Ancestry-based statistical methods for genetic studies

今天在人类基因组中观察到的变异是遗传过程的结果，比如突变，随着时间的推移作用于我们祖先的DNA。来自不相关个体的基因样本的“基因谱系”是一棵祖先树，它捕捉了这些遗传事件的历史。如果基因谱系是已知的，它可以提供有关有助于观察特征的突变位置的信息。然而，真正的基因谱系是无法知道的，因为时间尺度是在数万年的数量级上。我感兴趣的是如何使用遗传数据来重建这个祖先的历史，以及如何将家谱纳入统计方法，以将观察到的特征与遗传变异联系起来。我的研究项目的总体愿景——利用统计理论和群体遗传学——是开发、测试和实施基因定位的创新统计方法，利用基因谱系来模拟个体之间的遗传相似性。这是通过两个长期目标来实现的。我的研究计划的第一个目标是提出并测试一个灵活的“基因定位”统计，可以处理不同类型的反应变量，人群中因果遗传变异的可变频率，包括建模中的非遗传变量。第二个目标是设计一个模型和算法，以概率抽样基因谱系，与在家庭中观察到的遗传数据兼容。这一雄心勃勃的目标代表了基因数据中两种类型的树的合并：家谱和基因谱系。本研究计划的中心目标是为人类和群体遗传学研究人员增加基于树的基因定位的效用和可用性。所有技术都将在加拿大和国外的研究人员免费获得的开源软件中实现。我开发的统计工具可以被其他学科的研究人员使用，比如人类健康和医学遗传学。本研究项目将为HQP本科生、研究生和博士后提供高质量的培训机会。所有HQP将获得先进的统计和计算技能，这将为他们未来的职业生涯做好准备，成为政府、行业或学术机构的专业统计学家。

The variation observed today in the human genome is a result of genetic processes, such as mutation, acting over time on the DNA of our ancestors. The 'gene genealogy' of a sample of genes from unrelated individuals is an ancestral tree that captures the history of these genetic events. If the gene genealogy was known, it could provide information about the location of mutations that contribute to observed characteristics. However, the true gene genealogy cannot be known since the time scale is on the order of tens of thousands of years. I am interested in how genetic data can be used to reconstruct this ancestral history and how the genealogy can be incorporated in statistical methods to relate observed characteristics with genetic variants. The over-arching vision of my research program - which draws on both statistical theory and population genetics - is to develop, test and implement innovative statistical methods for gene mapping that use the gene genealogy to model the genetic similarity between individuals. This is accomplished with two long-term objectives. The first objective of my research program is to propose and test a flexible `gene mapping' statistic that can handle different types of response variables, variable frequencies of the causal genetic variant in the population, and that includes non-genetic variables in the modelling. The second objective is to design a model and algorithm to probabilistically sample gene genealogies that are compatible with genetic data observed on families. This ambitious objective represents a merging of two types of trees in genetic data: family trees and gene genealogies. The central goal of this research program is to increase the utility and availability of tree-based gene mapping for researchers in human and population genetics. All techniques will be implemented in freely-available open source software for researchers in Canada and abroad. The statistical tools that I develop can be used by researchers in other disciplines, such as in human health and medical genetics. This research program will provide high-quality training opportunities for undergraduate, graduate and postdoctoral HQP. All HQP will acquire advanced statistical and computational skills, which will prepare them for their future careers as professional statisticians in government, industry or in an academic setting.