Leveraging identity-by-descent information in large-scale population sequencing

描述(由申请人提供)：这个项目旨在开发一种方法，在这种方法中，可以在看似不相关的个体之间推断出特定染色体片段的共同祖先基因流的模式，并将其用于对与精神分裂症等疾病相关的通过测序发现的罕见突变的分析。血统同一性(IBD)是指两个或多个个体各自携带一段延伸的单倍体序列，该序列是单个祖先单倍型的直接拷贝或后代，该单倍型居住(或曾经居住)在这些个体的最近共同祖先中。在大样本中，发现数以千计的例子并不少见，在这些例子中，看似不相关的个体，其基因组的一小部分，与父母和后代的亲缘关系一样密切。在对罕见和常见遗传变异进行大规模、基于人群的研究的背景下，我们提出，在来自测序的罕见突变和多态的数据集之上分层个体内IBD共享的地图可以帮助解决将遗传变异与常见疾病的风险联系起来的艰巨挑战。具体地说，我们建议在测序研究中使用IBD共享信息来1)识别可能的从头开始和非常新的(私人)突变，2)确定优先顺序 稀有变异可能会对功能产生影响，3)允许额外的未测序样本根据稀有等位基因与测序个体共享IBD的可能性，对稀有等位基因进行优先排序。我们将把这里开发的方法应用于两个大型的精神分裂症测序研究，包括超过6,000个个体的全外显子组数据和超过14,000个基因组范围的SNP数据。此处制定的统计方法将作为PLINK/SEQ软件包的一部分加以实施和分发。

DESCRIPTION (provided by applicant): This project aims to develop ways in which the patterns of shared ancestral gene-flow for specific chromosomal segments can be inferred between seemingly-unrelated individuals and used to empower analyses of rare mutations discovered by sequencing, with respect to association with diseases such as schizophrenia. Identity-by-descent (IBD) implies that two or more individuals each carry an extended stretch of haploid sequence that is a direct copy, or descendant, of a single, ancestral haplotype that resides (or once resided) in a recent common ancestor of those individuals. In large samples it is not unusual to find many thousands of instances in which seemingly unrelated individuals are, for some fraction of their genome, related exactly as closely as are parent and offspring. In the context of large, population-based studies of rare and common genetic variation, we propose that layering a map of intra-individual IBD sharing on top of datasets of rare mutation and polymorphism from sequencing can help in the daunting challenge of relating genetic variation to risk for common disease. Specifically, we propose to use IBD sharing information in sequencing studies to 1) identify likely de novo and very recent (private) mutations, 2) prioritize rare variants for likely functional impact and 3) allow additional un-sequenced samples to prioritize rare alleles according to the likelihood they are causal given their IBD sharing with sequenced individuals. We will apply the methods developed here to two large schizophrenia sequencing studies, with whole-exome data on over 6,000 individuals and genome-wide SNP data on over 14,000. The statistical approaches developed here will be implemented and distributed as part of the PLINK/Seq software package.