Pathway-based Metabolomics Data Analysis

7项目总结 该申请要求支持开发用于分析代谢组学数据的新软件工具。 拟议的工具将分析代谢组学数据的背景下，代谢途径获得的 BioCyc数据库集合。我们工作的一个关键动机是， 代谢组学数据的背景下，现有的代谢知识，以及途径数据库，准确 绘制数千种生物体的代谢图是途径信息的逻辑来源。已知代谢 网络为解释代谢组学数据提供了一个有价值的框架， 数据通常来自于将代谢物相互连接的生化反应中的模式。 然而，生物体代谢网络的大规模和高连通性，以及大量的路径- 大多数生物体内的代谢途径，对利用现有代谢途径的概念提出了重大的认知障碍。 分析代谢组学数据的知识。这项工作的一个主要动机是降低这些认知障碍- 通过识别代谢网络的子集，让用户集中注意力。我们将评估 遵循两个建议的软件工具，将其应用于重新分析现有的代谢组学数据集， 由我们的团队创造的。 我们的第一个目标是开发软件来计算一组代谢物的最小途径覆盖集，即， 在提交的代谢组学数据集中包含代谢物的最小数量的代谢途径 由用户。我们的第二个目标是开发软件来计算代谢物组的反应生成树，以阐明 代谢物之间的机械关系。作为前面以路径为中心的算法的同伴， 该算法将有机体的全部代谢视为图，并将计算最小树 代谢组学数据集中的代谢物。最小路径覆盖集和反应 生成树将使用我们的Pathway Tools软件中的现有途径显示算法来显示。毛皮- 因此，我们的新软件将集成到Pathway Tools中，Pathway Tools是一个在 生物信息学

7 Project Summary This application requests support for the development of new software tools for analyzing metabolomics data. The proposed tools will analyze metabolomics data in the context of metabolic pathways obtained from the BioCyc database collection. A key motivation underlying our work is that it is advantageous to interpret metabolomics data in the context of existing metabolic knowledge, and that pathway databases that accurately map the metabolism of thousands of organisms are a logical source of pathway information. Known metabolic networks provide a valuable framework for interpreting metabolomics data because patterns in metabolomics data often result from patterns among the biochemical reactions that interconnect metabolites. However, the large size and high connectivity of organism metabolic networks, and the large number of path- ways within most organisms, present significant cognitive barriers to this notion of using existing metabolic knowledge to analyze metabolomics data. A major motivation of this work is to lower these cognitive barri- ers by identifying subsets of the metabolic network for the user to focus their attention on. We will evaluate the following two proposed software tools by applying them to re-analyze existing metabolomics datasets previously generated by our group. Our first aim is to develop software to compute minimal pathway covering sets for a set of metabolites, that is, a minimal number of metabolic pathways that contain the metabolites within a metabolomics dataset submitted by a user. Our second aim is to develop software to compute reaction spanning trees for metabolite sets to eluci- date mechanistic relationships among metabolites. As a companion to the preceding pathway-centric algorithm, this algorithm will consider the full metabolism of the organism as a graph, and will compute a minimal tree that spans the metabolites in a metabolomics dataset. Both the minimal pathway-covering sets and the reaction spanning trees will be displayed using existing pathway display algorithms in our Pathway Tools software. Fur- thermore, our new software will be integrated into Pathway Tools, which is a highly used software package in bioinformatics.