Statistical methods for estimation of copy number from next-generation sequencing

描述(由申请人提供)： 这项申请涉及董事会挑战领域(08)基因组学和特定的挑战主题，08-DA-102改进的生物信息学分析，用于深度测序。随着下一代测序技术的进步继续降低测序成本，预计未来几年进行全基因组测序的人类样本数量将大幅增加。除了检测序列变异外，这些数据还可以用来估计DNA拷贝数变异，并随后检查拷贝数与表型之间的相关性。在这项提案中，我们的目标是开发一系列计算步骤和综合分析流水线，以从下一代测序数据中准确估计拷贝数。这涉及对测序数据的有效处理，包括适当的比对程序和对实验伪影的校正。为了估计沿染色体位置的拷贝数，我们将开发新的分割程序，既适用于单个样本，也适用于多个样本，以利用测序数据的特殊性质。重要的是，我们还解决了实验设计中的问题，特别是测序深度(基因组覆盖)和阅读长度对拷贝数分布的分辨率和准确性的影响。我们使用了来自Solexa、Solid和CompleteGenome等多个平台的数据进行研究。本提案中开发的管道将在一个强大的分布式计算系统上实施，并将免费提供给研究界。因此，该项目的结果将能够有效地从全基因组测序数据中提取拷贝数，并将促进下一代测序技术的快速转换，以识别与正常或疾病表型相关的结构变异。

DESCRIPTION (provided by applicant): This application addresses board Challenge Area (08) Genomics and specific challenge topic, 08-DA-102 Improved Bioinformatics Analysis for Deep Sequencing. The number of human samples undergoing whole-genome sequencing is expected to increase dramatically in the next few years, as advances in next-generation sequencing technologies continue to lower the cost of sequencing. In addition to detection of sequence variation, these data can be used to estimate DNA copy number variation and subsequently to examine correlation between copy number and phenotype. In this proposal, we aim to develop a series of computational steps and integrated analysis pipeline for accurate estimation of copy number from next-generation sequencing data. This involves efficient processing of the sequencing data, including appropriate alignment procedures and correction for experiment artifacts. For estimation of the copy number along chromosomal location, we will develop novel segmentation procedures, both for a single sample and for multiple samples, to take advantage of the specific nature of sequencing data. Importantly, we also address issues in experimental design, especially the effect of depth of sequencing (genome coverage) and read length on the resolution and accuracy of copy number profiles. We use data from a number of platforms including Solexa, SOLiD, and CompleteGenomes for our studies. The pipeline developed in this proposal will be implemented on a powerful distributed computing system and will be made available freely to the research community. The results of this project will thus enable efficient extraction of copy number from whole-genome sequencing data and will facilitate rapid translation of next-generation sequencing technology to identify structural variations associated with normal or disease phenotypes.