Novel Stemness-Based Classification of HCC Patients for Understanding mRNAsi Subgroups and Phenotype Model

To investigate the mechanism underlying the two mRNAsi subgroups and elaborate on the function of the mRNAsi phenotype model, we developed a novel stemness-based classification of HCC patients using multiple bioinformatic analyses. Firstly, we conducted differential expression analysis to identify differentially expressed genes (DEGs) between tumor and adjacent tissue in the TCGA-LIHC and GSE14520 cohorts (with ∣Fold Change∣> 1.2 and P < 0.05). We then performed correlation analysis between all DEGs and mRNAsi score (with person correlation coefficient, ∣R∣> 0.3 and P < 0.05). This resulted in the identification of 19 mRNAsi phenotype-related DEGs, including 13 genes that were increased (Risk, HR > 1) and six genes that were decreased (Protect, HR < 1) in the mRNAsi-high group (Figure 2A).

To account for the heterogeneity of stemness characteristics, we used unsupervised consensus clustering to construct a specific stemness-based classification of HCC patients based on the expression patterns of the 19 mRNAsi phenotype-based DEGs. This resulted in the categorization of patients into three subgroups: the stemness subtype C1 (109 cases, 29.9%), the stemness subtype C2 (138 cases, 37.8%), and the stemness subtype C3 (118 cases, 32.3%). As shown in Figure 2B, 2C, and 2D, patients in the three subgroups exhibited distinct expression patterns of the 19 mRNAsi phenotype-related DEGs.

Furthermore, Kaplan-Meier curve analysis showed that patients in stemness subtype C3 had a significantly worse prognosis compared to those in stemness subtypes C1 and C2 (Figure 2E). These findings suggest that the novel stemness-based classification of HCC patients is a valuable tool for investigating the mechanism underlying mRNAsi subgroups and for elaborating on the function of the mRNAsi phenotype model.