利用小样本表达谱数据构造前列腺癌基因网络  

Reconstruction of Large Prostate Cancer Gene Network from Small Size Microarray Data

作　　者：付旭平[1] 曹明[2] 莫文娟[1] 姜梅[1] 李瑶[1] 

机构地区：[1]复旦大学生命科学学院,上海200433 [2]上海市智能信息处理重点实验室,上海200433

出　　处：《复旦学报（自然科学版）》2010年第1期1-8,F0002,共9页Journal of Fudan University：Natural Science

基　　金：国家高技术研究发展计划(863计划)资助项目(2006AA02Z324)

摘　　要：用25张前列腺癌芯片表达谱数据,基于基因间的表达相似性和基因的差异表达程度,筛选出287条基因,用模糊聚类的方法将其聚成65个类,用贝叶斯模型构造了一个大型的基因网络,并对这个网络的有效性进行了验证.结果表明,基于本文提出的策略,利用小样本量的表达谱数据构造大型基因网络是可行的.Based on the identification of differentially expressed genes and the analysis of gene co-expression, 287 genes were chosen from 25 prostate cancer cDNA microarrays. These genes were grouped into 65 clusters with Fuzzy-C-Means method, and then a large gene network was reconstructed with Bayesian model. Some confirmations of the gene network were performed. It is shown that the proposed strategy for reconstruction of a large gene network from small size microarray data is feasible.

关 键 词：基因网络 CDNA芯片 模糊聚类 贝叶斯模型 

分 类 号：Q811.4[生物学—生物工程]