Collaborative Development of Biomedical Ontologies and Terminologies

描述（由申请人提供）：定义实体及其之间关系的本体的发展已成为现代生物医学工作的必要条件。本体论帮助人类和计算机管理在生物学和医学中无处不在的迅速增长的数据。注释、检索和集成高通量数据集、处理自然语言和构建决策支持系统的需求，使许多研究人员社区开始构建大型本体。迄今为止，由于缺乏促进大规模本体和词汇表的分布式、协作工程的方法和工具，这些本体开发小组受到了限制。在这项建议中，我们概述了三个具体目标。首先，我们将探讨协作本体工程所必需的基本计算方法。我们将研究表示不同协作工作流的方法，关于变更和概念历史的信息，信任和来源，以及记录决策制定和设计原理的方法。对现有本体开发项目的实证分析将为我们构建协作开发工作流程的模型提供信息，这些工作流程将指导创作、审查和管理生物医学本体的过程。其次，我们将使用第一个特定目标的结果来构建cProtigi，这是一组健壮的、可定制的、交互式的工具，以支持分布式用户在他们的协作工作中构建和编辑术语和本体。第三，我们将在现实世界的背景下评估我们的工作，大规模的本体工程项目，包括国家自闭症研究数据库的自闭症本体；世卫组织《国际疾病分类》第11次修订；由美国国立卫生研究院支持的众多研究人员正在开发的生物医学研究本体论；以及NCI正在开发的BiomedGT。想象研究人员能够创造独立工作的生物医学本体已经不再可行。我们将研究的协作方法和我们将建立的工具将带来更多的机会，以支持遍布BISTI、CTSAs、ncbc和无数需要强大、可扩展本体的生物医学倡议的各种数据和知识密集型活动。公共卫生相关性：现代医学以知识为基础的性质要求使用本体和术语来处理和整合数据。本体开发本身成为一个协作过程，更大的研究社区的成员对新兴的本体做出贡献并发表评论。我们计划扩展Protigi本体编辑器（目前使用最广泛的本体编辑器，拥有近10万注册用户），以支持本体的协作开发，并通过在世界卫生组织的ICD-11开发和其他环境中部署它们来评估新工具。

DESCRIPTION (provided by applicant): The development of ontologies that define entities and the relationships among them has become essential for modern work in biomedicine. Ontologies help both humans and computers to manage the burgeoning data that are pervasive in biology and medicine. The need to annotate, retrieve, and integrate high-throughput data sets, to process natural language, and to build systems for decision support has set many communities of investigators to work building large ontologies. To date, these groups of ontology developers have been limited by the lack of methods and tools that facilitate distributed, collaborative engineering of large-scale ontologies and vocabularies. In this proposal, we outline three specific aims. First, we will explore basic computational methods that are essential for collaborative ontology engineering. We will investigate methods for representing diverse collaborative workflows, information about changes and concept history, trust, and provenance, and for recording decision making and design rationale. Empirical analysis of existing ontology-development projects will inform our construction of models for collaborative development workflows that will guide the processes of authoring, reviewing, and curating biomedical ontologies. Second, we will use the results from our first specific aim to build cProtigi, a set of robust, customizable, interactive tools to support distributed users in their collaborative work to build and edit terminologies and ontologies. Third, we will evaluate our work in the context of real-world, large-scale ontology-engineering projects, including the autism ontology of the National Database for Autism Research; the 11th revision of the WHO's International Classification of Diseases; the Ontology for Biomedical Investigations, under development by a wide range of NIH-supported researchers; and BiomedGT, under development by NCI. It is no longer feasible to imagine that investigators can create biomedical ontologies working independently. The collaborative methods that we will study and the tools that we will build will lead to expanded opportunities to support the diverse data- and knowledge-intensive activities that pervade BISTI, the CTSAs, the NCBCs, and myriad biomedical initiatives that require robust, scaleable ontologies. PUBLIC HEALTH RELEVANCE: The knowledge-based nature of modern medicine requires the use of ontologies and terminologies to process and integrate data. Ontology development itself becomes a collaborative process, with members of the larger research community contributing to and commenting on emerging ontologies. We plan to extend the Protigi ontology editor-the most widely used ontology editor today, with almost 100,000 registered users-to support collaborative development of ontologies and to evaluate the new tools by deploying them at the World Health Organization for the development of ICD-11 and in other settings.