| Background:Esophageal squamous cell carcinoma,as one of the most common malignant tumors in Asians,causes many deaths every year and brings a huge burden to the country and society.Because the clinical manifestations of early esophageal cancer are easy to be ignored,patients with esophageal cancer are not diagnosed in the early stage On count of the high-speed advance of technology applied to human genome,a great deal of biological data comes into being,and bioinformatics analysis is deep and wide applied to the study of kinds of cancers.This research is on the strength of the means of bioinformatics analysis to explore novel crucial genes in connection with oesophageal squamous cell carcinoma and build a prognostic model.It may be helpful for the early detection and subsequent clinical treatment of esophageal squamous cell carcinoma.Methods:The gene expression data of esophageal squamous cell carcinoma were obtained from public databases,such as GEO database,TCGA database,and related clinical information such as age,gender,tumor TNM stage,etc.were also collected.In the tumor group and the normal tissue group,bioinformatics analysis was used to identify differentially expressed genes(DEGs).p-value is less than 0.05.Weighted gene co-expression network analysis(WGCNA),this analysis is a biological algorithm that classifies genes into different modules,calculates the correlation between modules and specific clinical features,and obtains the corresponding target genes.Based on this statistical method,a group of genes that are significantly correlated with tumors were further screened from the differential genes in the TCGA group,and then intersected with the differential genes in the GEO group,and finally our tumor-related gene set was obtained,meanwhile,functional enrichment analysis was conduct.Finally,according to the patients in the GEO group,univariate Cox analysis method,lasso regression analysis method and multivariate Cox analysis method were successively used for this gene set to obtain genes significantly related to the prognosis of esophageal squamous cell carcinoma,and the risk model was constructed based on the obtained genes.A prognostic model was utilized to predict survival in patients with esophageal squamous cell carcinoma,and ROC curves and TCGA panels were applied to test the reliability of this model.Results:785 differentially tumor-related genes were finally obtained,and functional enrichment analysis showed that these cancer-related genes were highly correlated with tumor characteristics and various metabolic processes.7 of these genes were significantly associated with the prognosis of esophageal squamous cell carcinoma,namely AMBP,ANO1,GNMT,PLAU,RPL3L,CTBP1-AS,and SLC39A6.A prognostic model was established based on these 7 genes,and the ROC curve showed good survival prediction performance in both TCGA and GEO cohorts.Prognostic models can classify patients into high-risk and low-risk groups using risk score.Univariate and multivariate Cox regression analysis showed that the risk score model based on 7 genes could independently predict overall survival(OS)in ESCC.Finally,a nomogram was established to predict OS in patients with esophageal squamous cell carcinoma.Conclusion:The purpose of this study is to screen out the key genes associated with esophageal squamous cell carcinoma by means of bioinformatics,and these genes may become new targets for the diagnosis and treatment of esophageal squamous cell carcinoma in the future.Finally,a risk score prognostic model was built using these genes,and the performance of the model was validated in both training and experimental groups.A nomogram based on this model may be a reliable tool for predicting the prognosis of patients with esophageal squamous cell carcinoma. |
PDF Full Text Request