| Objective: Endometrial cancer is a common gynecological malignancy.Most patients are at early stage and have a good prognosis.While,a late clinical stage at diagnosis,special pathological types and insensitive response to treatment may lead to adverse clinical outcomes.In order to identify the mechanism of tumor development and screen more appropriate treatments,breakthroughs have been made in the field of mining of tumor-related genes in recent years.Increasing evidence suggests that pathogenicity abnormalities arise from genetic interactions in many complex network relationships.Therefore,this study aims to make a comprehensive analysis of gene function,pathways and mutual relationships,and explore a gene expression model for judging the prognosis of endometrial cancer patients,so as to provide new ideas for the rational treatment of endometrial cancer and the improvement of the prognosis of patients.Methods: Two datasets from GEO public database: GSE36389 and GSE115810 were included and merged,including G1 samples as control group,G2 and G3 samples as experimental groups,performed differential expression and GO enrichment analysis of genes between the two groups.The WGCNA method was applied to cluster genes with similar expression patterns,and all genes within the module highly associated with grade traits were extracted as candidate genes for subsequent analysis.Gene expression and clinical information of endometrial cancer patients were obtained using the TCGA public database,and univariate analysis was performed to identify the genes associated with prognosis among the candidate genes.LASSO regression analysis was used to construct the gene prognosis model of histological grade of endometrial cancer patients and verify the model,evaluate the efficacy of the model and conduct the difference analysis of clinical indicators.Subsequently,differentially expressed genes between high and low risk groups were extracted for GO enrichment analysis,based on which a protein interaction network was constructed.Results: A gene expression matrix with a total sample number of 37 was obtained,which was divided into 2 groups according to the endometrial cancer grade and used for differential expression analysis and GO enrichment analysis.37 samples,1655 genes were included for WGCNA analysis,resulting in seven modules,the Green module is the most strongly associated with endometrial cancer grade progression,containing 157 genes.The gene expression data and clinical survival data of endometrial cancer patients were downloaded through TCGA public database to extract all genes included in the Green module and screen the differentially expressed genes between tumor and normal tissue,and further select 137 DEGs,of which 93 were downregulated and 44 were upregulated.Twenty one genes associated with endometrial cancer prognosis were identified using univariate analysis.LASSO regression analysis yielded a prognostic model constructed with 11 key genes.The Kaplan-Meier analysis demonstrated a significant decrease in overall survival at high risk compared to low risk(P <0.001),and the overall survival AUC at 5 years in this prognostic model was greater than each clinical parameter and can be used as an independent risk factor for endometrial cancer patients.Models performed well in the validation group.The results suggested that the risk score also increased with the increase of grade,and the risk score was significantly different(P <0.05 among each grade group).A total of 363 genes were differentially expressed between the high and low risk groups,which were subsequently subjected to functional enrichment analysis and protein interaction network analysis.Conclusion: Based on WGCNA and LASSO analysis,we first identified a prognostic model related to endometrial cancer grade,and this prognostic model may be an independent prognostic factor,which is conducive to further understanding of the molecular pathogenesis and progression of endometrial cancer. |
PDF Full Text Request