荧光交换标记法基因表达芯片的数据挖掘  

Data Mining of Dye-Swap Labeled Gene Expression Microarrays

作　　者：廖之君[1] 马文丽[1] 梁爽[1] 刘华[1] 张海燕[1] 陈数珍[1] 郑文岭[1] 

机构地区：[1]南方医科大学基因工程研究所

出　　处：《福建医科大学学报》2008年第3期236-239,共4页Journal of Fujian Medical University

基　　金：广东省生物芯片重点实验室基金资助项目(2004B60144)

摘　　要：目的探讨荧光染料交换标记设计的基因芯片数据挖掘方法,并对低剂量电离辐射影响人成纤维细胞基因表达谱数据进行分析。方法应用GeneSifter在线软件和Panther生物学信息数据库,对下载于NCBI的GEO数据库的8个样品GSM(包含4个时间点),选择正确的参数设置上载数据,运用ANOVA方法进行数据挖掘,并对差异表达基因进行功能归类分析。结果获得203条差异表达基因,合并相同基因名后为176条基因。双向聚类和主成分分析发现,样品的24 h时间点基因表达谱与前3个时间点有显著差异,功能归类分析提示,多个生物通路如细胞周期、核酸代谢、DNA代谢等被显著激活。结论应用这种方法可以挖掘荧光交换标记的微阵列数据,低剂量电离辐射对人成纤维细胞基因表达有时间累积效应,可能引起DNA损伤、细胞周期阻滞等变化,诱导细胞凋亡。Objective To explore the methodology in analysis of data from dye-swap labeled microarray, and to profile the gene expression spectrum in human diploid fibroblasts exposed to low dose ionizing radiation. Methods GeneSifter suite and Panther database were used to analyze 8 GSM data （4 time points） downloaded from GEO data repository at NCBI. After parameter optimization, ANOVA was applied to perform data mining. Differentially expressed genes were also classified functionally. Results A total of 203 differentially expressed genes were obtained. The number was reduced to 176 after pruning away the redundant gene names. Two-way clustering and principal component analysis showed that gene expression profile at 24-hour time point differed significantly from 3 other time points. Functional classification of these genes showed that several biological processes such as cell cycle, nucleic acid metabolism and DNA metabolism, were significantly activated by irradiation. Conclusion This method can be applied to mine dye-swap-labeled microarray data. Low dose ionizing radiation might exert time-cumulative effects on gene expression profiles in human fibroblasts, implicating a mechanism by DNA damage, cell cycle arrest, and eventually apoptosis.

关 键 词：荧光 成纤维细胞 辐射 电离 芯片分析技术 自动数据处理 

分 类 号：R346[医药卫生—基础医学]