ABI Development: Computational infrastructure for phylogenetic analysis in the R environment

在过去的几十年里，系统发育树（用来表示基因序列、种群或物种之间的祖先-后代关系的图表）已经成为越来越广泛的生物学学科（如生态学、进化生物学、分子生物学、比较基因组学和发育学）的科学研究中不可或缺的一部分。系统发生学也开始在生物学以外的不同研究领域得到应用，如比较语言学、人类学和法医学。同时，被称为“R”的计算环境已经发展成为科学计算的主要平台之一。R的强大之处在于它提供了成千上万的函数库（称为“包”），这些函数库是由非R核心开发团队成员的计算科学家和开发人员开发和免费分发的。由于这些贡献，R已经成长为操作、可视化和分析系统发育树的卓越软件平台。也就是说，从R中的数据实际估计系统发育树的能力仍然相对有限。该奖项的主要目标是实质性地扩展可用于在R科学计算环境中进行系统发育估计的工具包。该工具包将使用开放开发模型构建，并将作为免费和开源资源分发给国内和国际科学界。用R开发科学计算软件的主要优点之一是，通过允许将数据分析管道的所有阶段构建到来自开源组件的单个、完全可复制的工作流中，从而极大地促进了科学的再现性。这项研究将创造新的、开源的、自由分发的软件工具，这些工具将在整个系统发育科学界和其他领域具有广泛的效用。总体目标是r中系统发育的计算基础设施开发。这些研究人员将特别关注四个具体目标：(1)计算树之间距离的方法的开发和应用；(2)发展新的共识和平均树方法；(3)基于似然的联合物种树估计方法的发展；最后，(4)可视化工具在树集和系统发育网络中的应用。它们特别关注这些研究目标，因为它们以特别有趣的方式相互作用，因为它们反映了需要大量研究的领域；然而，对这些主题的研究可能会在根本新的领域产生意想不到的发现。该奖项还将支持博士后和研究生培训，后者将通过针对年轻开发人员的密集系统发育方法训练营进行培训。这一培训目标的目的是帮助形成下一代计算系统发育生物学家，并将针对来自传统上在科学领域代表性不足的群体的研究生参与。该项目的系统发育工具将在CRAN网站https://cran.r-project.org/web/packages/phytools/index.htmlThis上发布，该奖项反映了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，认为值得支持。

Over the past several decades, phylogenetic trees, the graphs that are used to represent the ancestor-descendant relationships between gene sequences, populations, or species, have become integral in scientific research across an increasingly widening range of biological disciplines such as ecology, evolutionary biology, molecular biology, comparative genomics, and development. Phylogenies too have begun to see application outside of biology in such diverse fields of study as comparative linguistics, anthropology, and forensic science. Meanwhile, the computational environment known as 'R' has developed into one of the leading platforms for scientific computing. The strength of R comes from the literally thousands of contributed function libraries (called "packages") developed and distributed for free by computational scientists and developers that are not members of the R core development team. Because of these contributions, R has grown into the preeminent software platform for manipulating, visualizing, and analyzing phylogenetic trees. That said, the capability to actually estimate phylogenetic trees from data in R remains relatively limited. The major objective of this award is to substantially extend the toolkit that is available to undertake phylogeny estimation in the R scientific computing environment. This toolkit will be built using an open development model and will be distributed as a free and open-source resource to the national and international scientific communities. Among the major advantages of developing software for scientific computing in R is that so doing dramatically facilitates scientific reproducibility by allowing all stages of the data analysis pipeline to be built into a single, fully reproducible workflow from open source components.This research will create new, open-source, and freely distributed software tools that will be of broad utility throughout the phylogenetic scientific community and beyond. The general goal is computational infrastructure development for phylogenetics in R. These investigators will focus on four specific objectives in particular: (1) the development and application of methods to compute distances between trees; (2) the development of new consensus and average tree methods; (3) the development of joint species-tree estimation methods based on likelihood; and, finally, (4) the application of visualization tools for sets of trees and phylogenetic networks. They focus specifically on these research objectives because they interact in particularly interesting ways and because they reflect areas in which substantial research is needed; however, research on these topics will likely result in unanticipated discoveries in fundamentally new areas. The award will also support postdoctoral and graduate training, the latter via an intensive, phylogeny methods boot camp for young developers. This training objective will be undertaken with the aim of helping to form the next generation of computational phylogenetic biologists and will target for participation graduate students from groups traditionally underrepresented in the sciences. Phylogenetic tools from this project will be hosted on the CRAN at https://cran.r-project.org/web/packages/phytools/index.htmlThis 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.