Integration of single-cell and bulk RNA-sequencing to analyze the heterogeneity of hepatocellular carcinoma and establish a prognostic model.

backgroundThe highly heterogeneous nature of hepatocellular carcinoma (HCC) results in different responses and prognoses to the same treatment in patients with similar clinical stages.

aimsThus, it is imperative to investigate the association between HCC tumor heterogeneity and treatment response and prognosis. METHODS AND

resultsAt first, we downloaded scRNA-seq, bulk RNA-seq, and clinical data from TCGA and GEO databases. We conducted quality control, normalization using SCTransform, dimensionality reduction using PCA, batch effect removal using Harmony, dimensionality reduction using UMAP, and cell annotation-based marker genes on the scRNA-seq data. We recognized tumor cells, identified tumor-related genes (TRGs), and performed cell communication analysis. Next, we developed a prognostic model using univariable Cox, LASSO, and multivariate Cox analyses. The signature was evaluated using survival analysis, ROC curves, C-index, and nomogram. Last, we studied the predictability of the signature in terms of prognosis and immunotherapeutic response for HCC, assessed a variety of drugs for clinical treatment, and used the qRT-PCR analysis to validate the mRNA expression levels of prognostic TRGs.

conclusionTo conclude, this study expounded upon the influence of tumor cell heterogeneity on the prediction of treatment outcomes and prognosis in HCC. This, in turn, enhances the predictive ability of the TNM staging system and furnishes novel perspectives on the prognostic assessment and therapy of HCC.