Fig. 1: Expression invariance of oncogenes.
From: Globally invariant behavior of oncogenes and random genes at population but not at single cell level
Various types of statistical analysis were performed on bulk datasets to show the expression invariance of cancer genes. The analysis for transcriptomics samples is on the right panels, and the analysis for the proteomics samples is on the left panels. a Scatterplots between normal samples, tumor samples, and normal vs tumor samples for liver cancer and ovarian cancer, with the remaining types presented in Supplementary Fig. 1. Regular genes are represented by gray dots, CGC genes by blue dots, and CSO genes by red dots. b Pearson correlation for the expression levels of CGC genes (blue), CSO genes (red), CGC-sized sampled random genes (purple), CSO-sized sampled random genes (orange). Ovarian cancer was selected as an example for both proteome and transcriptome here, and the rest of the cancer types are in Supplementary Fig. 2. c PCA plots for whole dataset normal samples (light blue circles), whole dataset tumor samples (light orange circles), CGC genes normal samples (dark blue circles), CGC genes tumor samples (dark orange circles), CGC-sized random sampling of genes from normal samples (dark blue stars), CGC-sized random sampling of genes from tumor samples (dark orange stars), and the rest of the cancer types are in Supplementary Fig. 4.
