Fig. 2

Gene expression analysis and risk ratio’s. a depicts the cumulative number of differentially expressed genes (y-axis) as a mean with 95% CI (of 1000 times 200 random sampling) plotted against the p value used as cut-off to define differential expression (x-axis). b is a visualization of the genes that display significant differences between tumor stages in the whole group (ANOVA p < 0.05, n = 2764) using a t-SNE algorithm with clear separation of the immune (CMS1), mesenchymal (CMS4) and epithelial subtypes (CMS2/3). c displays the correlation between disseminated disease (stage III-IV) (x-axis) and CMS4 (y-axis) signature scores (r = 0.77, p < 0.001). d shows risk ratio’s for progression to advanced stages of disease (stage III-IV) and a 95% confidence interval comparing the different CMS subtypes