置换检验在微阵列数据分析中的应用.doc

置换检验在微阵列数据分析中的应用作者简介: 王婷(1984-),女,山西省永济市人,研究实****员,主要从事微阵列数据分析和贝叶斯统计研究,发表论文10余篇。电话:(05616)85748456
通迅作者:曾平,E-mail: ******@xzmc.
王婷,曾平
(徐州医学院公共卫生学院,徐州 221002)
摘要:目的研究基于置换检验的微阵列数据多重假设检验和错误发现率的估计。方法采用置换检验获得统计量和p值,采用插入估计方法和微阵列显著性分析方法估计错误发现率,采用bootstrap抽样估计零假设的比例,Hedenfalk的乳腺癌微阵列被用来展示本文的方法。结果如选择作为拒绝域,;微阵列显著性分析方法中△=,。结论置换检验适合于微阵列数据,能够用来进行错误发现率控制和估计。
关键词:微阵列数据;多重假设检验;错误发现率;置换检验;乳腺癌
中图分类号: 文献标识码:A
Application of Permutation Test in Microarray Data Analysis
WANG Ting, ZENG Ping*
(School of Public Health, Xuzhou medical college, Xuzhou 221002, China)
Abstract:Objective To investigate the multiple hypotheses testing in microarray data based on permutation test and estimate the false discovery rate (FDR). Methods The original statistics and p values were obtained via permutation test, then FDR was estimated using plug-in method and significance analysis of microarrays (SAM), bootstrap sampling was applied to estimate the proportion of null hypotheses, Hedenfalk′s breast cancer microarray data was illustrated as an example. Results The plug-in estimated FDR was if the rejection region was specified as, SAM yielded an estimated FDR of under △ equal and the proportion of null hypotheses was estimated as . Conclusion Permutation test can be applied to microarray data analysis for controlling and estimating f