Statistical Methods for Experimental Genome Populations

描述（申请人提供）：项目摘要：实验模型生物，如小鼠，大鼠和果蝇提供了宝贵的见解人类疾病。这是因为我们可以在实验群体中操纵遗传和环境因素。生物学的限制限制了我们操纵遗传系统和推断因果关系的能力（相对于工业实验环境）。我们的提案解决了由于用于研究人类疾病的实验基因组群体中的这些限制而带来的独特挑战。我们将开发统计设计和分析方法，以降低实验成本，有效利用现有资源，并更好地推断因果关系时，我们有不完全的控制分配的遗传因素，以个别生物。我们关注三种类型的遗传扰动集合：（1）实验杂交（源自杂交两种或更多种品系，如回交、互交和重组近交系群体），（2）渐渗系群体（如同源品系和基因组标记的小鼠），和（3）非随机化基因组群体（如近交系集合）。我们的具体目标是：（1）发展高效、稳健的杂交试验基因分型策略。（2）发展渐渗系群体的统计设计和分析工具。(3)在实验生物的遗传关联研究中，发展检测因果因素的统计方法。相关性：科学家通过研究实验构建的动物种群（如小鼠）发现导致人类疾病易感性的遗传因素。我们将开发统计方法来有效地构建和使用这些群体。我们的方法将降低实验成本，并提高对疾病易感性的因果遗传因素的检测。

DESCRIPTION (provided by applicant): Project Summary: Experimental model organisms such as mice, rats, and Drosophila provide valuable insights into human diseases. This is because we can manipulate genetic and environmental factors in experimental populations. Biological constraints limit our ability to manipulate genetic systems, and infer causation (relative to industrial experimental settings). Our proposal addresses unique challenges due to these constraints in experimental genome populations used to study human diseases. We will develop statistical design and analysis methods that will reduce experimental cost, make efficient use of existing resources, and better infer causation when we have incomplete control over the assignment of genetic factors to individual organisms. We focus on three types of collections of genetic perturbations: (1) experimental crosses (derived from crossing two or more strains such as backcross, intercross, and recombinant inbred line populations), (2) introgression line populations (such as consomic strains, and genome-tagged mice), and (3) non- randomized genome populations (such as inbred strain collections). Our specific aims are: (1) To develop efficient and robust genotyping strategies in experimental crosses (2) To develop statistical design and analysis tools for introgression line populations. (3) To develop statistical methods for detecting causal elements in genetic association studies in experimental organisms. Relevance: Scientists discover genetic elements contributing to human disease susceptibility by studying experimentally-constructed animal populations (such as mice). We will develop statistical methods to construct and use these populations efficiently. Our methods will reduce experimental cost, and improve detection of causal genetic elements contributing to disease susceptibility.