三种miRNA对肝癌预后预测的生物信息学分析模型

A Model of Three miRNA on the Prognosis Prediction of Hepatocellular Carcinomaby Bioinformatics Analysis

目的通过TCGA数据库构建一种miRNA预后模型,预测肝癌的预后。方法通过edgeR软件包对miRNA和mRNA的表达数据进行分析和标准化,并与生存率相结合分为两组。通过单因素和多因素Cox回归分析构建3种与总生存率明显差异相关的miRNA,使用ROC曲线评价其可靠性。通过miRDB、TargetScan和miRTarBase 3个数据库筛选出miRNA的靶基因,并运用GO和KEGG探究其相关作用机制。使用Cytoscape3.7.2和String数据库筛选出前10个中枢基因。结果共筛选出47个靶基因与生存预后显著相关;ROC曲线结果显示,三组的AUC分别为0.789、0.730、0.763,说明模型预测HCC患者预后的能力;回归分析显示,该模型可独立影响肿瘤的预后情况,包括T分期、临床分期及远处转移的存在。结论hsa-miR-3682-3p、hsa-miR-9-5p和hsa-miR-139-5p三种miRNA的生物学分析模型可以预测HCC患者的预后,且可以作为HCC中独立的预后因素。

Objective To construct a miRNA prognostic model by TCGA database to predict the prognosis of liver cancer.Methods The expression data of miRNA and mRNA were analyzed and standardized by edgeR software package,and they were divided into two groups combined with survival rate.Single-factor and multi-factor Cox regression analysis was used to construct three miRNAs related to significant differences in overall survival rate,and ROC curve was used to evaluate their reliability.The target genes of miRNA were screened from miRDB,TargetScan and miRTarBase databases,and the related mechanism was explored by GO and KEGG.The top 10 central genes were screened using Cytoscape 3.7.2 and String databases.Results A total of 47 target genes were found to be significantly associated with survival and prognosis.ROC curve showed that the AUC of the three groups were 0.789,0.730 and 0.763,respectively,indicating the ability of the model to predict the prognosis of HCC patients.Regression analysis showed that the model could independently affect the prognosis of tumors,including T staging,clinical staging and distant metastasis.Conclusion The biological analysis models of hsa-miR-3682-3p,hsa-miR-9-5p and hsa-miR-139-5p can predict the prognosis of HCC patients and can be used as independent prognostic factors in HCC.