KNOWLEDGE BASED BIOLOGICAL MODELING INFORMATION SYSTEM

基于知识的生物建模信息系统

• 批准号: 3421491
• 负责人: G. OVERTON
• 金额: $ 27.76万
• 依托单位: UNIVERSITY OF PENNSYLVANIA
• 依托单位国家: 美国
• 财政年份: 1991
• 资助国家: 美国
• 起止时间: 1991-09-30 至 1993-09-29
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/3421491
• 关键词: DNA; artificial intelligence; computer assisted sequence analysis; computer system design /evaluation; eukaryote; genetic library; genetic regulatory element; information retrieval; information systems; molecular biology; nucleic acid sequence

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出了 一个“生物知识矩阵”（矩阵）来满足需求 用于组织和智能获得财富的手段 面对生物医学研究人员的信息（C.F.模型 生物医学研究，国家学院出版社，1985年）。 到 实施矩阵将需要融合人造 智能（AI）和数据库技术以及 了解生物学知识的结构。 然而， 最好直到 有关知识限制领域的试点研究 证明了该概念的可行性和效用。 一 相当普遍的兴趣领域是核酸和 蛋白质序列文库已成为无价的研究 生物医学研究界的大部分地区的工具。 除了图书馆本身，还有大量计算机 已经编写了程序来分析序列数据， 导致了重大的新发现。 这些实用性 典型的实验室研究者的图书馆可能很大 通过增强知识类型来增强它们 矩阵的管理和推理系统设想。 这项拟议工作的目的是为矩阵建造一个测试床 在现有序列库之上的概念，使两者受益 矩阵的更大目标和 序列库用户。 此过程中的步骤将包括： （1）以关系形式实施GenBank数据库 连接到AI环境； （2）创建一系列 与数据库有关的知识库，从简单开始 分类法和发展为更复杂的表示（On 较窄的测试域）； （3）创建一组用户界面 面向分子生物学家的需求； （4）调查 针对知识基础的各种推论技术 以类比为类型的特定推理，最终目的是允许 用户在本地建立和检验假设。