Genome-scale metabolic reconstructions include all known biochemical reactions occurring in a cell. A typical application is the prediction of potential drug targets for cancer treatment. The precision of these predictions relies on the definition of the objective function. Generally, the biomass reaction is used to illustrate the growth capacity of a cancer cell. Today, seven human biomass reactions can be identified in published metabolic models. The impact of these differences on the metabolic model predictions has not been explored in detail. We explored this impact on cancer metabolic model predictions and showed that the metabolite composition and the associated coefficients had a large impact on the growth rate prediction accuracy, whereas gene essentiality predictions were mainly affected by the metabolite composition. Our results demonstrate the importance of defining a consensus biomass reaction compatible with most human models, which would contribute to ensuring the reproducibility and consistency of the results.

基因组规模的代谢重建包括发生在细胞中的所有已知生化反应。一个典型的应用是预测癌症治疗的潜在药物靶点。这些预测的精度取决于目标函数的定义。通常，生物质反应用于说明癌细胞的生长能力。今天，可以在已发表的代谢模型中识别出七种人类生物量反应。尚未详细探讨这些差异对代谢模型预测的影响。我们探讨了这种对癌症代谢模型预测的影响，并表明代谢物组成和相关系数对增长率预测准确性有很大影响，而基因重要性预测主要受代谢物组成的影响。