PAIR: paired allelic log-intensity-ratio-based normalization method for SNP-CGH arrays
PAIR: paired allelic log-intensity-ratio-based normalization method for SNP-CGH arrays
复制标题
DOI：
10.1093/bioinformatics/bts683
复制
发表时间：
2013-02-01
期刊：
影响因子：
5.8
通讯作者：
Fang, Zhide
中科院分区：
文献类型：
--
作者：
Yang, Shengping;Pounds, Stanley;Fang, Zhide
动机：标准化在 DNA 拷贝数分析中至关重要。我们提出了一种新方法来正确识别基因组中的双拷贝探针，以获得单核苷酸多态性阵列标准化的代表性参考。该方法基于二态隐马尔可夫模型。与文献中目前大多数可用的方法不同，所提出的方法不需要假设二拷贝状态探针的百分比在基因组中占主导地位，只要确实存在二拷贝探针即可。结果：真实数据分析和仿真研究表明，该算法的成功之处在于：（i）对于具有主导二拷贝状态的样本，其性能与当前方法（例如CGHnormaliter和popLowess）相同，而对于具有较少主导二拷贝状态的样本，其性能优于这些方法； (ii) 它可以识别杂合性的复制中性丢失；结果：真实数据分析和模拟研究表明，该算法是成功的，因为（i）对于具有主导二拷贝状态的样本，它的性能与当前方法（例如 CGHnormaliter 和 popLowess）一样好，并且对于具有不太主导二拷贝状态的样本，其性能优于这些方法； (ii) 它可以识别杂合性的复制中性丢失； (iii) 就所使用的计算时间而言，它是高效的。
Motivation: Normalization is critical in DNA copy number analysis. We propose a new method to correctly identify two-copy probes from the genome to obtain representative references for normalization in single nucleotide polymorphism arrays. The method is based on a two-state Hidden Markov Model. Unlike most currently available methods in the literature, the proposed method does not need to assume that the percentage of two-copy state probes is dominant in the genome, as long as there do exist two-copy probes. Results: The real data analysis and simulation study show that the proposed algorithm is successful in that (i) it performs as well as the current methods (e.g. CGHnormaliter and popLowess) for samples with dominant two-copy states and outperforms these methods for samples with less dominant two-copy states; (ii) it can identify the copy-neutral loss of heterozygosity; and (iii) it is efficient in terms of the computational time used.Results: The real data analysis and simulation study show that the proposed algorithm is successful in that (i) it performs as well as the current methods (e. g. CGHnormaliter and popLowess) for samples with dominant two-copy states and outperforms these methods for samples with less dominant two-copy states; (ii) it can identify the copy-neutral loss of heterozygosity; and (iii) it is efficient in terms of the computational time used.