Risk Evaluation Of The Poor Prognosis In Breast Cancer Based On Bioinformatics

Objective:Breast cancer(BC)is one of the most common cancers and the leading cause of cancer death in women worldwide.Breast cancer is common in women,accounting for 18%of all cancers in women,with a median age of 64 years,according to one million new cancer cases reported worldwide.By 2021,the incidence of breast cancer is expected to increase to 85 cases per 100,000 women.Therefore,it is important to search for biomarkers that play a key role in the prognosis of BC.Based on bioinformatics analysis,this paper is of great significance to search for molecular markers associated with poor prognosis of BC and to evaluate its prognostic risk.Methods:BC-related data sets were downloaded from Gene Expression Omnibus(GEO),and differential Gene analysis was performed using GEO2R.Then,the Venn graph software was used to obtain the common differential Gene Expression(DEG)in the above three data sets.These DEG were then analyzed using the An NOTated,Visualized,and Comprehensive Discovery Database(DAVID),including Molecular Function(MF),Cell Composition(CC),Biological Processes(BP),and the Kyoto Protocol Encyclopedia of Genes and Genomes(KEGG)pathway;Establish a protein-protein interaction(PPI)network and then use cell-based MCODE(molecular complex detection)for other analyses of DEG to identify some core genes.In addition,these core DEG were imported into the Kaplan Meier plotter online database,and Kaplan--Meier was used to analyze the overall survival of these genes to obtain important prognostic information(P<0.05);We further verified DEG expression between BC cancer tissues and normal breast tissues by gene expression profile interaction analysis(GEPIA)(P<0.05).Finally,the differential genes obtained in the previous step were reanalyzed and enriched by KEGG pathway.Results:1.Data downloaded from GEO in this study included 28 normal tissue samples and 180 BC tissue samples.A total of 185 common DEG were identified by differential analysis,including 118 down-regulated genes in BC tissues(log₂FC<-1.5)and 67 upregulated genes(log₂FC>1.5).2.A total of 185 DEG differentially expressed genes were imported into the DEG PPI network complex using the String database and Cytoscape software,which contained 137 nodes and 1156 edges,including 74 down-regulated genes and 63up-regulated genes.Further analysis using cellular MCODE showed that 43 central nodes were identified out of 137 nodes,and these central nodes were all up-regulated genes.3.Survival data of 43 core genes were identified by Kaplan Meier plotter.The survival rate of 31 genes was significantly reduced,while the survival rate of 12 genes was not significantly different(P<0.0001).Then,GEPIA was used to mine the expression levels of 31 genes between cancer and normal people.The results reported that 30 of 31 genes in BC samples were overexpressed compared with normal breast tissue samples(P<0.05).4.In order to understand the possible pathways of the 29 selected DEG,the enrichment of KEGG pathway was re-analyzed by David(P?<0.05).The results showed that four genes(AURKA,CDC20,PTTG1 and CCNA2)were significantly enriched in the cell cycle pathway(P<0.05).5.The data downloaded from TCGA includes 113 normal tissues and 1109 BC tissue samples.According to the above experiments,4 genes with significant prognosis were obtained.Cox multivariate regression analysis was performed on the above genes,and a linear prediction model covering 4 genes was obtained.Risk score==0.431×Exp_CCNA2+0.342×Exp_AURKA+0.632×Exp_PTTG1.The ROC curve area AUC=0.803,indicating that the 4-gene model has good sensitivity and specificity for the prognosis of BC patients.6.Use Human Protein Alts to verify the protein level expression of 4 genes in breast cancer.Conclusion:1.Based on three different microarray datasets,four DEG(AURKA,CDC20,PTTG1,and CCNA2)were identified between BC tissue and normal breast tissue.The results suggest that these four genes may play a key role in the process of BC.