利用TCGA数据库构建肾透明细胞癌相关miRNA预后模型

高艾,王昕苑,苏依琳,苏龙龙,张建辉,牛晓辰

(山西医科大学,太原 030000)



摘要:

利用TCGA数据库中肾透明细胞癌的miRNA与mRNA数据及临床信息,构建由miRNA组成的预后风险评分模型,并筛选与生存预后相关的miRNA-mRNA调控关系对,为研究提供理论依据。下载并整理TCGA[JP+1]数据库中肾透明细胞癌的miRNA与mRNA数据；对数据进行差异分析,将差异表达的miRNA与临床信息进行合并,利用单因素与多因素Cox回归分析,构建预后模型并进行模型评价；对模型中的miRNA进行靶基因预测,结果与差异表达的mRNA进行取交集,构建miRNA-mRNA调控网络；对网络中的mRNA进行生存分析,筛选生存相关的miRNA-mRNA调控关系对。共得到49个差异表达的miRNA与3 613个差异表达的mRNA；预后模型计算公式为:风险值(risk score)=hsa-miR-21-5p表达量×0.603＋hsa-miR-1251-5p表达量×-0.093；调控网络中共纳入31个miRNA-mRNA调控关系对；对mRNA进行生存分析,共得到7个有价值的关系对。所构建预后模型可有效预测肾透明细胞癌患者生存预后情况,筛选到的miRNA-mRNA调控关系对可为相关研究与治疗提供参考。

关键词:  肾透明细胞癌  miRNA-mRNA  预后模型  TCGA数据库

DOI：10.12113/202008003

分类号:R737.11

文献标识码:A

基金项目:



Exploiting the TCGA database to establish a prognostic model of miRNAs associated with kidney renal clear cell carcinoma

GAO Ai, WANG Xinyuan, SU Yilin, SU Longlong, ZHANG Jianhui, NIU Xiaochen

(Shanxi Medical University, Taiyuan 030000, China)

Abstract:

HJ1.5mm]To provide theoretical basis for further research,a prognostic risk score model of kidney renal clear cell carcinoma was constructed based on the miRNA and mRNA expression profiles and clinical information in the TCGA database, and the miRNA-mRNA regulatory relationships related to survival prognosis were selected. The miRNA and mRNA expression profiles of kidney renal clear cell carcinoma were downloaded from the TCGA database and collated. Differential analysis was performed on the data, and the differentially expressed miRNAs were combined with clinical information. Univariate and multivariate Cox regression analyses were used to construct the prognostic model and conduct model evaluation. Target genes of miRNAs in the model were predicted, and the results were intersected with differentially expressed mRNAs to construct the miRNA-mRNA regulatory network. Survival analysis was performed on the mRNAs in the network to screen the miRNA-mRNA regulatory pairs related to survival. A total of 49 differentially expressed miRNAs and 3 613 differentially expressed mRNAs were obtained and the formula of prognostic model was: risk score=hsa-miR-21-5p expression×0.603＋hsa-miR-1251-5p expression×-0.093. Thirty-one miRNA-mRNA regulatory pairs were included in the network and seven miRNA-mRNA regulatory pairs were selected. The prognostic model can effectively predict the survival and prognosis of patients with kidney renal clear cell carcinoma, and the selected miRNA-mRNA regulatory pairs can provide reference for related research and clinical treatment.

Key words:  Kidney renal clear cell carcinoma  miRNA-mRNA  Prognostic model  TCGA database