Module selection and module scores of BRCA and combined LUAD-LSCC cancer datasets. Module scores are calculated for each sample and each module by taking the log2 of the average abundance of all members in a module in a sample for (A) BRCA and (B) combined LUAD and LSCC cohorts. C, modules with significant correlations with sample annotations. DHX9, FAT1, PDGFRA, POLQ, RARA, STK11 indicate gene mutations.