Identification Of Biomarkers For Breast Cancer Prognosis Based On CeRNA Mechanism

Growing evidence highlights the role of differentially expressed genes as competing endogenous RNAs（ce RNAs）in tumor development and progression.Nonetheless,studies of key biomarkers in ce RNA networks capable of predicting breast cancer（BC）prognosis are still lacking.The ce RNA network construction method based on differential gene interaction pairs not only narrows the research scope of target genes,but also provides key potential biomarkers for evaluating the prognosis of breast cancer patients,which will help us increase our understanding of the involvement of breast cancer patients Understanding of ce RNA mechanisms in early development.The purpose of this study is to identify the potential genes that can predict the overall survival（OS）of breast cancer patients based on the differentially expressed genes screened by the ce RNA mechanism and to construct a corresponding model that can predict the prognostic risk of patients.In this study,the breast cancer Count expression profile data downloaded from the TCGA database were firstly grouped according to breast cancer tissues and paracancerous tissues,and the expression profile data and clinical data of lnc RNA,m RNA and mi RNA were obtained,and the RNA sequencing data were differentiated.Genetic analysis.A breast cancer-related dysregulated ce RNA network was constructed based on differential gene interaction pairs and visualized using Cytoscape software.Then,GO and KEGG enrichment analysis of differential m RNAs in the ce RNA network was performed to obtain their functions and related pathways.The m RNAs that have certain effects in the process of breast cancer and the possible mechanism of action of these biomarkers were comprehensively analyzed.For further exploration,survival analysis was performed on the ce RNA network,and the key DEm RNA and DEmi RNA related to breast cancer were correlated with overall survival（OS）according to Kaplan-Meier and log-rank test,and were significantly correlated with OS of breast cancer patients（p<0.05）.Univariate Cox proportional hazards regression model was used to analyze the lnc RNA,mi RNA and m RNA of Ce RNA network nodes in the TCGA-BRCA cohort.Further Lasso regression analysis was performed for key genes with P<0.05 in univariate Cox regression and TCGA-BRCA cohort samples.A risk scoring formula was established to carry out risk scoring for the training set TCGABRCA cohort,and the cutoff value of the risk score was selected according to the Step Miner dichotomy,and the patients were divided into high and low risk groups.Kaplan-Meier analysis was performed for both high and low risk groups in the validation set（GSE20685）.The ROC curve was drawn to test the efficacy of the model.Finally,to test the independence of the key gene marker model with other clinical variables,Cox univariate and multivariate analyses were performed on samples from the TCGA-BRCA cohort that retained comprehensive prognostic information.Through differential gene analysis of breast cancer Count expression profile data,a total of 167 differentially expressed lnc RNAs（107 up-regulated and 60 downregulated）,169 mi RNAs（96 up-regulated and 73 down-regulated）and 1831 m RNAs（620 1 up-regulated and 1211 down-regulated）.According to the interaction between differential genes,a breast cancer-related ce RNA regulatory network composed of 131 lnc RNAs,11 mi RNAs and 131 m RNAs was drawn using Cytoscape.The results of functional enrichment analysis showed that the differentially expressed m RNAs were mainly enriched in immune-related pathways.According to Kaplan-Meier and log-rank test,5 key DEm RNAs and 3 key DEmi RNAs in breast cancer were significantly associated with overall survival（OS）（P<0.05）.Univariate Cox proportional hazards regression was used to obtain seven genes TP63,BACH2,LIFR,TBC1D4,AK3,TCF7L2,CACNA2D1 associated with breast cancer prognosis.Lasso regression analysis shows that when the number of variables is 6,the root mean square deviation of the model is relatively small,so a 6-gene breast cancer prognosis prediction risk score model can be established: risk score=（-0.062）* TP63 expression value+（-0.035）*LIFR expression value+（-0.060）*TBC1D4 expression value+（-0.0170）*AK3expression value+（-0.078）*TCF7L2 expression value+（-0.059）*CACNA2D1expression value.Survival analysis showed that compared with the low-risk group,the high-risk group had more patients with lower OS,while the low-risk group had a higher survival rate（P<0.05）.The ROC curve was used to test the performance of the model,and the AUC values of 1 year,3 years and 5 years were 0.688,0.611 and 0.594,respectively.The AUC value shows that the model has a certain test efficiency,and its true test ability needs to be further confirmed by clinical cohort studies.Cox univariate and multivariate analysis showed that 6-gene risk score,age and clinical stage were all independent factors that significantly affected OS in breast cancer patients.