Identification A MiRNA Signature As Prognostic Biomarker Of NSCLC By Bioinformatics Analysis

Objective: A large number of studies have proved that the expression level of mi RNA in tissue and circulation may play a crucial role in tumorigenesis.The aim of our study was to screen some mi RNA signatures as Prognostic biomarkers of NSCLC by bioinformatics analysis and to analyze biological functions of target genes of mi RNA.Method: The mi RNA sequencing dataset and corresponding clinical parameters of NSCLC were obtained from TCGA(961 tumor samples and 90 normal samples).The mi RNA differentially expressed in normal tissues and tumor tissues was obtained through the "edge R" package of R language.A total of 906 tumor samples were included according to the screening criteria.All samples data were randomly divided into the training set(454)and the validation set(452).In the training set,Univariate cox regression analysis was used to screen prognostic mi RNA related of NSCLC(P<0.05)and multivariate regression analysis was used to build of risk model.K-M survival analysis and ROC curve analysis were performed in each group(the training set,the validation set and the whole data set)to verify the prediction accuracy of the risk model.Multi-indicator ROC curve analysis and independent prognostic analysis and clinical correlation analysis were performed to evaluate the prognostic value of the prognostic signature in the whole data set.Target genes of mi RNA were predicted using online database which inclusive of Target Scan,mi RDB and mi RTar Base.The regulation network of mi RNA-m RNA was established by Cytoscape 3.6.1 software.GO enrichment analysis and KEGG signaling pathway analysis were used to assess the biological function of mi RNA target genes.Result: A total of 387 differentially expressed mi RNAs were detected.In the training set,Cox regression analysis showed that risk models constructed by(hsa-mi R-615-3p;hsa-mi R-556-3p;has-mi R-1293;hsa-mi R-4661-5p;hsa-mi R-99a-3p;hsa-mi R-215-5p;hsa-mi R-187-3p)to predicting prognosis of NSCLC.The samples were divided into high and low risk groups according to the median risk score.The K-M curve analysis of the three data sets was statistically significant.The overall survival rate of the high-risk group was lower than that the low-risk group(the training set 3 year K-M: HR:1.736 95%CI=1.357-2.222,p<0.001;the training set 5 year K-M:HR:1.814 95%CI=1.390-2.369,p<0.001;the validation set 3 year K-M: HR:1.61795%CI=1.267-2.064,p < 0.001,the validation set 5 year K-M: HR:1.69795%CI=1.298-2.217,p < 0.001;all data set 3 year K-M: HR:1.69995%CI=1.429-2.020,p < 0.001,all data set 5 year K-M: HR:1.79195%CI=1.484-2.161,p<0.001).The ROC analysis results(the training set: 3year AUC=0.651,5year AUC=0.650;the validation set: 3year AUC=0.619;5year AUC=0.589;all data sets:3year AUC=0.635,5year AUC=0.617)The assessment of the survival rate of NSCLC indicated that the prediction results of all three groups were of high accuracy.It could be used as an independent prognostic factor(HR:1.761 95%CI=1.514-2.050;p<0.001).Compare the accuracy of risk model with other clinical traits in predicting the prognosis of NSCLC by Multi-indicator ROC.An area under the curve(AUC)of risk model(AUC=0.888)、gender(AUC=0.416)、T(AUC=0.577)、M(AUC=0.519)、N(AUC=0.334).The results of clinical correlation analysis showed that hsa-mi R-1293,hsa-mi R-187-3p,hsa-mi R-615-3p,hsa-mi R-4661-5p,hsa-mi R-215-5p in the risk model was associated with clinical stage p<0.05.Hsa-mir-615-3p,hsa-mir-1293,and hsa-mir-187-3p were associated with tumor size p<0.05,and hsa-mir-215-5p was associated with lymph nodes and metastasis p<0.05.The target genes of mi RNA in the model have total of 201 m RNA were predicted.These functions were significantly enriched in multiple biological processes and pathways,including cell proliferation and cell migration regulation.Moreover,ACSL4,APLN,PXMP4 and TMEM164 were correlated with survival state.The top 12 hub genes(ADCY9,CNR1,ADRB2,CXCL2,ITPR1,PPARGC1 A,ACSL1,CACNA1 C,NMUR1,KAT2 B,CCL21,FCGR3A)in protein protein interaction(PPI)network were screened by string database and Cytoscape 3.6.1.Conclusion: our study found that the mi RNA risk model can be used as a prognostic biomarker for NSCLC.