Deubiquitinase-associated mRNA signature model predicts prognosis and influences the immune microenvironment in patients with hepatocellular carcinoma: An observational study.

Liver hepatocellular carcinoma (LIHC) represents a category of malignant neoplasms that present a considerable risk to public health. Recent studies have increasingly focused on the biological roles of messenger RNAs (mRNAs) linked to deubiquitinating enzymes in the context of LIHC. These deubiquitinating enzyme-associated mRNAs have been utilized to construct a prognostic model for this type of cancer. Prognostic mRNAs associated with LIHC were identified through univariate Cox regression and co-expression analysis. A clinical risk prediction model was established utilizing multivariate Cox regression and least absolute shrinkage and selection operator analysis, resulting in the stratification of patients into high-risk and low-risk categories. The model's accuracy and clinical significance were assessed through various methodologies, including receiver operating characteristic curves, area under the curve calculations, univariate and multivariate Cox regression analyses, principal component analysis, t-distributed stochastic neighbor embedding, Kaplan-Meier Plotter, gene set enrichment analysis, tumor mutation burden analysis, immune infiltration analysis, and drug sensitivity prediction. The UALCAN database was employed to validate the aberrant expression of the identified characteristic genes, and consistency clustering analysis was conducted to delineate and compare the molecular subtypes of LIHC. The risk model we developed exhibited robust predictive capabilities, with the high-risk cohort demonstrating reduced survival rates across various clinical contexts. This group also presented a more pronounced tumor mutation burden, exhibited stronger correlations with immune cell populations, and displayed heightened activation of numerous immune checkpoints. Notably, the characteristic genes (CBX2, ERGIC3, GNL2) were found to be aberrantly overexpressed in the cancer genome atlas cohort, correlating with unfavorable prognostic outcomes, and may play a role in tumor invasion and metastasis. Consistency clustering analysis revealed 3 distinct subtypes (C1, C2, C3), with subtype C3 showing elevated activation levels at the majority of immune checkpoints in comparison to subtypes C2 and C1, as well as increased sensitivity to pharmacological agents such as 5-fluorouracil and afatinib. The prognostic assessment model developed in this research offers an innovative approach for the identification of novel prognostic markers in patients diagnosed with (LIHC).