基于Notch通路相关基因结肠癌预后模型的构建及验证  

Construction and validation of colon cancer prognosis model based on Notch pathway related genes

作　　者：王俊杰 詹雪冰 罗倩 况云舒 陶香香[1] 朱晓群[1] 梁箫 赵同洋 陈冰[1] Wang Junjie;Zhan Xuebing;Luo Qian;Kuang Yunshu;Tao Xiangxiang;Zhu Xiaoqun;Liang Xiao;Zhao Tongyang;Chen Bing(Department of Pathology,Wannan Medical College,Wuhu 241002,Anhui,China)

机构地区：[1]皖南医学院病理学教研室,安徽芜湖241002

出　　处：《右江民族医学院学报》2022年第3期373-381,共9页Journal of Youjiang Medical University for Nationalities

基　　金：安徽省卫生健康委科研项目(AHWJ2021b059);国家级大学生创新创业训练项目(202110368030,202110368068);活性生物大分子研究安徽省重点实验室自主研究课题(LAB202005)。

摘　　要：目的构建一个基于Notch通路相关基因的结肠腺癌预后模型并通过列线图绘制进行验证。方法下载TCGA结肠腺癌患者的mRNA表达量和临床病理资料。GSEA(gene set enrichment analysis)分析Notch信号通路相关基因集,筛选癌与癌旁组织中差异表达的基因。单因素与多因素Cox比例回归模型进行预后相关mRNAs的筛选,并构建基于mRNAs表达谱的预后模型和列线图,通过生存分析C-index、ROC曲线和校准曲线评估其预测价值。从GEO数据库中下载验证队列GSE29621,对预后模型预测患者预后的有效性进行验证。使用在线网站人类蛋白图谱(HPA)对预后模型内基因进行蛋白表达情况验证。结果从TCGA中下载444例结肠腺癌患者mRNA表达数据和临床病理资料,排除临床资料不全的患者信息,最终纳入385例患者的信息作为研究对象。从GSEA中3个Notch信号通路基因集中获取Notch相关基因,并与TCGA表达数据结合,得到Notch相关基因的表达量,最后进行Notch相关基因的差异基因的筛选。使用GSEA在结肠腺癌mRNAs表达谱中筛选出391个Notch信号通路相关差异表达基因(P<0.05)。利用单变量Cox回归分析筛选出14个结肠腺癌预后相关Notch通路基因,进一步多变量Cox回归分析构建出5种mRNAs(CDHR2,KRT8P12,NEURL1B,SELE,FSTL3)组成的预后模型。ROC曲线和生存分析显示,高Notch通路相关基因风险评分与较差的生存结果显著相关(AUC=0.748,P<0.05)。Notch通路相关的基因评分被证明是一个独立的预后因素。构建了具有临床病理特征和Notch通路相关基因评分的列线图,进一步预测结肠腺癌患者的预后,在C-index、ROC曲线和校准曲线上也表现良好(C-index=0.794,AUC=0.969)。结论构建的包含5个Notch信号通路相关基因的结肠腺癌预后模型具有良好的预后预测效果,有望作为评估结肠腺癌患者预后的指标。Objective To construct a prognostic model of colonic adenocarcinoma based on Notch pathway-related genes and verify it by nomogram.Methods The mRNA expression and clinicopathological data of patients with TCGA colonic adenocarcinoma were downloaded.GSEA(gene set enrichment analysis)was used to analyze the gene set related to Notch signaling pathway and screen differentially expressed genes in cancer and adjacent tissues.Univariate and multivariate Cox proportional regression models were used to screen prognosis related mRNAs,and the prognostic model and nomogram was constructed based on mRNAs expression profile.The survival analysis was applied to evaluate predictive value of C-index,ROC curve and calibration curve.The validation queue GSE29621 was downloaded from GEO database to verify the effectiveness of prognostic model in predicting patient prognosis.The protein expression of genes in the prognostic model was verified using the online Human Protein Atlas(HPA).Results The mRNA expression data and clinicopathological data of 444 patients with colonic adenocarcinoma were downloaded from TCGA.Patients with incomplete clinical data were excluded,and 385 patients were finally included as the study subjects.Multiple Notch-related genes were extracted from three Notch signaling pathway genes in GSEA and combined with TCGA expression data to obtain the expression level of Notch-related genes.Finally,differential genes of Notch related genes were screened.GSEA was used to screen out 391 differentially expressed genes related to Notch signaling pathway in mRNAs expression profile of colonic adenocarcinoma(P<0.05).Univariate Cox regression analysis was used to screen 14 prognostic Notch pathway genes in colonic adenocarcinoma.Multivariate Cox regression analysis was further used to construct prognostic models of 5 mRNAs(CDHR2,KRT8P12,NEURL1B,SELE,FSTL3).ROC curve and survival analysis showed that high Notch pathway related gene risk score was significantly associated with poor survival outcome(AUC=0.748,P<0.05).Notch pa

关 键 词：临床预后模型 结肠腺癌 NOTCH信号通路 列线图 

分 类 号：R735.35[医药卫生—肿瘤]