Construction of A Novel Prognostic Model with Molecular Targets for Hepatocellular Carcinoma Based on Bioinformatics Applications

Abstract

We aimed to analyze the differentially expressed genes associated with hepatocellular carcinoma (HCC) by bioinformatics method and to identify potential molecular targets for immunotherapy and molecular indicators for predicting HCC prognosis. Gene Expression Omnibus (GEO) was used to download the hepatocellular carcinoma related microarray data. The R language’s Limma tool was used to identify the genes with differential expression. For differentially expressed genes, GO (Gene Ontology) enrichment analysis, KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway analysis, and protein–protein interaction analysis were carried out. PPI (protein–protein interaction) regulation network construction. In order to further evaluate HCC specific differentially expressed genes, HCC specific expression analysis was carried out at the same time by merging with other tumour RNA-seq transcriptome data in the TCGA database. The relationship between immune-related LncRNA and independent risk factors was examined using univariate and multivariate Cox regression as well as Least absolute shrinkage and selection operator (LASSO) analysis. The proportional Hazards model (COX model) was utilized to model the chosen important genes and predict the prognosis. We obtained important genes through additional screening, and the GSE6764 validation set discovered that the expression of these genes decreased with increasing tumor stage (P < 0.05). The prognosis analysis of the gene model revealed that the high-risk group had a dismal outcome. COX modelling was carried out for important genes. Meanwhile, the GSE76427 and GSE54236 validation sets validated the model’s survival analyses. By analyzing the gene expression profile of HCC utilizing cuttingedge bioinformatics techniques including Cox Regression and LASSO analysis, we were able to screen out the important modules and essential genes, create a predictive model for HCC, and propose possible biomarkers for the prediction of HCC.