Establish a prognostic prediction model of HCC based on the comprehensive analysis of scRNA-seq and bulck RNA-seq
Objective To integrate batch and single-cell RNA-seq data for data mining of cancer-associated fibro-blasts(CAFs)and to explore the relationship between CAF characteristics and prognosis of hepatocellular carcinoma(HCC).Methods HCC scRNA-seq data was obtained from the Gene Expression Omnibus(GEO)database.The scRNA-seq data was analyzed using the Seurat and Monocle 2 software packages to identify cell clusters and differentia-tion trajectories.Additionally,enrichment analysis was performed on the marker gene sets specifically expressed in all cell clusters.RNA-seq gene expression data was then integrated with the corresponding clinical information to identify CAF characteristics.Univariate Cox regression and least absolute shrinkage and selection operator(LASSO)regression analyses were conducted to screen for prognostic-related CAF feature genes,construct a prognostic model,delineate risk groups,and establish a nomogram to validate the predictive efficacy of the model.Results Through the scRNA-seq and RNA-seq data integration analysis,we identified seven HCC cell clusters,and identified the prognosis related 15 CAF genes.By single factor Cox regression and LASSO regression analysis to screen the TTK,EZH2,EME1,SLC7A11,DNAJC6,PNCK,TERB2,S100A8 and PTPRD-AS1 as CAF trait genes.On the basis of these genes we build and veri-fy prognosis characteristics,and patients were grouped according to the characteristics of risk score,the survival time of patients with low risk group was obviously longer than high-risk group.In addition the ROC curve and nomogram risk score model can better assess the prognosis of patients with liver cancer.Conclusion Based on the risk of CAF signa-ture can effectively predict the prognosis of HCC,the signature helps for individualized treatment in patients with HCC.
Liver cancerCAF featurescRNA analysis technologyPrognosis prediction model
NETL
NSTL
万方数据
