| Objective:Esophageal cancer usually does not have obvious symptoms at the early stage of the disease,and the patients’ clinical diagnosis time is relatively late.Surgical treatment for advanced esophageal cancer is difficult to achieve good results,which leads to a significant decrease in patient survival rate.It can be seen that the prediction of surgical prognosis is an important decisive factor to ensure a high survival rate of patients.This article uses the Weighted Gene Coexpression Network Analysis(WGCNA)method to study the potential molecular mechanisms of esophageal cancer(ESCA),in order to identify the corresponding hub genes and provide a basis for the early diagnosis,treatment,and prognosis of ESCA.Methods:Download the Cancer Gene Atlas(TCGA)of esophageal cancer and its RNA sequencing and clinical feature data,combine the differences in related gene expression levels,and use WGCNA to construct a prognostic model for esophageal cancer.Draw the Receiver Operating Characteristic Curve(ROC)and Kaplan Meier curve evaluation models for patients.Draw ROC curves and screen out key genes closely related to esophageal cancer,and analyze and evaluate the predictive ability of candidate biomarkers and classification prediction models.Results:Ten gene modules related to the clinical characteristics of ESCA patients were identified,from which turquoise color modules and black modules highly related to cancer incidence were obtained,and eight hub genes were screened: chromosome related kinesin KIF4,chymotrypsin like elastase family member 2A(CELA2A),DNA replication factor CDT1,cyclin dependent kinase inhibitor 3(CDKN3)Ribonucleoside diphosphate reductase subunit M2(RRM2),Myb proto oncogene 2(MYBL2),pepsinogen 3(PGA3)and pepsinogen 4(PGA4).Conclusion : WGCNA identified some modules related to esophageal cancer(turquoise module has the highest correlation)and hub genes,which will play a certain reference significance for the follow-up study of molecular mechanism and targeted treatment of esophageal cancer. |