基于多元方差分析的成对肿瘤SNP array数据分段算法
首发时间:2013-08-27
摘要:单核苷酸多态性微阵列(SNP array)技术是近年来获得快速发展的一种高通量生物芯片技术,可以有效地对肿瘤细胞中的染色体变异进行检测。本文针对癌症药物治疗前后肿瘤的染色体变异的成对SNP array数据,提出了一种基于多元方差分析二维统计量的全新染色体异常区域分段算法。对模拟SNP array数据的测试表明,该算法可以精确地将成对肿瘤数据异常区域进行分段,其结果明显好于与现有的CBS算法。同时,ROC性能曲线分析显示本文算法具有较好的抗噪性能。对赫赛汀治疗前后的成对乳腺癌SNP array数据的分析结果显示,该算法可准确地检测出重要致癌基因ERBB2在治疗前后的拷贝数变化。上述结果表明这种基于多元方差分析的算法是一种有效的SNP array数据分析工具。
关键词: SNP array 多元方差分析 分段算法 肿瘤 生物信息学
For information in English, please click here
paired tumors SNP array data segmentation algorithm based on MANOVA
Abstract:As a high-throughput biochip, single nucleotide polymorphism microarray (SNP array) technique is rapidly developed in recent years, which makes it possible to effectively detect the tumor chromosome aberrations. In this article, based on multivariate analysis of variance (MANOVA), we proposed a novel chromosome aberration region segmentation algorithm focusing on the pre- and post- treatment paired tumor SNP array data. Evaluation on simulation SNP array data shows that the proposed algorithm segments the alteration regions in tumor sample with higher accuracy, when compared with current CBS algorithm. Besides, analysis of ROC curve illustrates that proposed algorithm possesses a high robustness against noise. Results on breast cancer samples, which are obtained from pre- and post- treatment using Herceptin, demonstrate new algorithm can precisely detect the difference of copy numbers in the oncogene ERBB2 before and after treatment. All above results indicate this MANOVA-based algorithm is an effective SNP array data analysis tool.
Keywords: SNP array MANOVA segmentation algorithm tumor bioinformatics
基金:
论文图表:
引用
No.****
同行评议
共计0人参与
勘误表
基于多元方差分析的成对肿瘤SNP array数据分段算法
评论
全部评论0/1000