基于TCGA、GEO数据库联合挖掘结肠癌诊断和预后生物标志物  

Mining Diagnostic and Prognostic Biomarkers of Colon Cancer Based on TCGA and GEO Databases

作　　者：黄睿 安随虎 张泽娟 高亚珍 麻江涛 李鹏达 刘金英[1] HUANG Rui;AN Suihu;ZHANG Zejuan;GAO Yazhen;MA Jiangtao;LI Pengda;LIU Jinying(Key Laboratory of Environmental ecology and population health in northwest minority areas,Northwest Minzu University,Lanzhou 730030,China)

机构地区：[1]西北民族大学,西北民族地区环境生态与人群健康国家民委重点实验室,甘肃兰州730030

出　　处：《西北民族大学学报（自然科学版）》2024年第2期37-45,共9页Journal of Northwest Minzu University(Natural Science)

基　　金：甘肃省教育厅教育科技创新项目(2022B-072);甘肃省自然科学基金项目(22JR5Ra189);中央高校专项资助项目(31920220112)。

摘　　要：目的:应用GEO和TCGA数据库中的结肠癌基因表达数据,通过生物信息学手段挖掘影响结肠癌诊断和预后的关键基因,以期找到结肠癌早期诊断及预后相关的生物标志物,为结肠癌的临床诊断及结肠癌药物研发提供分子基础.方法:检索GEO基因表达数据库,下载符合条件的结肠癌患者基因芯片数据集,同时下载TCGA中结肠癌的转录组测序数据.应用R语言及相应的R包进行数据处理及统计学分析.首先通过R语言将下载的GEO基因芯片数据集和TCGA转录组数据集整理为基因表达矩阵,应用“limma”包进行基因表达差异分析,找出每个数据集的差异表达基因;再利用RRA包对这些差异表达基因进行优先级排序,找出在这些数据集中共同上调和共同下调的基因以及根据其变化倍数进行的优先级排序表.同时对这些共同上调和下调基因进行GO和KEGG信号通路富集分析,根据TCGA中的病人临床信息对上调和下调中排名前20的基因进行生存相关分析及蛋白质-蛋白质相互作用分析.结果:从TCGA结肠癌数据集、GSE44861数据集、GSE33113数据集、GSE39582数据集中分别筛选出差异基因4002,349,8917,1697个,接着使用RRA算法得到85个共同显著性上调基因和141个共同显著性下调基因,这些共同上调基因和共同下调基因参与多种信号通路,在肿瘤中的变化倍数高达1000倍,其中有多种下调基因参与金属离子的代谢.还发现GUCA2A,CA4,GCG,CXCL1,CXCL8,CLCA1六个基因的表达水平与生存时间之间存在显著性正相关性.结论:CLCA1,GUCA2A,GCG,CXCL1和CXCL8与结肠癌的发展和预后可能相关.Objective:The colon cancer gene expression data in GEO and TCGA databases to mine the key genes affecting the diagnosis and prognosis of colon cancer by bioinformatics means so as to provide biomarkers related to the early diagnosis or prognosis of colon cancer,and provide molecular basis for the clinical diagnosis of colon cancer,also to improve researching and developing new drugs for colon cancer.Methods:Retrieved the GEO gene expression database,downloaded the qualified gene chip data set of colon cancer patients and the transcriptome sequencing data of colon cancer in TCGA.The R language and corresponding R packages are used for data processing and statistical analysis.Firstly,the downloaded GEO gene chip dataset and TCGA transcriptome dataset were organized into gene expression matrix by R language.Then used the“limma”package to perform gene expression differential analysis to find the differentially expressed genes in each dataset;then used the“Robust Rank Aggregation(RRA)”package to prioritize these differentially expressed genes to find common genes in these datasets.A table of co-up-regulated and co-down-regulated genes and their prioritization according to their fold change.At the same time,the researchers performed GO and KEGG signaling pathway enrichment analysis for these co-up-regulated and down-regulated genes,and performed survival correlation analysis and protein-protein interaction analysis for the top 20 genes based on patient clinical information in TCGA.Results:The researchers downloaded the TCGA colon cancer dataset,GSE44861 dataset,GSE33113 dataset,and GSE39582 dataset,and screened out 4002,349,8917,and 1697 differential genes,and then used the RRA algorithm to get 85 co-significantly up-regulated genes and 141 co-down-regulated genes.The researchers found that these genes were involved in a variety of signaling pathways,and the fold change in tumors was as high as 100 times.Among them,a variety of down-regulated genes were involved in metal ion metabolism.In addition,the researche

关 键 词：结肠癌 GEO数据库 TCGA数据库 生物信息学 

分 类 号：R735.35[医药卫生—肿瘤]