KNOWLEDGE BASED BIOLOGICAL MODELING INFORMATION SYSTEM

基于知识的生物建模信息系统

• 批准号: 3450947
• 负责人: G. OVERTON
• 金额: $ 29.14万
• 依托单位: UNISYS
• 依托单位国家: 美国
• 财政年份: 1987
• 资助国家: 美国
• 起止时间: 1987-09-30 至 1990-09-29
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/3450947
• 关键词: artificial intelligence; computer system design /evaluation; evolution; genetic library; growth /development; information retrieval; information system analysis; molecular biology; morphology; physiology

The Committee on Models for Biomedical Research has proposed a "matrix of biological knowledge" (Matrix) to address the need for a means to organize and intelligently access the wealth of information that confronts biomedical researchers (c.f. Models for Biomedical Research, National Academy Press, 1985). To implement the matrix will require a fusion of artificial intelligence (AI) and database technologies as well as an understanding of the structure of biological knowledge. However, a full implementation of the Matrix is best not attempted until pilot studies on a restricted domain of knowledge have demonstrated the feasibility and utility of the concept. One domain of considerable general interest is that of nucleic acid and protein sequence libraries, which have become invaluable research tools for a large segment of the biomedical research community. In addition to the libraries themselves a large number of computer programs have been written to analyze the sequence data, which has resulted in major new discoveries. The utility of these libraries to the typical laboratory investigator could be greatly enhanced by augmenting them with the types of knowledge management and reasoning systems envisioned for the Matrix. The goal of this proposed work is to build a testbed for the Matrix concept on top of the existing sequence libraries, benefitting both the larger goals of the Matrix and the immediate needs of sequence library users. Steps in this process will include: (1) implementation of the GenBank database in a relational form, interfaced to an AI environment; (2) creation of a series of knowledge bases relating to the database, starting with simple taxonomies and progression to more complex representations (on narrower test domains); (3) creation of a set of user interfaces oriented to the needs of molecular biologist; and (4) investigation of various techniques of inference against the knowledge base, in particular reasoning by analogy, with the ultimate aim of allowing users to build and test hypotheses locally.

生物医学研究模型委员会提议 一个“生物知识矩阵”（Matrix）来解决这个需求 寻找一种组织和智能获取财富的方法 生物医学研究人员面临的信息（参见模型 生物医学研究，国家科学院出版社，1985）。 到 实现矩阵需要人工融合 智能（AI）和数据库技术以及 了解生物知识的结构。 然而， 最好不要尝试完全实施矩阵，直到 对有限知识领域的试点研究 论证了该概念的可行性和实用性。 一 人们普遍感兴趣的领域是核酸和 蛋白质序列库，已成为无价的研究 适用于大部分生物医学研究界的工具。 除了图书馆本身之外，还有大量的计算机 已编写程序来分析序列数据，其中 取得了重大新发现。 这些的用处 对于典型的实验室研究人员来说，图书馆可能会极大地 通过增加知识类型来增强它们 为矩阵设想的管理和推理系统。 这项工作的目标是为 Matrix 构建一个测试平台 现有序列库之上的概念，使双方受益 矩阵的更大目标和当前的需求 序列库用户。 此过程的步骤将包括： (1) GenBank数据库的关系形式的实现， 连接到人工智能环境； (2)创建系列 与数据库相关的知识库，从简单的开始 分类法和更复杂表示的进展（关于 更窄的测试域）； (3) 创建一组用户界面 以分子生物学家的需求为导向； (4) 调查 针对知识库的各种推理技术， 特殊的类比推理，最终目的是让 用户在本地建立和测试假设。