Steffensen Jon Lund, Dufault-Thompson Keith, Zhang Ying
Department of Cell and Molecular Biology, College of the Environment and Life Sciences, University of Rhode Island, Kingston, Rhode Island, United States of America.
PLoS One. 2018 Feb 15;13(2):e0192891. doi: 10.1371/journal.pone.0192891. eCollection 2018.
The metabolism of individual organisms and biological communities can be viewed as a network of metabolites connected to each other through chemical reactions. In metabolic networks, chemical reactions transform reactants into products, thereby transferring elements between these metabolites. Knowledge of how elements are transferred through reactant/product pairs allows for the identification of primary compound connections through a metabolic network. However, such information is not readily available and is often challenging to obtain for large reaction databases or genome-scale metabolic models. In this study, a new algorithm was developed for automatically predicting the element-transferring reactant/product pairs using the limited information available in the standard representation of metabolic networks. The algorithm demonstrated high efficiency in analyzing large datasets and provided accurate predictions when benchmarked with manually curated data. Applying the algorithm to the visualization of metabolic networks highlighted pathways of primary reactant/product connections and provided an organized view of element-transferring biochemical transformations. The algorithm was implemented as a new function in the open source software package PSAMM in the release v0.30 (https://zhanglab.github.io/psamm/).
个体生物体和生物群落的新陈代谢可被视为一个通过化学反应相互连接的代谢物网络。在代谢网络中，化学反应将反应物转化为产物，从而在这些代谢物之间转移元素。了解元素如何通过反应物/产物对进行转移，有助于通过代谢网络识别主要化合物连接。然而，此类信息并不容易获取，对于大型反应数据库或基因组规模的代谢模型而言，往往难以获得。在本研究中，开发了一种新算法，利用代谢网络标准表示中可用的有限信息自动预测元素转移的反应物/产物对。该算法在分析大型数据集时显示出高效性，并且在与人工整理的数据进行基准测试时提供了准确的预测。将该算法应用于代谢网络的可视化，突出了主要反应物/产物连接的途径，并提供了元素转移生化转化的有序视图。该算法在开源软件包PSAMM的v0.30版本（https://zhanglab.github.io/psamm/）中作为一个新功能实现。
