基于TCGA数据筛选潜在的宫颈癌miRNA预后标志物

Screening potential prognostic markers of cervical cancer miRNA based on TCGA data

目的 基于TCGA数据库中miRNA表达量及临床资料,寻找与宫颈癌患者生存相关的microRNA(miRNA or miR)生物标记物,构建模型以预测宫颈癌患者的生存期。方法 从TCGA数据库中提取出宫颈癌样本的miRNA表达及患者临床信息完整的数据,将样本随机分为训练集和验证集,通过统计学及生物信息学方法(Kaplan-Meier生存曲线、单因素Cox回归分析、Lasso回归、logRanK检验,多因素Cox回归建模,受试者工作特征曲线检验)筛选出与宫颈癌预后相关的miRNA。结果 在训练集中,通过单因素Cox生存分析,Lasso回归压缩建模得出5个miRNA与宫颈癌预后有关(P<0.05),多因素Cox回归分析表明在训练集和验证集中该模型都属于独立预后因素(P<0.05),ROC曲线表明该模型能够对预后做良好预判。基因本体论,通路,疾病富集分析都表明该模型中的miRNA标志物靶基因富集在与肿瘤相关的生物过程,通路,疾病上。结论 本研究构建的宫颈癌患者miRNA标志物预后模型,具有较好的可信性和特异性,对miRNA在宫颈癌预后判断具有较高借鉴作用。

Objective To find the microRNA biomarkers related to the survival of patients with cervical cancer based on the expression of miRNA in TCGA database and clinical data, and to construct a model to predict the survival time of patients with cervical cancer.Methods The miRNA expression of cervical cancer samples and the data of complete clinical information of patients were extracted from TCGA database, and the samples were randomly divided into training set and verification set.The miRNA related to the prognosis of cervical cancer were screened by statistical and bioinformatics methods(Kaplan-Meier survival curve, univariate Cox regression analysis, Lasso regression, logRanK test, multivariate Cox regression modeling, receiver working characteristic curve test).Results In the training set, through univariate Cox survival analysis and Lasso regression compression modeling, five miRNA were related to the prognosis of cervical cancer(P<0.05).Multivariate Cox regression analysis showed that the model was an independent prognostic factor in both the training group and the verification group(P<0.05).ROC curve showed that the model could predict the prognosis well.Gene ontology, pathway and disease enrichment analysis all show that the target genes of miRNA markers in this model are enriched in tumor-related biological processes, pathways and diseases.Conclusions The prognostic model of miRNA markers in patients with cervical cancer constructed in this study has good credibility and specificity, and can be used as a reference for miRNA in judging the prognosis of cervical cancer.