Statistical Methods for Relating Sequence Data to Phenotype

描述（由申请人提供）：当今生物学面临的最重要挑战之一是理解遗传变异。了解基因型变异如何转化为表型变异是我们理解进化的基础，对人类健康以及农业和保护具有巨大的实际意义。现在正在进行的大量全基因组关联研究就是明证。该项目的长期目标是开发利用序列级数据的关联映射方法。该项目有3个主要目标：第一，发展理论方法，以允许有效分析序列水平的遗传。我们建议研究不同的实验设计和数据插补方法对研究功效的影响，旨在找到优化检测与表型变异相关的遗传变异的能力的设计。我们还建议开发的方法，使独特的挑战和机遇所提出的序列水平的数据。第二，拷贝数变异[CNV]数据进化的群体遗传学模型的发展。我们的建议将开发模型，使我们能够评估所提出的拷贝数变化机制的效用，拷贝数变化模式对附近序列多态性模式的影响，并为未来基于模型的单倍型推断方法提供关键的理论基础。第三，发展理论方法，以允许在感兴趣的性状的分布与数据的全局特征（例如遗传祖先或位置）相关的情况下有效分析序列水平数据。我们的重点是混合模型和基于聚类的方法的集成。当今生物学面临的最重要挑战之一是了解个体之间的遗传变异如何转化为我们可以看到或测量的变异，例如人类的血压或水稻的耐旱性。我们的建议旨在开发方法，帮助我们使用DNA序列水平的数据来了解人类表型的遗传原因，如疾病易感性。

DESCRIPTION (provided by applicant): One of the most important challenges facing biology today is to make sense of genetic variation. Understanding how genotypic variation translates into phenotypic variation is fundamental to our understanding of evolution, and has enormous practical implications for human health as well as for agriculture and conservation. Witness the large number of genome-wide association studies now underway. The long-term objective of this project is to develop methods for association mapping methods that exploit the power of sequence-level data. The project has 3 main aims: First, the development of theoretical methods to allow efficient analysis of sequence- level genetic. We propose to investigate the effect of different experimental designs and data imputation methods on the power of the study, aiming to find designs that optimize the ability to detect genetic variation that is associated with phenotypic variation. We also propose to develop methods that allow for the unique challenges and opportunities presented by sequence-level data. Second, the development of population genetics models for the evolution of copy number variation [CNV] data. Our proposal will develop models that will allow us to assess the utility of proposed mechanisms for change in copy number, the effects of patterns of copy number variation on patterns of polymorphism in nearby sequence, and will also provide key theoretical under-pinnings for future model-based methods for haplotype inference, for example. Third, the development of theoretical methods to allow efficient analysis of sequence- level data in situations where the distribution of traits of interest is correlated with global features of the data (such as genetic ancestry or location). Our focus is on the integration of mixed-models and cluster-based methods. One of the most important challenges facing biology today is to understand how genetic variation between individuals translates into variation we can see or measure, like blood pressure in humans, or drought tolerance in rice. Our proposal seeks to develop methods that will help us use DNA sequence-level data to understand the genetic causes of human phenotypes such as disease susceptibility.