Construction of a Novel Prognostic lncRNA Signature Based on Edge Perturbation Network in Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is an aggressive liver cancer with poor prognosis and few available therapeutic options. Long non-coding RNAs (lncRNAs) have been identified as promising prognostic biomarkers. In this study, we constructed an innovative prognostic model for HCC by employing the edge perturbation network method. Firstly, RNA transcriptomic data, along with clinical and mutation information, were retrieved from the Cancer Genome Atlas (TCGA) database. By means of edge perturbation network approach, we identified 104 lncRNAs potentially associated with HCC prognosis. Subsequently, the prognostic model was established by selecting seven pivotal lncRNAs through univariate Cox, least absolute shrinkage and selection operator (LASSO), and multivariate Cox regression analysis. The results of Kaplan-Meier survival analysis combined with receiver operating characteristic (ROC) curve demonstrated that the proposed prognostic model had a strong prognostic performance (5-year area under the curve (AUC) = 0.78, 0.75, and 0.77 for the training, testing, and the entire cohort, respectively). Subsequently, univariate and multivariate Cox regression analyses of risk scores and clinical features confirmed that the risk score (RS) is an independent prognostic factor for HCC. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) were utilized to identify the signaling pathways related to the HCC in this system. Patients in the low-risk group exhibited better prognosis and lower mutation rates. In conclusion, the proposed lncRNA-based model provides a reliable tool for HCC prognosis and offers new insights for patient management and treatment.