AF: Medium: Algorithms for Scalable Phylogenetic Network Inference

物种系统发育学模拟物种如何分裂和分化，并提供对基本生物现象和过程的重要洞察，包括生物多样性和特征进化，而基因树提供对蛋白质结构和功能以及系统生物学的洞察。测序技术和组装方法的进步以及全基因组数据集的可获得性为估计物种系统发育和基因树的准确性提供了变革性的改进的可能性。系统发育网络扩展了系统发育树，以提供网状进化史的适当模型。网状进化描述了通过两个祖先谱系的部分合并而形成一个谱系的过程。最近开发的方法允许对系统发育网络进行统计推断，以便解释在基因组进化过程中可能发挥作用的其他过程。然而，这些方法只能处理不到几个基因组。该奖项将开发从序列数据和基因树估计中估计大规模系统发育网络的方法。该奖项将刺激计算机科学和统计学的研究，并将对进化生物学产生重大影响。该奖项将为PhyloNet软件包贡献开源代码。将向社区提供讲座和教程，介绍在该奖项中取得的进展和使用PhyloNet的情况。该奖项将为培训学生和博士后在前沿、跨学科的算法研究方面提供充足的机会。该项目将通过五项活动进行，这些活动在整个奖项的一生中交织在一起。(1)开发新的算法技术，用于系统发育网络的可伸缩推理，允许分析具有数十个甚至数百个基因组的数据集。(2)PhyloNet软件包中所有方法的实现和实现。(3)从精度和计算要求两个方面对这些方法进行了全面的评价。(4)对学生和博士后研究员的指导和培训。(5)通过开放源码软件包、同行评议期刊和会议记录、讲座和教程以及课程材料来传播结果。这一奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

Species phylogenies model how species split and diverge and provide important insight into fundamental biological phenomena and processes, including biodiversity, and trait evolution, while gene trees provide insight into protein structure and function as well as systems biology. Advances in sequencing technologies and assembly methods and the availability of whole-genome datasets have opened up the possibility of transformative improvements in accuracy for estimating species phylogenies and gene trees. Phylogenetic networks extend phylogenetic trees to provide an appropriate model of reticulate evolutionary histories. Reticulate evolution describes the origination of a lineage through partial merging of two ancestor lineages. Recently developed methods allow for statistical inference of phylogenetic networks in order to account for other processes that could be at play during the evolution of the genomes. However, these methods can handle fewer than a handful of genomes. This award will develop methods for estimating large-scale phylogenetic networks from sequence data as well as gene tree estimates. The award will stimulate research in computer science and statistics and will have a major impact on evolutionary biology. The award will contribute open-source code to the PhyloNet software package. Lectures and tutorials will be given to the community on the developments made in the award and on the use of PhyloNet. The award will provide ample opportunities for training students and post-doctoral fellows in cutting-edge, interdisciplinary algorithmic research.The project will be carried out through five activities that are intertwined throughout the lifetime of the award. (1) Development of novel algorithmic techniques for scalable inference of phylogenetic networks that allow for analyzing data sets with tens and even hundreds of genomes. (2) Implementation and of all methods in the PhyloNet software package. (3) Thorough evaluation of the methods in terms of accuracy and computational requirements. (4) Mentoring and training of students and post-doctoral fellows. (5) Dissemination of the results through an open-source software package, publications in peer-reviewed journals and conference proceedings, lectures and tutorials, and course materials.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.