基于TRPC相关基因构建膀胱癌预测模型  

Construction of bladder cancer prediction model based on TRPC-related genes

作　　者：周汇栋 臧婧蕾 黎汝琦 刘静红 申隆江 郭抗萧 陈涛 Zhou Hui-dong;Zang Jing-lei;Li Rv-qi;Liu Jing-hong;Shen Long-jiang;Chen Tao(Changsha Hospital Affiliated to Hunan Normal University,Changsha 410006,China;Changsha Health Vocational College,Changsha 410100,China;Changsha Central Hospital,Changsha 410018,China)

机构地区：[1]湖南师范大学附属长沙医院,长沙410006 [2]长沙卫生职业学院,长沙410100 [3]长沙市中心医院,长沙410018

出　　处：《湖南师范大学学报（医学版）》2022年第3期201-206,共6页Journal of Hunan Normal University(Medical Sciences)

基　　金：湖南省卫生健康委课题(202204053210)。

摘　　要：目的:通过分析经典瞬时受体电位通道(TRPC)相关基因,构建膀胱癌亚型及相关预后模型,旨在为膀胱癌患者提供较准确的预后判断.方法:从GENECARD获取TRPC相关基因,并从TCGA数据库下载膀胱癌患者的RNAseq数据、临床数据.对TRPC相关基因进行富集分析,运用NMF算法将膀胱癌样本进行肿瘤亚型分组并进行临床信息统计及生存分析.运用LASSO回归筛选构建预后模型,并绘制列线图和进行验证.结果:从GENECARD获取了364个TRPC相关基因,运用NMF算法将BLCA患者分为三组,结果显示,C3患者的预后最差,C2患者的预后最好,且不同分组之间有统计学差异.发现在不同分组中年龄、性别、AJCC分期、分级、T分期、N分期均有统计学差异.通过LASSO回归及交叉验证,最终筛选出6个基因构建相关预后模型,依据计算出的风险值将患者分为高危、低危两组,经统计两组生存有明显差异,且风险值为膀胱癌独立预后因素,最后绘制了列线图及校正曲线.结论:本研究基于TRPC相关基因构建了膀胱癌预测模型,经验证该模型较传统病理及TNM分期有较高准确性,并筛选出影响膀胱癌患者预后的关键基因,可为后续机制研究提供理论基础.Objective Through the analysis of TRPC related genes,bladder cancer subtypes and related prognostic mod-els were constructed in order to provide more accurate prognosis judgment for patients with bladder cancer.Methods TRPC-related genes were obtained from GENECARD,and RNAseq and clinical data of bladder cancer patients were downloaded from TCGA database.TRPC-related genes were enriched and analyzed.Bladder cancer samples were divided into tumor subtypes by NMF algorithm and clinical information statistics and survival analysis were carried out.The prognostic model was constructed by LASSO regression screening,and the line diagram was drawn and verified.Results 364 TRPC related genes were obtained from GENECARD,BLCA patients were divided into three groups by NMF algorithm.The results showed that C3 patients had the worst prognosis and C2 patients had the best prognosis,and there were statistical differences among different groups.It was found that there were statistical differences in age,sex,AJCC stage,grade,T stage and N stage among different groups.Through LASSO regression and cross-validation,six genes were selected to construct the related prognostic model.According to the cal-culated risk value,the patients were divided into high-risk group and low-risk group.There was significant difference in sur-vival between the two groups,and the risk value was an independent prognostic factor for bladder cancer.Finally,the line chart and correction curve were drawn.Conclusion This study constructed a bladder cancer prediction model based on TRPC-re-lated genes,which was proved to be more accurate than traditional pathology and TNM staging,and screened out the key genes that affect the prognosis of bladder cancer patients,which can provide a theoretical basis for follow-up mechanism research.

关 键 词：膀胱癌 肿瘤亚型 预测模型 生物信息学 

分 类 号：R737.14[医药卫生—肿瘤]