Design and analysis of sequencing studies for gene mapping in families

描述（由申请人提供）：尽管在过去的几年中已经确定了人类复杂性状的许多常见变异，但这些性状的大部分遗传力仍然无法解释。下一代测序目前被用于揭示遗传变异的全谱，特别关注鉴定与复杂性状相关的低频变异（例如，1-5%之间的次要等位基因频率（MAF））和罕见变异（例如，MAF<1%）。然而，由于频率低和等位基因异质性，相关罕见变异的鉴定极具挑战性。因此，开发有效的设计和高效的分析和计算工具来解决这些困难至关重要。虽然病例对照研究被广泛用于常见变异的关联研究，但由于家系中因果罕见变异的富集，家系设计为罕见变异分析提供了有效的替代方案。此外，家族研究在存在群体分层的情况下是稳健的，这是一个必不可少的属性，因为常规使用的常见变异的方法可能无法校正罕见变异的群体分层。为 在家族测序研究中，关键的一步是从序列数据推断潜在的基因型，不准确的基因型调用可能导致孟德尔不一致和关联研究的功率损失。为了解决这些挑战，在本申请中，我们建议开发一套全面的统计和计算方法，用于根据家族测序数据进行基因型/单倍型推断以及家族中的罕见变异关联分析。使用这些方法，我们将进行模拟，以调查成本效益的各种家庭设计相比，病例对照研究，以提高功率检测罕见的变异协会。我们还将把我们的方法应用于我们的阿米什家族研究中的序列数据和我们的合作者关于多种复杂性状的数据集。将发布便于用户使用和记录良好的软件包供公众使用。

DESCRIPTION (provided by applicant): Although numerous common variants have been identified for human complex traits in the past few years, a large proportion of the heritability o these traits remains unexplained. Next-generation sequencing is currently being employed to uncover the full spectrum of genetic variations with a particular focus on identifying low frequency variants (e.g. minor allele frequency (MAF) between 1-5%) and rare variants (e.g. MAF<1%) associated with complex traits. However identification of associated rare variants is extremely challenging due to the low frequency and allelic heterogeneity. Therefore it is crucial to develop effective designs and efficient analytical and computational tools to address these difficulties. Although case-control studies were extensively used for association studies of common variations, family designs provide an effective alternative for rare variant analysis due to the enrichment of causal rare variants in pedigrees. In addition, family studies are robust in the presence of population stratification, a property that is essential since routinely used methods for common variants may fail to correct for population stratification of rare variants. For family sequencing studies, a critical step is to infer underlying genotypes from sequence data and inaccurate genotype calls can lead to Mendelian inconsistencies and power loss of association studies. To address these challenges, in this application we propose to develop a comprehensive suite of statistical and computational methods for genotype/haplotype inference from family sequencing data and for rare variant association analysis in families. Using these methods we will carry out simulations to investigate cost efficiency of various family designs in comparison with case-control studies for improved power of detecting rare variant associations. We will also apply our methods to sequence data in our Amish family study and datasets from our collaborators on multiple complex traits. User-friendly and well-documented software packages will be released for public use.