基于TCGA数据库Wnt相关长链非编码RNA构建肾乳头状细胞癌预后模型  

A novel prognosis model for papillary renal cell carcinoma based on Wnt-associated lncRNAs in TCGA database

作　　者：唐国军 洪余德 赵崇玉 李辽源[1] Tang Guojun;Hong Yude;Zhao Chongyu;Li Liaoyuan(Department of Urology,the Third Affiliated Hospital of Sun Yat-sen University,Guangzhou 510630,China)

机构地区：[1]中山大学附属第三医院泌尿外科,广州510630

出　　处：《中华腔镜泌尿外科杂志（电子版）》2023年第3期270-275,共6页Chinese Journal of Endourology(Electronic Edition)

基　　金：国家自然科学基金(81972419,81772753)。

摘　　要：目的构建一种Wnt通路相关的长链非编码RNA(LncRNAs)的肾乳头状细胞癌(pRCC)预后模型。方法经癌症基因组图谱(TCGA)获取pRCC患者的转录组数据和对应的临床信息。采用共表达网络分析、COX回归和Lasso分析方法筛选出Wnt相关的LncRNA用于构建风险预后模型。此外,通过绘制受试者工作特征(ROC)曲线验证模型的预测性能。结果构建了由5个Wnt相关的LncRNAs(CT66,PRECSIT,AL352984.1,AC124798.1,AC011503.2)组成的肾乳头状细胞癌预后模型。生存分析结果表明,该模型在TCGA训练组(P=0.002)、测试组(P=0.013)和完整组(P<0.001)中均具有较好的有效性。将风险评分作为独立预后指标时,完整组患者5年总生存率的ROC曲线下面积(AUC)为0.708。结论由5个Wnt相关lncRNA构建的新的预后模型有助于评估pRCC患者的预后。Objective To construct a prognostic model of Wnt associated long non-coding RNAs(LncRNAs)in papillary renal cell carcinoma(pRCC).Methods The data for ranscriptome and clinical information of pRCC patients were obtained by The Cancer Genome Atlas(TCGA).Wnt-associated LncRNAs were selected by co-expression network analysis,COX regression,and Lasso analysis to build the risk prognosis model.In addition,the predictive performance of this model is verified by receiver operating characteristic(ROC)curves.Results Five Wnt-associated LncRNAs(CT66,PRECSIT,AL352984.1,AC124798.1,and AC011503.2)were chosen to construct a prognostic model of pRCC.Survival analyses showed that the model have predictive validity in TCGA training group(P=0.002),testing group(P=0.013)and combined group(P<0.001).When the risk score was used as an independent prognostic indicator,the area under ROC curve(AUC)for 5-year overall survival in the combined group was 0.708,indicating acceptable predictive performance.Conclusion The prognosis model including 5 Wnt-associated LncRNAs is helpful to evaluate the prognosis of pRCC patients.

关 键 词：WNT 肾乳头状细胞癌 肾肿瘤 癌症基因组图谱 大数据 生物信息学分析 计算生物学 长链非编码RNA 预后模型 

分 类 号：R737.11[医药卫生—肿瘤]