摘要
目的构建基于Scissor算法的肝内胆管癌(intrahepatic cholangiocarcinoma,ICC)新型预后预测模型。方法分别从癌症基因组图谱(TCGA)数据库和美国国立生物技术信息中心基因表达(GEO)数据库下载ICC高通量测序数据集和ICC单细胞数据集(GSE151530)。运用R语言Scissor包筛选与ICC预后相关的细胞,并计算差异表达基因(differentially expressed genes,DEGs)。通过功能富集分析和基因相互作用检索工具(search tool for the retrieval of interacting genes,STRING)对DEGs进行蛋白质-蛋白质相互作用(protein-protein interaction,PPI)网络分析。通过肿瘤免疫功能障碍和排斥(tumor immune dysfunction and exclusion,TIDE)算法推断肿瘤免疫逃逸评分,并计算TIDE评分与DEGs的相关性。最后通过多因素Cox回归分析筛选得到关键基因并构建ICC新型预后预测模型。结果筛选得到肿瘤微环境中的604个与ICC不良预后相关的细胞并进一步明确了366个DEGs,包括281个上调基因和85个下调基因(P<0.05,|Fold change|>1.5),主要富集于程序性死亡受体1信号通路、白细胞介素10信号通路和肿瘤坏死因子信号通路等。STRING网络分析构建了4个模块化PPI网络,通过计算各模块基因与病人TIDE评分的相关性,共鉴定出18个与免疫治疗密切相关的关键基因(P<0.05),最后使用多因素Cox回归分析确定了3个既与ICC免疫响应相关又与其预后相关的枢纽基因CDK1、FCGR2A和CTSD,并基于此构建了预后预测模型[1年生存曲线下面积(area under the curve,AUC)为0.672,2年生存AUC为0.692,3年生存AUC为0.742],同时于外部数据集验证了该模型效果(1、2、3年生存AUC分别为0.584、0.651、0.668)。结论基于CDK1、FCGR2A和CTSD建立的预后预测模型具备良好性能,并可为ICC病人能否获益于免疫治疗提供重要参考价值。
Objective To construct a novel prognostic prediction model for intrahepatic cholangiocarcinoma(ICC)based upon Scissor algorithm.Methods A high-throughout bulk sequencing dataset of ICC was downloaded from the database of Cancer Genome Atlas(TCGA)and a high-quality single cell dataset acquired from the gene expression omnibus(GEO)database of National Center for Biotechnology Information Carcinoma.In R software,Scissor package was employed for selecting cell subpopulations related to the prognosis in tumor microenvironment of ICC and the differentially expressed genes(DEGs)were identified through the FindMarkers function in the Seurat package.Then ClusterProfiler was utilized for enrichment analysis based upon gene ontology(GO),Kyoto Encyclopedia of Genes and Genomes(KEGG),Hallmark and Reactome.By applying STRING,protein-protein interaction(PPI)network of DEGs were analyzed and further plotted by Cytoscape software.The immune escape score of each patient tumor in the TCGA dataset was estimated by tumor immune dysfunction and exclusion(TIDE)tool and the correlation of individual DEGs with TIDE score was calculated.Finally key genes were screened and their prognostic impact on ICC was verified in the literature dataset(DOI:10.1016/j.ccell.2021.12.006).Finally the key genes were screened out by multi-factor Cox regression analysis and a prediction model was constructed.Results A total of 604 cells associated with adverse prognosis of ICC were screened and 366 DEGs were further identified,including 281 up-regulated genes and 85 down-regulated genes(P<0.05,|Fold change|>1.5),which were mainly enriched in PD-1,IL-10 and tumor necrosis factor signaling pathways.STRING network analysis constructed 4 modular PPI networks.Through calculating the correlation between each module gene and patient TIDE score,a total of 18 key genes closely related to immunotherapy were identified(P<0.05),three pivotal genes CDK1,FCGR2A and CTSD were identified by multivariate Cox regression analysis.And a prognostic prediction model(1-year survival AUC=0.672,2-year survival AUC=0.692,3-year survival AUC=0.742)was constructed based upon multivariate Cox regression analysis.And model effect was verified on external datasets(1-year survival AUC=0.584,2-year survival AUC=0.651,3-year survival AUC=0.668).Conclusion The prognostic prediction models based upon CDK1,FCGR2A and CTSD have good performance and can provide important references for ICC patients benefiting from immunotherapy.
作者
宋一粟
张惠忠
疏文志
苏仁义
魏绪勇
徐骁
Song Yisu;Zhang Huizhong;Shu Wenzhi;Su Renyi;Wei Xuyong;Xu Xiao(Department of Hepatobiliary&Pancreatic Surgery,Affiliated First Municipal People′s Hospital,Zhejiang University School of Medicine,Zhejiang Hangzhou 310006,China;Key laboratory of Integrated Oncology&Intelligent Medicine of Zhejiang Province,Zhejiang Hangzhou 310006,China;Zhejiang University School of Medicine,Zhejiang Hangzhou 310058,China;Department of Hepatobiliary&Pancreatogastric Surgery,Jinhua Guangfu Cancer Hospital,Zhejiang Jinhua 321111,China)
出处
《腹部外科》
2023年第5期384-392,共9页
Journal of Abdominal Surgery
基金
国家自然科学基金重大研究计划(92159202)。
关键词
肝内胆管癌
单细胞转录组学
免疫治疗
预测模型
Intrahepatic cholangiocarcinoma
Single-cell transcriptomics
Immunotherapy
Predictive model
