Models to infer evolution in declining populations and across space

进化、人口统计学和遗传学相互作用，形成了一个决定种群命运的复杂系统。更好地理解这个复杂的系统（种群生物学）具有内在价值——例如，揭示我们物种的进化史——并指导保护生物多样性、改善作物、控制害虫和病原体的实际尝试。我的研究计划的长期愿景是利用数学模型开发关于种群生物学的更深层次的直觉和新颖的见解，这对于从大量数据（例如，数千个基因组）中进行复杂推理和推断都是必不可少的。在这个理论种群生物学的更广泛的框架内，本提案试图解决两个基本问题，使用数学建模，计算机模拟和基因组数据的组合。1)进化是如何影响人口减少的？我们可能正处于第六次物种大灭绝之中，许多物种的数量都在下降。与此同时，我们也受到我们试图消灭的病原体和害虫的攻击。这些人口下降是如何受到进化的影响的？理论表明，进化将对种群数量的下降产生两种截然不同的影响：1)增加遗传漂变，导致灭绝漩涡；2)快速适应进化，使种群免于灭绝。如果有的话，我们怎么知道是哪一种作用于自然种群呢？这一建议概述了两种新的方法来推断进化是如何在人口下降中起作用的，这两种方法都建立在人口遗传学的最新发展之上。一种方法是利用基因组中等位基因频率的微小变化来共同估计种群规模的变化以及适应与漂变的相对强度。另一种是从全基因组谱系中重建最近的人口学和选择。两者都有望在目前的汇总统计方法上取得巨大进步。2)人口如何在景观中移动和适应？今天的遗传模式是由选择和祖先的地理运动形成的，那么我们如何从当代基因组中提取关于过去选择和分散的信息呢？在大量样本中推断基因组中每个位点的基因谱系（树状序列）的新发现为解决这一问题开辟了许多新途径。这一提议描述了四种利用树序列中包含的大量信息来推断过去的选择和扩散的新方法。第一种方法推导出遗传祖先的可能位置，并利用这些推断出大范围的地理历史，例如冰川避难所之间的混合物。第二种方法扩展了这种方法，以检测物种范围的持续扩张和收缩。第三种方法联合估计了有利等位基因的选择和扩散，以描述有利等位基因的空间分布。第四部分旨在重建表型曲线的进化，如人类身高的纬度梯度。

Evolution, demography, and genetics interact to form a complex system that determines the fate of populations. Better understanding this complex system (population biology) holds intrinsic value -- e.g., unraveling the evolutionary history of our species -- and guides practical attempts to conserve biodiversity, improve crops, and control pests and pathogens. The long-term vision of my research program is to develop deeper intuition and novel insights about population biology using mathematical models, which are essential for both complex reasoning and inference from vast amounts of data (e.g., thousands of genomes). Within this broader framework of theoretical population biology, this proposal seeks to address two fundamental questions, using a combination of mathematical modelling, computer simulation, and genomic data. 1) How does evolution affect declining populations? We are in the midst of what is perhaps a sixth mass extinction, with many populations in demographic decline. At the same time we are also barraged by pathogens and pests that we attempt to extinguish. How are these demographic declines impacted by evolution? Theory suggests that evolution will have one of two contrasting effects on declining populations: i) increasing genetic drift leading to an extinction vortex and ii) rapid adaptive evolution rescuing populations from extinction. How can we tell which, if either, is acting on natural populations? This proposal outlines two novel methods for inferring how evolution acts in declining populations, both building on recent developments in population genetics. One uses small changes in allele frequencies across the genome to jointly estimate a changing population size and relative strength of adaptation vs drift. The other reconstructs recent demography and selection from genome-wide genealogies. Both promise large advances over current summary statistic approaches. 2) How do populations move and adapt across the landscape? Present-day genetic patterns have been shaped by selection and the geographic movements of ancestors, so how do we extract information about past selection and dispersal from contemporary genomes? The newfound ability to infer gene genealogies at every site in the genome (tree-sequences) for a large number of samples opens up many new ways to approach this problem. This proposal describes four new methods that capitalize on the massive amount of information contained in tree-sequences to infer past selection and dispersal. The first derives the likely locations of genetic ancestors and uses these to deduce large scale geographic histories, such as admixture between glacial refugia. The second extends this method to detect ongoing expansions and contractions of species' ranges. The third jointly estimates selection and dispersal to describe the spatial spread of beneficial alleles. And the fourth aims to reconstruct the evolution of phenotypic clines, such as the latitudinal gradient in human height.