Turnover number predictions for kinetically uncharacterized enzymes using machine and deep learning

The turnover number kcat, a measure of enzyme efficiency, is central to understanding cellular physiology and resource allocation. As experimental kcat estimates are unavailable for the vast majority of enzymatic reactions, accurate computational prediction is highly desirable. Existing machine learning models are often organism-specific or provide accurate predictions only for enzymes highly similar to proteins in the training set. Here, we present TurNuP, a general, organism-independent model that predicts turnover numbers for natural reactions of wild-type enzymes. We represent complete chemical reactions using differential reaction fingerprints and enzymes using a modified, re-trained Transformer-based protein representation. TurNuP outperforms previous models and generalizes well even to enzymes dissimilar to the training set. Using TurNuP-predicted kcat values to parameterize metabolic models improves proteome allocation predictions. A TurNuP web server is provided to facilitate broad use.

Alexander Kroll, Yvan Rousset, Xiao-Pan Hu, Nina A. Liebrand, Martin J. Lercher