Fig. 2

Macro-scale pan-cancer subgroups reveal clinical relevance. a Patient distribution of 12 cancer types (COAD and READ are treated as one type) in pan-cancer subgroups with k = 3. For a patient set of each cancer type in a subgroup, the significance of enrichment was evaluated using a chi-squared test (see Methods). P-values lower than 1 × 10⁻¹⁰ were set as 1 × 10⁻¹⁰ for convenient visualization. The number denotes the size of the corresponding patient cohort of each cancer type in this subgroup. b-f Kaplan-Meier survival curves of patients in significantly enriched subgroups and the remaining ones for cancer types OV b, LAML c, HNSC (d), LUAD (e), and LUSC f were plotted, respectively. P-values were derived from the log-rank test. g-h Pie plots demonstrating the distributions of tumor stage and grade on patients in significantly enriched subgroups and others for cancer types KIRC (g) and UCEC h. P-values denoting the significant level of the difference between the distributions of two groups were calculated by Fisher’s exact test