Hu Jianhua
Department of Biostatistics, Division of Quantitative Science, University of Texas M.D. Anderson Cancer Center, Houston, TX, USA.
Bioinformatics. 2008 Oct 1;24(19):2193-9. doi: 10.1093/bioinformatics/btn372. Epub 2008 Aug 12.
Microarray experiments can be used to help study the role of chromosomal translocation in cancer development through cancer outlier detection. The aim is to identify genes that are up- or down-regulated in a subset of cancer samples in comparison to normal samples.
We propose a likelihood-based approach which targets detecting the change of point in mean expression intensity in the group of cancer samples. A desirable property of the proposed approach is the availability of theoretical significance-level results. Simulation studies showed that the performance of the proposed approach is appealing in terms of both detection power and false discovery rate. And the real data example also favored the likelihood-based approach in terms of the biological relevance of the results.
R code to implement the proposed method in the statistical package R is available at: http://odin.mdacc.tmc.edu/~jhhu/cod-analysis/.
微阵列实验可用于通过癌症离群值检测来帮助研究染色体易位在癌症发展中的作用。目的是识别与正常样本相比在一部分癌症样本中上调或下调的基因。
我们提出了一种基于似然性的方法，该方法旨在检测癌症样本组中平均表达强度的变化点。所提方法的一个理想特性是可获得理论显著性水平结果。模拟研究表明，所提方法在检测能力和错误发现率方面的表现都很有吸引力。并且实际数据示例在结果的生物学相关性方面也支持基于似然性的方法。
在统计软件包R中实现所提方法的R代码可在以下网址获取：http://odin.mdacc.tmc.edu/~jhhu/cod-analysis/ 。
