Collaborative Research: ABI Innovation: An Ontology-Based System for Querying Life in a Post-Taxonomic Age

提高我们对生命的理解，无论是像我们自己这样的个体物种的生物学，还是控制生物多样性的机制和过程，关键取决于对来自许多不同生物的生物数据的整合、查询和汇总。直到今天，实现这一目标的最基本和最常见的方法依赖于生物体的名称，使它们成为查询和管理我们的生物知识和数据的支柱之一。然而，传统上使用的基于林奈命名法的生物和生物群的名称，在整合和交流数据时，其实用性受到两个主要限制。首先，由于它们是简单的文本字符串，因此名称的创造者和应用者想要表达的含义对于机器来说是不可访问的。因此，确切地说，一个名称应该包括或不应该包括哪种生物往往是模棱两可的，因此名称的使用往往不一致。第二，有许多生物类群还没有，也可能永远不会有林奈的名字，但它们的分子或宏观特征已经被发现，构成了有价值的生物学知识。该项目旨在通过生成一种称为“系统参照”的机制来解决这些问题，该机制允许通过机器可解释的独特遗传模式定义来参考任何具有共同进化血统的生物群体，这种模式将该群体与所有其他群体区分开来。这个项目建立在十多年来对系统发育分类学的理论和应用研究的基础上。随着最近对通用系统发育树的合成和不断更新，系统参考将在生命树的交流、整合和查询生物数据方面具有直接和广泛的实际应用。与权威的命名法命名相比，系统参考的目标是用户可以立即为他们希望交流发现的任何共享进化血统的群体构建系统参考。除了开发系统发育命名标准和支持工具外，该项目还旨在创建课堂教学材料和课程，以培训生物学学生系统发育命名法的理论基础和实际应用。该项目将开发的系统参考机制使用为Web开发的标准和工具，特别是Web本体语言（OWL）、本体和机器推理。具体来说，这项工作旨在将系统参考作为OWL类表达式，由充分和必要条件组成，机器推理器可以使用它来识别子类和类实例。本体论和推理技术在生物知识整合和发现方面已经显示出强大的力量，并且越来越多地被用于进化研究。作为该项目的一部分，将研究、实现和测试在本体论框架内构建和计算系统参考的规范。系统参考的设计目标是，生命之树上的任何元素，无论是节点、分支还是进化枝，都可以以一种明确的方式被引用，并且具有由进化关系模式定义的完全可计算的语义。本研究的主要目标包括创建OWL中系统参考和系统发生编码与推理的形式化规范；通过领域专家验证的小规模测试来确定规范的正确性；最后将该方法扩展到大规模生物多样性数据资源导航概念验证应用中。具体来说，系统参考文献将被开发并整合到开放生命之树、ARBOR和由生命谱系计划新资助的项目中。该项目的结果将在http://www.phyloref.org上公布。

Improving our understanding of life, whether the biology of individual species such as our own, or the mechanisms and processes governing biodiversity at large, critically depends on integrating, querying, and aggregating biological data from many different organisms. To this day, the most fundamental and common way to accomplish this relies on organism names, making these one of the pillars of querying and managing our biological knowledge and data. However, the traditionally used names for organisms and groups of organisms, which are based in Linnaean nomenclature, suffer from two major limitations to their usefulness when it comes to integrating and communicating data. First, because they are simple text-strings, the meaning intended by those who coin a name and those who apply it is inaccessible to machines. As a result, exactly which organism a name is or is not meant to include is often ambiguous, and names are therefore often applied inconsistently. Second, there are many groups of organisms that do not yet and may never have a Linnaean name, but for which molecular or macroscopic characteristics have been discovered that constitute valuable biological knowledge. This project aims to address these issues by generating a mechanism, called phyloreferencing, that allows referring to any group of organisms of shared evolutionary descent by a machine-interpretable definition of the unique pattern of descent that distinguishes the group from all others. This project builds on more than a decade's worth of theoretical and applied research into phylogenetic taxonomy. With the recent synthesis and continuous update of a universal phylogenetic Tree of Life, phyloreferences will have immediate and broad practical applications for communicating, integrating, and querying biological data across the Tree of Life. In contrast to authoritative nomenclatural naming, the goal for phyloreferences is that users can construct them instantly for any group of shared evolutionary descent for which they wish to communicate discoveries. In addition to developing the phyloreferencing standard and supporting tools, this project also aims to create classroom-ready teaching materials and curricula for training biology students in the theoretical underpinnings and practical applications of phylogenetic nomenclature. The phyloreferencing mechanism that this project will develop uses standards and tools developed for the Web, specifically the Web Ontology Language (OWL), ontologies, and machine reasoning. Specifically, this work aims to develop phyloreferences as OWL class expressions consisting of sufficient and necessary conditions, which machine reasoners can use to identify subclasses and class instances. Ontology and reasoning technologies have already shown their power for biological knowledge integration and discovery and are increasingly being adopted for evolutionary research as well. As part of this project, a specification for constructing and computing with phyloreferences within an ontological framework will be researched, implemented, and tested. Phyloreferences will be designed with the goal that any element, whether node, branch, or clades, on the Tree of Life can be referenced in a way that is unambiguous and has fully computable semantics defined by patterns of evolutionary relatedness. The main objectives of this research include creating a formal specification for phyloreference and phylogeny encoding and reasoning in OWL; ascertaining correctness of the specification using small-scale tests verifiable by domain experts; and finally scaling the approach to a large-scale biodiversity data resource navigation proof-of-concept application. Specifically phyloreferences will be developed and integrated into Open Tree of Life, ARBOR and projects newly funded by the Genealogy of Life program. Results of the project will be available at http://www.phyloref.org.