Metabolomics analysis: Finding out metabolic building blocks

In this paper we propose a new methodology for the analysis of metabolic networks. We use the notion of strongly connected components of a graph, called in this context metabolic building blocks. Every strongly connected component is contracted to a single node in such a way that the resulting graph is a directed acyclic graph, called a metabolic DAG, with a considerably reduced number of nodes. The property of being a directed acyclic graph brings out a background graph topology that reveals the connectivity of the metabolic network, as well as bridges, isolated nodes and cut nodes. Altogether, it becomes a key information for the discovery of functional metabolic relations. Our methodology has been applied to the glycolysis and the purine metabolic pathways for all organisms in the KEGG database, although it is general enough to work on any database. As expected, using the metabolic DAGs formalism, a considerable reduction on the size of the metabolic networks has been obtained, specially in the case of the purine pathway due to its relative larger size. As a proof of concept, from the information captured by a metabolic DAG and its corresponding metabolic building blocks, we obtain the core of the glycolysis pathway and the core of the purine metabolism pathway and detect some essential metabolic building blocks that reveal the key reactions in both pathways. Finally, the application of our methodology to the glycolysis pathway and the purine metabolism pathway reproduce the tree of life for the whole set of the organisms represented in the KEGG database which supports the utility of this research.本文提出一种用于代谢网络分析的全新方法论。本研究采用图论中的强连通分量（strongly connected components）概念，并将其在本研究语境中称为代谢构建模块（metabolic building blocks）。将每个强连通分量收缩为单个节点后，最终得到的图为有向无环图（directed acyclic graph, DAG），本文中将其称为代谢DAG，其节点数量已大幅缩减。作为有向无环图的特性，该结构可展现代谢网络的背景拓扑，清晰揭示代谢网络的连通特性，以及桥（bridges）、孤立节点与割点（cut nodes）的分布情况。综上，该拓扑结构可为功能性代谢关联的挖掘提供关键信息。本方法论已针对KEGG数据库（Kyoto Encyclopedia of Genes and Genomes）中的所有物种，应用于糖酵解（glycolysis）与嘌呤代谢通路（purine metabolic pathways）的分析，尽管该方法通用性极强，可适配任意数据库。正如预期，通过代谢DAG形式化方法，可实现代谢网络规模的大幅缩减，其中嘌呤代谢通路因本身规模相对更大，缩减效果尤为显著。作为概念验证，我们通过代谢DAG及其对应的代谢构建模块所捕获的信息，成功提取出糖酵解通路与嘌呤代谢通路的核心结构，并识别出若干关键代谢构建模块，可揭示两条通路中的核心反应过程。最后，将本方法论应用于糖酵解通路与嘌呤代谢通路的分析，成功复现了KEGG数据库中收录的所有物种的生命之树，这一结果验证了本研究的实用价值。