Fig. 6.

Application of TAPINTO to assess the characterizations of tumor heterogeneity. (A) TAA scores derived from TAPINTO were able to identify oncogenic pathways. The TAA score was used with the gene set enrichment analysis to identify 28 significantly enriched hallmark gene sets associated with tumor progression from MSigDB. The enrichment plots of 6 well-known cancer hallmarks are shown. (B) The disease enrichment analyses of the genes with fold changes in the top 10% calculated by 3 methods (TAPINTO, LIMMA, and DIDS) provided different degrees of identification for subtypes of breast cancer. The circle dots represent the gene ratio, and the P values are displayed with colors. (C) The survival risk was analyzed based on TAPINTO and potential TAAs (ERBB2, CXCL9, KCNJ3, MMP11, SQLE, and SLC7A2) that had a score of 6 in the binary logistic regression. The 1,097 breast cancer samples with overall survival (OS) annotation from TCGA were divided into 2 groups: high group (risk scores ≥ median) and low group (risk scores < median). Their OS at 8 years was calculated and displayed in a Kaplan–Meier plot, and the significance was determined by the log-rank test (P = 0.00047).