Mack Sean G, Sriram Ganesh
Department of Chemical and Biomolecular Engineering, University of Maryland, College Park, MD, USA.
Metab Eng Commun. 2020 Dec 2;12:e00154. doi: 10.1016/j.mec.2020.e00154. eCollection 2021 Jun.
Genome-scale stoichiometric models (GSMs) have been widely utilized to predict and understand cellular metabolism. GSMs and the flux predictions resulting from them have proven indispensable to fields ranging from metabolic engineering to human disease. Nonetheless, it is challenging to parse these flux predictions due to the inherent size and complexity of the GSMs. Several previous approaches have reduced this complexity by identifying key pathways contained within the genome-scale flux predictions. However, a reduction method that overlays carbon atom transitions on stoichiometry and flux predictions is lacking. To fill this gap, we developed NetFlow, an algorithm that leverages genome-scale carbon mapping to extract and quantitatively distinguish biologically relevant metabolic pathways from a given genome-scale flux prediction. NetFlow extends prior approaches by utilizing both full carbon mapping and context-specific flux predictions. Thus, NetFlow is uniquely able to quantitatively distinguish between biologically relevant pathways of carbon flow within the given flux map. NetFlow simulates C isotope labeling experiments to calculate the extent of carbon exchange, or carbon yield, between every metabolite in the given GSM. Based on the carbon yield, the carbon flow to or from any metabolite or between any pair of metabolites of interest can be isolated and readily visualized. The resulting pathways are much easier to interpret, which enables an in-depth mechanistic understanding of the metabolic phenotype of interest. Here, we first demonstrate NetFlow with a simple network. We then depict the utility of NetFlow on a model of central carbon metabolism in . Specifically, we isolated the production pathway for succinate synthesis in this model and the metabolic mechanism driving the predicted increase in succinate yield in a double knockout of . Finally, we describe the application of NetFlow to a GSM of lycopene-producing , which enabled the rapid identification of the mechanisms behind the measured increases in lycopene production following single, double, and triple knockouts.
基因组规模的化学计量模型（GSMs）已被广泛用于预测和理解细胞代谢。GSMs及其产生的通量预测已被证明在从代谢工程到人类疾病等领域中不可或缺。然而，由于GSMs固有的规模和复杂性，解析这些通量预测具有挑战性。先前的几种方法通过识别基因组规模通量预测中包含的关键途径来降低这种复杂性。然而，缺乏一种将碳原子转换叠加到化学计量和通量预测上的简化方法。为了填补这一空白，我们开发了NetFlow算法，该算法利用基因组规模的碳映射从给定的基因组规模通量预测中提取并定量区分生物学相关的代谢途径。NetFlow通过利用全碳映射和特定背景的通量预测扩展了先前的方法。因此，NetFlow能够独特地定量区分给定通量图中生物学相关的碳流途径。NetFlow模拟碳同位素标记实验，以计算给定GSM中每种代谢物之间的碳交换程度或碳产量。基于碳产量，可以分离并轻松可视化流向或来自任何感兴趣代谢物或任何一对感兴趣代谢物之间的碳流。由此产生的途径更易于解释，从而能够对感兴趣的代谢表型进行深入的机制理解。在这里，我们首先用一个简单的网络演示NetFlow。然后，我们描述NetFlow在[具体物种]中心碳代谢模型上的效用。具体来说，我们在该模型中分离了琥珀酸合成的生产途径以及在[具体基因]双敲除中驱动预测的琥珀酸产量增加的代谢机制。最后，我们描述了NetFlow在产番茄红素[具体物种]的GSM上的应用，这使得能够快速识别在单敲除、双敲除和三敲除后测得的番茄红素产量增加背后的机制。
