Four Novel MiRNAs Predict Survival In Pancreatic Cancer Using Bioinformatics Analysis

Aims:A great quantities of miRNAs were significantly up-regulated or down-regulated in pancreatic cancer with many of which having been proved relevant with the tumorigenesis,progression and metastasis.Our study aimed to distinguish the differentially expressed miRNAs between the pancreatic cancer and normal pancreatic tissues by analyzing the TCGA database.Methods:The raw miRNA sequencing data and clinical information were downloaded from TCGA database(https://cancergenome.nih.gov/).182 samples consisting of 178 tumor tissues and 4 matched normal tissues were included in our research through the standard that the samples should have both the miRNA sequencing data and complete clinical information.The processing of the miRNA sequencing data and the analysis of the two group tissues were through limma package in R language.The differential miRNAs were obtained by the filtering standard P<0.05 and log2|FC|>1.The Kaplan-Meier curve and Log-rank method were used to analyze the survival and the differential miRNAs by Graphpad software.Then significant miRNAs were appraised by the univariate and multivariate Cox regression method by SPSS.The t-test was used to compare the clinical features and the independent miRNAs by SPSS.Three websites comprising TargetScan(http://www.targetscan.org/),miRDB(http://www.mirdb.org/miRDB/)and Diana(http://diana.imis.athena-innovation.gr/)were used to predict the targets of the significant miRNAs.In order to increase the reliability,the overlapped target were identified by the Venn diagram.The functional enrichment analysis including Gene Ontology(GO)and Kyoto Encyclopedia of Genes and Genomes(KEGG)pathway were performed by The Database for Annotation,Visualization and Integrated Discovery(DAVID)bioinformatics tool(https://david.ncifcrf.gov/)with the cut-off criteria(gene counts>4 and P<0.05).STRING(a database of known and predicted protein interactions)was used to predict protein interactions among the targets.Firstly,the 473 targets were submitted to the STRING website to get PPI data.Then,the PPIs,with median confidence scores greater than 0.4,were selected for constructing PPI networks by cytoscape.Hub-proteins,the important nodes for the protein interactions in PPI network,are a small number of proteins(hubs)that have many interaction partners.In this study,we use the MCC in the cytoHubba to present the top 50 hub genes.Results:25 miRNAs containing 4 up-regulated and 21 down-regulated miRNAs were obtained by setting the cut-off criteria(P<0.05 and log2|FC|>1.0)and we identified four new miRNAs(mima-342,mima-424,mirna-766,mima-3613)which were closely associated with the survival of the pancreatic cancer patients through the Kaplan-Meier and survival method and log-rank test.What's more,the univariate and multivariate Cox regression model revealed that three miRNAs including mirna-424,mirna-766 and mirna-3613 were all independent prognostic factor of pancreatic cancer.The functional enrichment analysis of the predicted targets of the 4 miRNAs implied that these novel miRNAs might be related with the p53 signaling pathway,the ubiquitin mediated proteolysis,cell cycle apoptosis and many pathways in cancer.The string database and cytoscape software were used to construct the PPI network and identify the top 50 hub genes.Conclusion:Taken together,our research pointed out four novel miRNAs that could serve as the prognostic markers in pancreatic cancer.