Establishment Of Prognostic Model For Deubiquitination Based On Bioinformatics Analysis In Cervical Cancer

Objective The main purpose of this study is to investigate the correlation between deubiquitination and the prognosis of cervical cancer patients,and to construct a prognosis model related to cervical cancer deubiquitination.To provide certain ideas for the individualized diagnosis and treatment of cervical cancer patients and the study of the molecular mechanism.Methods Cervical Cancer gene expression data were obtained through The Cancer Genome Atlas(TCGA)database.Obtain deubiquitination gene data from the iUUCD 2.0Database.Based on the DEGs,single sample Gene Set Enrichment Analysis(ssGSEA)to obtain cervical cancer deubiquitination related scores.Weighted Correlation Network Analysis(WGCNA)was used to cluster different modules,and the modules with the strongest correlation with deubiquitination were selected,and key genes in the modules were identified.Cox regression analysis and other statistical methods were used to obtain prognostic genes to obtain cervical cancer deubiquitination risk score and construct prognostic model.KM survival Curve and Area Under Curve(AUC)were used to predict the overall survival rate of patients to evaluate the prediction quality of this model.Cervical cancer related data were obtained from Gene Expression Omnibus database(GEO)as a test set to further confirm the predictive ability of this model.Finally,this model can be used as an independent prognostic factor for cervical cancer.Results DNA microarray expression data of 306 cervical cancer samples were obtained from the TCGA database,and 293 cases had complete clinical data after integration.98 human deubiquitination enzymes(DUBs)data with expression information were obtained and integrated from the iUUCD2.0 database.A total of 47 DEGs of cervical cancer deubiquitination were obtained.PPI Networks(Protein-Protein Interaction Networks)were analyzed and constructed.Enrichment analysis showed that these genes were mainly involved in biological processes such as post-translational modification and hydrolysis of proteins.Based on the above DEGs,cervical cancer samples of the TCGA training set were analyzed by ssGSEA,and cervical cancer deubinization related scores were obtained.The R-surv＿cutpoint was used to divide the training set into high and low groups based on the optimal node,and the results showed that there was a significant difference in KM survival curve between groups(p=0.032).The weighted co-expression network analysis of protein-coding genes in cervical cancer samples was conducted,including 4177 genes.The parameter GS(Gene Significance)<0.5 and MM(Module Membership)< 0.6,then 2,067 key genes were screened.Using univariate Cox regression,143 of the above key genes were associated with OS(p<0.05).A risk score composed of24 genes was obtained after dimensionality reduction by Lasso regression analysis to construct a prognosis model related to deubiquitination of cervical cancer.Kaplan-Meier analysis showed that the model could significantly distinguish patients in the high-risk and low-risk groups of the TCGA training concentration(p< 0.0001),and the area under ROC curve of the model in the training set of one year,three years and five years was0.79,0.83 and 0.91,respectively,indicating good prediction effect.In addition,in order to better evaluate the efficacy of the model,we selected GSE52903 from GEO database as the test set.Kaplan-Meier analysis showed that the model could also distinguish between patients in the high risk group and the low risk group in the test set(p=0.017).Multivariate Cox regression was used to prove the deubiquitination related prognostic model could be used as an independent prognostic factor for cervical cancer.Conclusion1.Deubiquitination is associated with the prognosis of cervical cancer.2.This study integrated a cervical cancer risk score model based on deubiquitination related genes,which is an independent prognostic factor for patients with cervical cancer and has clinical significance in the prognosis assessment of patients with cervical cancer.3.Risk score model is related to T stage and BMI index in TNM stage,in which the later T stage,the higher the score;BMI less than 24kg/m2 has a high score.The risk scoring model was related to immune cell infiltration.The distribution of central memory T cells was more in the high-risk group and activated T cells was more in the low-risk group.