Data Mining And Preliminary Analysis Of Prognostic Related Genes In Esophageal Squamous Cell Carcinoma Based On Public Databases

Objective:The purpose of this study was to analyze the data sets of GSE23400,GSE20347 and GSE45670 of esophageal squamous cell carcinoma by bioinformatics methods,The genes related to prognosis of esophageal squamous cell carcinoma(ESCC)were screened out,and the genes were verified to provide valuable biomarkers for the treatment of patients with ESCC.Methods:In this study,the data set was retrieved from the gene expression comprehensive database(GEO),and the co-expression of different genes in the selected ESCC data set was analyzed.Then the biological information of differentially expressed genes in multiple data sets was analyzed by the pathway enrichment analysis in "Kyoto Encyclopedia of Genes and Genomes"(KEGG).The prognostic value of related genes in the top 5 pathways of KEGG pathway analysis was verified in the cancer genome atlas database(TCGA).Combining TCGA results and literature review,the differentially expressed genes that are ultimately used for validation were determined.The prognostic value of identified differentially expressed genes in patients with ESCC was evaluated by immunohistochemistry in our own cohort.Results:Three data sets of GSE23400,GSE20347 and GSE45670 ESCC were retrived from GEO.There were 201 overlapping differentially expressed genes with statistical differences in the three data sets,among which 152 were up-regulated genes and 49 were down-regulated genes.KEGG pathway enrichment analysis of up-regulated overlapping differentially expressed genes.Up-regulated overlapping differentially expressed genes areenriched in 19 KEGG pathways,including pathways such as cell cycle,cancer,DNA replication,extracellular matrix-receptor interactions,and small molecule RNA in cancer.We evaluated all overlapping differentially expressed genes for the top 5 pathways involved in KEGG in the ESCC cohort in the TCGA database.The results showed that the high expression level of CDC25B(P=0.022),CXCL8(P=0.004),FZD6 mRNA(P=0.033)and low expression level of MCM4 mRNA(P=0.020)in patients with ESCC and the overall survival of patients(OS)and progression-free survival(PFS)were negatively correlated.We found that the prognostic value of FZD6 in ESCC patients has not been evaluated before.Finally,we used immunohistochemical method to detect the expression level of FZD6,and analyzed the relationship between its expression level and patient prognosis.The results found that FZD6 high expression was negatively correlated with overall survival of ESCC patients(P=0.005).COX regression analysis showed that N stage(P=0.026)and FZD6 expression level(P=0.001)were independent prognostic factors of OS in ESCC patients.Conclusion: In this study,GEO was analyzed through a variety of bioinformatics techniques,and genes related to the prognosis of CDC25 B,CXCL8,FZD6,and MCM4 in patients with ESCC were screened.And immunohistochemical analysis confirmed that the high expression of FZD6 in patients with ESCC was significantly associated with poor prognosis.