基于双基因分析的结肠癌标志基因选择  

Marker Gene Selection of Colon Cancer Based on Two-gene Analysis

作　　者：李建更[1] 高志坤[1] 严志[2] 阮晓钢[1] 

机构地区：[1]北京工业大学人工智能与机器人研究所,北京100124 [2]北京肿瘤分子生物学实验室,北京市肿瘤防治研究所,北京大学临床肿瘤学院,北京100036

出　　处：《中国生物医学工程学报》2009年第5期691-695,共5页Chinese Journal of Biomedical Engineering

基　　金：国家自然科学基金重点项目(60774077);北京市教育委员会发展计划资助项目

摘　　要：针对结肠癌标志基因选择问题,引入一种最高得分对(TSP)方法,处理一组包含40个肿瘤和22个正常样本的结肠癌微阵列数据,得到标志基因对(VIP,DARS)并构建双基因分类器。结果显示,该分类器对62例样本的分类准确率可达93.55%。进一步,利用荧光实时定量PCR在10个独立样本上检验所选基因对的分类效果,正确率为100%,表明该方法能有效地选择结肠癌标志基因,对临床诊断有积极的意义。For the selection of marker gene for colon cancer, a method based on top scoring pair （TSP） was introduced to process a microarray dataset including 40 tumor and 22 normal colon tissue samples. A pair of genes, VIP and DARS, was obtained as marker genes. According to a two-gene classifier constructed by the two genes, the 62 samples were distinguished with the accuracy of 93.55 %. Furthermore, this pair of genes was validated with quantitative real-time PCR on an additional 10 independent samples, and gained the accuracy of 100%. The outcome showed that this method could select marker genes for colon cancer effectively and was promising in clinical diagnosis.

关 键 词：标志基因 结肠癌 双基因分类器 荧光实时定量PCR(Q-RT-PCR) 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] Q617[自动化与计算机技术—控制科学与工程]