Centre for Functional Genomics and Bio-Chips, Institute for Biochemistry and Molecular Genetics, Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia.
ESIGELEC Graduate School of Engineering, Rouen, France.
Comput Biol Med. 2022 Jun;145:105428. doi: 10.1016/j.compbiomed.2022.105428. Epub 2022 Mar 23.
COVID-19 presents a complex disease that needs to be addressed using systems medicine approaches that include genome-scale metabolic models (GEMs). Previous studies have used a single model extraction method (MEM) and/or a single transcriptomic dataset to reconstruct context-specific models, which proved to be insufficient for the broader biological contexts. We have applied four MEMs in combination with five COVID-19 datasets. Models produced by GIMME were separated by infection, while tINIT preserved the biological variability in the data and enabled the best prediction of the enrichment of metabolic subsystems. Vitamin D3 metabolism was predicted to be down-regulated in one dataset by GIMME, and in all by tINIT. Models generated by tINIT and GIMME predicted downregulation of retinol metabolism in different datasets, while downregulated cholesterol metabolism was predicted only by tINIT-generated models. Predictions are in line with the observations in COVID-19 patients. Our data indicated that GIMME and tINIT models provided the most biologically relevant results and should have a larger emphasis in further analyses. Particularly tINIT models identified the metabolic pathways that are a part of the host response and are potential antiviral targets. The code and the results of the analyses are available to download from https://github.com/CompBioLj/COVID_GEMs_and_MEMs.
COVID-19 是一种复杂的疾病，需要采用系统医学方法来解决，其中包括基因组规模代谢模型 (GEM)。以前的研究使用单一的模型提取方法 (MEM) 和/或单一的转录组数据集来重建特定于上下文的模型，事实证明，这对于更广泛的生物学背景来说是不够的。我们应用了四种 MEM 并结合了五个 COVID-19 数据集。GIMME 生成的模型根据感染情况进行了分离，而 tINIT 保留了数据中的生物学变异性，并能够最好地预测代谢子系统的富集。GIMME 预测一个数据集的维生素 D3 代谢被下调，而 tINIT 则预测所有数据集的维生素 D3 代谢被下调。tINIT 和 GIMME 生成的模型预测不同数据集的视黄醇代谢下调，而胆固醇代谢下调仅由 tINIT 生成的模型预测。这些预测与 COVID-19 患者的观察结果一致。我们的数据表明，GIMME 和 tINIT 模型提供了最具生物学相关性的结果，应该在进一步的分析中得到更大的重视。特别是 tINIT 模型确定了宿主反应的一部分代谢途径，这些途径可能是潜在的抗病毒靶点。分析的代码和结果可从 https://github.com/CompBioLj/COVID_GEMs_and_MEMs 下载。
