Machine Learning-Based Disulfidptosis-Related lncRNA Signature Predicts Prognosis, Immune Infiltration and Drug Sensitivity in Hepatocellular Carcinoma

Abstract

Abstract Disulfidptosis plays a crucial role in the development and progression of Hepatocellular Carcinoma (HCC). However, the significance of disulfidptosis-related Long non-coding RNAs (DRLs) in the prognosis and immunotherapy of HCC remains unclear. Based on The Cancer Genome Atlas (TCGA) database, we used Least Absolute Shrinkage and Selection Operator (LASSO) and Cox regression model to construct DRL Prognostic Signature (DRLPS)-based risk scores. Survival analysis was then performed and a nomogram was constructed. Moreover, we performed functional enrichment annotation, immune infiltration analyses and drug sensitivity analyses. Five DRLs, including AL590705.3, AC072054.1, AC069307.1, AC107959.3 and ZNF232-AS1, were identified to construct prognostic signature. DRLPS-based risk scores exhibited a better predictive efficacy of survival than conventional clinical features. The nomogram showed a high degree of congruence between the predicted survival and observed survival. Gene set were mainly enriched in cell proliferation, differentiation and growth function related pathways. Immune cell infiltration in the low-risk group was significantly higher than that in the high-risk group. Additionally, the high-risk group exhibited higher sensitivity to Afatinib, Fulvestrant, Gefitinib, Osimertinib, Sapitinib, and Taselisib. In conclusion, our study highlighted the potential utility of the constructed DRLPS in the prognosis prediction of HCC patients, which demonstrated promising clinical application value.