The variance captured by each PC is shown along the respective axis. PDAC samples are represented in red and normal samples in black. In all datasets, PDAC and normal samples formed separate clusters marked with ellipses. b Heatmap of fold change (Fc) of top 100 significantly differentially expressed genes identified from meta-analysis (abs(Fc) ≥ 2 in at least one of the datasets; false discovery rate ≤ 0.05). Datasets are shown as columns and genes are shown as rows. Relative fold changes are shown with a pseudocolor scale (–1 to 1), with red denoting upregulation and green denoting downregulation