Alkodsi Amjad, Louhimo Riku, Hautaniemi Sampsa
Brief Bioinform. 2015 Mar;16(2):242-54. doi: 10.1093/bib/bbu004. Epub 2014 Mar 5.
Somatic copy-number alterations (SCNAs) are an important type of structural variation affecting tumor pathogenesis. Accurate detection of genomic regions with SCNAs is crucial for cancer genomics as these regions contain likely drivers of cancer development. Deep sequencing technology provides single-nucleotide resolution genomic data and is considered one of the best measurement technologies to detect SCNAs. Although several algorithms have been developed to detect SCNAs from whole-genome and whole-exome sequencing data, their relative performance has not been studied. Here, we have compared ten SCNA detection algorithms in both simulated and primary tumor deep sequencing data. In addition, we have evaluated the applicability of exome sequencing data for SCNA detection. Our results show that (i) clear differences exist in sensitivity and specificity between the algorithms, (ii) SCNA detection algorithms are able to identify most of the complex chromosomal alterations and (iii) exome sequencing data are suitable for SCNA detection.
体细胞拷贝数改变（SCNAs）是影响肿瘤发病机制的一种重要结构变异类型。准确检测存在SCNAs的基因组区域对于癌症基因组学至关重要，因为这些区域可能包含癌症发展的驱动因素。深度测序技术可提供单核苷酸分辨率的基因组数据，被认为是检测SCNAs的最佳测量技术之一。尽管已经开发了几种算法来从全基因组和全外显子组测序数据中检测SCNAs，但它们的相对性能尚未得到研究。在此，我们在模拟和原发性肿瘤深度测序数据中比较了十种SCNAs检测算法。此外，我们评估了外显子组测序数据在SCNAs检测中的适用性。我们的结果表明：（i）各算法在敏感性和特异性方面存在明显差异；（ii）SCNAs检测算法能够识别大多数复杂的染色体改变；（iii）外显子组测序数据适用于SCNAs检测。
