Machine learning applied to enzyme turnover numbers reveals protein structural correlates and improves metabolic models. Issue 1 (December 2018)

• Record Type: Journal Article
• Authors: Heckmann, David; Lloyd, Colton; Mih, Nathan; Ha, Yuanchi; Zielinski, Daniel; Haiman, Zachary; Desouki, Abdelmoneim; Lercher, Martin; Palsson, Bernhard
• Is Part Of: Nature communications. Volume 9:Issue 1(2018)
• Page Start: 1
• Page End: 10
• Publication Date: 2018-12
• Subjects: Biology -- Periodicals; Physical sciences -- Periodicals; 505
• DOI: 10.1038/s41467-018-07652-6
• Languages: English
• ISSNs: 2041-1723

Abstract

Abstract Knowing the catalytic turnover numbers of enzymes is essential for understanding the growth rate, proteome composition, and physiology of organisms, but experimental data on enzyme turnover numbers is sparse and noisy. Here, we demonstrate that machine learning can successfully predict catalytic turnover numbers inEscherichia coli based on integrated data on enzyme biochemistry, protein structure, and network context. We identify a diverse set of features that are consistently predictive for both in vivo and in vitro enzyme turnover rates, revealing novel protein structural correlates of catalytic turnover. We use our predictions to parameterize two mechanistic genome-scale modelling frameworks for proteome-limited metabolism, leading to significantly higher accuracy in the prediction of quantitative proteome data than previous approaches. The presented machine learning models thus provide a valuable tool for understanding metabolism and the proteome at the genome scale, and elucidate structural, biochemical, and network properties that underlie enzyme kinetics. Experimental data on enzyme turnover numbers is sparse and noisy. Here, the authors use machine learning to successfully predict enzyme turnover numbers forE. coli, and show that using these to parameterize mechanistic genome-scale models enhances their predictive accuracy.