Geometry of gene cophylogenies as relates to genome evolution and speciation

描述（由申请人提供）：对微生物基因组和物种进化以及人类基因组进化的基本了解对公共卫生和医学科学非常重要。这一建议解决了共分化的概念，即一个基因或物种谱系的分化伴随着另一个基因或物种谱系的分化。在这个过程中，两个或两个以上的谱系彼此保持密切联系：基因与物种，宿主与病原体、寄生虫或共生体。在病原体和人类基因组中，越来越多的人认识到共同分化的偏差包括基因复制、物种间的基因横向转移、通过平衡选择保留祖先多态性以及通过新功能化加速进化。该项目将汇集互补的专业知识，并在代数几何和数学生物学、分子生物学和共生系统的进化、计算机科学和生物信息学方面对学生进行交叉训练。研究人员建议：(1)建立一种新的统计模型和相应的方法和算法来同时推导基因树对，以便对它们的共散或偏离共散进行严格的检验；(2)设计和开发有效实现这些新方法的软件；(3)将这些方法应用于从基因组序列中获得的大量基因，以便更好地评估物种形成和基因组进化的历史。与现有的基于独立构建的系统发育树的方法不同，本文提出了一种基于多面体和代数几何的统计模型来确定基因序列集是否具有共发散性。该方法将共同开发两个或多个树，以更好地反映共发散谱系的特性。该方法将与系统基因组学工具一起实现，以表征祖先基因组或物种，将最近的共同祖先物种建模为具有相关但不相同的基因树拓扑结构的相关基因谱系的“云”。由于这些技术的精确算法对于基因组规模的数据集在计算上是不切实际的，因此启发式和近似法也将被开发和测试。所提出的方法将适用于广泛的生物学问题，例如识别全基因组中的共发散和非共发散基因集，评估可能的宿主-寄生虫共发散和共同进化，以及测试多模块酶中模块的共发散。在这个项目中，来自所有代表学科的研究生和本科生团队将进行联合项目。他们将学习如何收集序列数据，设计和使用生物信息学工具，用数学和统计学分析生物数据来推断基因和物种的关系。此外，一个新的系统发育课程，建立在现有的跨学科信息学证书课程，将提供给研究生。

DESCRIPTION (provided by applicant): A fundamental understanding of microbial genome and species evolution, as well as human genome evolution, is important for public health and medical science. This proposal addresses the concept of codivergence, i.e., the divergence of one gene or species lineage concomitantly with the divergence of another. In this process two or more lineages stay closely associated with one another: In this process two or more lineages stay closely associated with one another: genes with species and hosts with pathogens, parasites or symbionts. Deviations from codivergence that are increasingly recognized in pathogen and human genomes include gene duplications, lateral gene transfers between species, retention of ancestral polymorphisms by balancing selection, and accelerated evolution by neofunctionalization. This project will bring together complementary expertise and cross-train students in algebraic geometry and mathematical biology, molecular biology and evolution of symbiotic systems, and computer science and bioinformatics. The investigators propose to: (1) develop a new statistical model and corresponding methods and algorithms for simultaneous derivation of pairs of gene trees to allow rigorous tests of their codivergence or deviation from codivergence; (2) design and develop software that efficiently implements these new methods; and (3) apply such methods to the large number of genes available from genome sequences in order to better assess the history of speciation and genome evolution. Unlike existing methods based on independently constructed phylogenetic trees, a novel statistical model is proposed based on polyhedral and algebraic geometry to determine whether sets of gene sequences exhibit codivergence. The method will develop two or more trees jointly, to better reflect the properties of codivergent lineages. This approach will be implemented together with phylogenomic tools to characterize ancestral genomes or species, modeling most-recent common ancestor species as "clouds" of associated gene lineages with related but nonidentical gene tree topologies. Because exact algorithms for these techniques will be computationally impractical for genome-scale data sets, heuristics and approximations will also be developed and tested. The proposed methods will be applicable to a broad array of biological problems, such as identifying codivergent and noncodivergent gene sets in whole genomes, evaluating possible host-parasite codivergence and coevolution, and testing codivergence of modules in multi-modular enzymes. In this project, teams of students, both graduate and undergraduate from all of the represented disciplines, will conduct joint projects. They will learn how to gather sequence data, design and use bioinformatic tools, and analyze biological data with mathematics and statistics to infer gene and species relationships. Additionally, a new phylogenetics course, building on an existing interdisciplinary Informatics Certificate curriculum, will be offered to graduate students.