A systems biology approach to unveil shared therapeutic targets and pathological pathways across major human cancers.

Cancer remains a major cause of global mortality. Despite their distinct clinical and molecular characteristics, different cancer types often share fundamental molecular mechanisms that remain underexplored. In this study, we systematically profiled transcriptomic data from four highly prevalent cancers, including breast, lung, colorectal, and prostate, to uncover shared molecular signatures with diagnostic and prognostic value. Using The Cancer Genome Atlas (TCGA) datasets and a rigorous integrative workflow, we combined differential expression analysis, Elastic Net-based feature selection, and weighted gene co-expression network analysis (WGCNA) to identify 179 cross-cancer signature genes linked to clinical traits. Protein-protein interaction (PPI) analysis and Markov Cluster Algorithm (MCL) clustering further refined these into 26 robust hub genes with strong diagnostic potential. Survival analyses demonstrated that these hub genes possess strong prognostic potential, while pan-cancer assessment revealed consistent dysregulation across more than 20 cancer types. Several hub genes also displayed context-dependent immunomodulatory roles within the tumor microenvironment. Notably, 16 hub genes showed strong associations with metastatic disease: some were consistently downregulated, suggesting tumor-suppressive functions, whereas others were upregulated in a cancer-specific manner, reflecting context-dependent oncogenic roles. We constructed a hub gene-based signature and demonstrated its potential as a prognostic marker in these four cancers types. This comprehensive analysis offers valuable insights into shared oncogenic mechanisms, contributing to improved diagnosis, prognosis, and targeted therapies across multiple cancer types.