Sequence-resolved structural variation of human genomes

要了解人类疾病的遗传基础，需要对全谱进行全面评估 人类的基因变异。基因组结构变异，包括更大的缺失、插入和倒置 (&gt；50个碱基)，由于与重复DNA有关，因此更难表征。大多数人 结构变异，包括常见的结构变异或SVS，还没有使用短读技术发现 全基因组数据集和标准SV呼叫者。过去三年测序技术的进展 然而，多年以来，这种变异的系统发现首次成为可能。这项建议 重点关注最复杂和未被确定的基因的发现、序列解析和基因分型 人类遗传变异的形式，包括多拷贝数变异(MCNV)、倒置和中间- 调整插入和删除的大小。我们的目标是一个由34个人类基因组组成的多样性小组，并将长期阅读的单链- 分子，使用10倍链接读数和Strand-Seq数据的实时测序数据，以便完全相和 序列-解析每个人类单倍型上的SVS。利用这些长时间读取的序列数据，我们进一步开发了一个 基于计算图的区分和组装大型mCNV下不同拷贝的方法 映射到高同一性的分段复制。最后，我们利用序列结构， 包括断点和拷贝之间的序列差异，以更准确地对这些变体进行基因分型 在一个由2800个人类基因组组成的多样性小组中，短读全基因组序列数据已经 可用。这项工作将开发新的方法来表征更复杂的人类基因变异形式 并提供了对它们的多样性、起源机制和突变特性的基本见解。这 研究还有一个额外的好处，那就是它将改善基因组组装，表征新的人类基因组 序列，识别一大类缺失的基因变异，并为我们提供系统地 作为疾病关联研究的一部分，探索这种形式的人类基因变异。

Understanding the genetic basis of human disease requires a comprehensive assessment of the full spectrum of human genetic variation. Genome structural variation, including larger deletions, insertions, and inversions (>50 bp), has been more difficult to characterize due to the association with repetitive DNA. The majority of structural variation, including common structural variants or SVs, has not yet been discovered using short-read whole-genome datasets and standard SV callers. Advances in sequencing technology over the last three years, however, have made the systematic discovery of this variation possible for the first time. This proposal focuses on the discovery, sequence resolution, and genotyping of the most complex and under-ascertained forms of human genetic variation, including multi-copy number variants (mCNVs), inversions, and intermediate- size insertions and deletions. We target a diversity panel of 34 human genomes and partition long-read single- molecule, real-time sequencing data using 10X linked reads and Strand-seq data in order to fully phase and sequence-resolve SVs on each human haplotype. Using these long-read sequence data, we further develop a computational graph-based approach to distinguish and assemble distinct copies underlying large mCNVs mapping to high-identity segmental duplications. Finally, we take advantage of the sequence structure, including breakpoints and sequence differences among the copies, to more accurately genotype these variants in a diversity panel of >2,800 human genomes where short-read whole-genome sequence data are already available. The work will develop new methods to characterize more complex forms of human genetic variation and provide fundamental insight into their diversity, mechanism of origin, and mutational properties. This research has the additional benefit that it will improve genome assembly, characterize new human genome sequence, identify a large class of missing genetic variation, and provide us with the ability to systematically explore this form of human genetic variation as part of disease-association studies.