A Physics-Guided Neural Network for Predicting Protein–Ligand Binding Free Energy: From Host–Guest Systems to the PDBbind Database

Cain, Sahar; Risheh, Ali; Forouzesh, Negin

doi:10.3390/biom12070919

Cited by 8 publications

(3 citation statements)

References 62 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The downside to acquiring data from Sabio-RK or BRENDA is it will have to be cleaned extensively as this data is of quite low quality due to misannotations, human error in curation, and high experimental variance [53,51,52]. Another possible approach for improving model performance is to integrate physical principles of molecule-protein interaction into enzymesubstrate prediction models, which could lead to further improvements in model performance [63,64]. The addition of 3D protein information could also be explored as this is critical for determining substrate preferences in many enzyme-catalyzed reactions [65][66][67][68].…”

Section: Discussionmentioning

confidence: 99%

VIPER: A General Model for Prediction of Enzyme Substrates

Campbell

2024

Preprint

View full text Add to dashboard Cite

Enzymes, nature’s catalysts, possess remarkable properties such as high stereo-, regio-, and chemo-specificity. These properties allow enzymes to greatly simplify complex synthetic processes, resulting in improved yields and reduced manufacturing costs compared to traditional chemical methods. However, the lack of experimental characterization of enzyme substrates, with only a few thousand out of tens of millions of known enzymes in Uniprot having annotated substrates, severely limits the ability of chemists to repurpose enzymes for industrial applications. Previous machine learning models aimed at predicting enzyme substrates have been hampered by poor generalization to new substrates. Here, we introduce VIPER (Virtual Interaction Predictor for Enzyme Reactivity), a model that achieves an average 34% improvement over the previous state-of-the-art model (ProSmith) in reaction prediction for unseen substrates. Furthermore, we present a novel benchmarking methodology for assessing the out-of-distribution generalization capabilities of enzyme-substrate prediction models. VIPER represents a significant advance towards the in silico prediction of enzyme-substrate compatibility, paving the way for the discovery of novel biocatalytic routes for the sustainable synthesis of high-value chemicals.

show abstract

Section: Discussionmentioning

confidence: 99%

VIPER: A General Model for Prediction of Enzyme Substrates

Campbell

2024

Preprint

View full text Add to dashboard Cite

show abstract

“…It is then not surprising that the recent explosive development of machine learning (ML) techniques, including deep neural networks (DNNs), − is already making a noticeable impact on this field, − including works aimed directly at improving the accuracy of the description of complex solvation effects. − It should be noted that the majority of these recent works combine quantum mechanics (QM)-based methodology with ML, while our interest here is purely classical approaches. Among recent purely ML-based approaches, a featurization algorithm, based on functional class fingerprints and implemented within the DeepChem ML framework, was used in ref to predict the hydration-free energies (HFEs) of a diverse set of 642 neutral small molecules available in FreeSolvarguably the largest public database of experimentally measured HFEs.…”

Section: Introductionmentioning

confidence: 99%

Improving the Accuracy of Physics-Based Hydration-Free Energy Predictions by Machine Learning the Remaining Error Relative to the Experiment

Bass,

Elder,

Folescu

et al. 2023

J. Chem. Theory Comput.

Self Cite

View full text Add to dashboard Cite

The accuracy of computational models of water is key to atomistic simulations of biomolecules. We propose a computationally efficient way to improve the accuracy of the prediction of hydration-free energies (HFEs) of small molecules: the remaining errors of the physics-based models relative to the experiment are predicted and mitigated by machine learning (ML) as a postprocessing step. Specifically, the trained graph convolutional neural network attempts to identify the "blind spots" in the physics-based model predictions, where the complex physics of aqueous solvation is poorly accounted for, and partially corrects for them. The strategy is explored for five classical solvent models representing various accuracy/speed trade-offs, from the fast analytical generalized Born (GB) to the popular TIP3P explicit solvent model; experimental HFEs of small neutral molecules from the FreeSolv set are used for the training and testing. For all of the models, the ML correction reduces the resulting root-mean-square error relative to the experiment for HFEs of small molecules, without significant overfitting and with negligible computational overhead. For example, on the test set, the relative accuracy improvement is 47% for the fast analytical GB, making it, after the ML correction, almost as accurate as uncorrected TIP3P. For the TIP3P model, the accuracy improvement is about 39%, bringing the ML-corrected model's accuracy below the 1 kcal/mol threshold. In general, the relative benefit of the ML corrections is smaller for more accurate physics-based models, reaching the lower limit of about 20% relative accuracy gain compared with that of the physics-based treatment alone. The proposed strategy of using ML to learn the remaining error of physics-based models offers a distinct advantage over training ML alone directly on reference HFEs: it preserves the correct overall trend, even well outside of the training set.

show abstract

“…Emerging data-driven models, on the other hand, are often accurate, yet not fully interpretable, and are likely to be overfitted. In [ 3 ], Cain et al explored the application of theory-guided data science in studying protein–ligand binding. A hybrid model is introduced by integrating a graph convolutional network (data-driven model) with the GBNSR6 implicit solvent (physics-based model).…”

mentioning

confidence: 99%