Abstract：
We now live in an era where metabolic data are increasingly accessible and available. Analysis of this data can be done using novel techniques (e.g., machine learning and artificial intelligence) but this does not alleviate scientists to use \"human intelligence\". The objective of this paper is to combine the information of a large database of biochemical reactions with a method and tool to make nutritional biochemistry more accessible to nutritionists. A script was developed to extract information from a database with more than 16 000 biochemical reactions so that it can be used for \"biochemical bookkeeping\". A system of more than 300 meta-reactions (i.e., the outcome reaction of a series of connected individual reactions) was constructed covering a wide range of metabolic pathways for macro- and micronutrients. Meta-reactions were constructed by identifying metabolic nodes, which are inputs or outputs of a metabolic system or that serve as connection points between meta-reactions. Complete metabolic pathways can be constructed by combining and balancing the meta-reactions using a simple Excel tool. To illustrate the use of meta-reactions and the tool in the teaching of nutritional biochemistry, examples are given to illustrate how much ATP can be synthesized from glucose, either directly or indirectly (i.e., via storage and mobilization or via transfer of intermediate metabolites between tissues and generations). To illustrate how meta-reactions and the tool can be used in research, nutrient balance data of the mammary gland of a dairy cow were used to construct a plausible pathway of nutrient metabolism of the whole mammary gland. The balance data included 34 metabolites taken up or exported by the mammary gland and 39 meta-reactions were used to construct a metabolic pathway that accounted for the uptake and output of metabolites. The results highlighted the importance of the synthesis of proline from arginine and the concomitant synthesis of urea by the mammary gland. It also raised the question of whether the availability of metabolic pathways or glucose uptake would be the more limiting factor for the synthesis of NADPH required for fatty acid synthesis. The availability of an open database with biochemical reactions, the concept of meta-reactions, and the provision of a tool allow users to construct metabolic pathways, which helps acquiring a more comprehensive and integrated view of metabolism and may raise issues that may be difficult to identify otherwise.
摘要：
我们现在生活在一个代谢数据越来越可访问和可用的时代。可以使用新技术(例如，机器学习和人工智能)，但这并不能缓解科学家使用“人类智能”的问题。本文的目的是将大型生化反应数据库的信息与方法和工具相结合，以使营养学家更容易获得营养生物化学。开发了一个脚本来从具有超过16.000个生化反应的数据库中提取信息，以便将其用于“生化簿记”。超过300个元反应的系统(即，一系列相互关联的个体反应的结局反应）构建了涵盖大量和微量营养素代谢途径的广泛途径。通过识别代谢节点构建Meta反应，它们是代谢系统的输入或输出，或者用作元反应之间的连接点。可以通过使用简单的Excel工具组合和平衡元反应来构建完整的代谢途径。为了说明在营养生物化学教学中使用元反应和工具，给出了一些例子来说明葡萄糖可以合成多少ATP，直接或间接(即，通过储存和动员或通过中间代谢物在组织和世代之间的转移)。为了说明元反应和工具如何在研究中使用，利用奶牛乳腺的营养平衡数据构建了整个乳腺营养代谢的合理途径。平衡数据包括乳腺吸收或输出的34种代谢物，并使用39种元反应来构建代谢途径，以解释代谢物的摄取和输出。结果强调了从精氨酸合成脯氨酸以及乳腺伴随的尿素合成的重要性。它还提出了代谢途径或葡萄糖摄取的可用性是否将是脂肪酸合成所需的NADPH合成的更多限制因素的问题。一个开放的生化反应数据库的可用性，元反应的概念，并提供一个工具允许用户构建代谢途径，这有助于获得更全面和综合的新陈代谢视图，并可能提出可能难以识别的问题。
