When Do Quantum Mechanical Descriptors Help Graph Neural Networks Predict Chemical Properties?
Preprint
- 4 April 2024
- preprint
- Published by American Chemical Society (ACS) in ChemRxiv
Abstract
Deep graph neural networks are extensively utilized to predict chemical reactivity and molecular properties. However, because of the complexity of chemical space, such models often have difficulty extrapolating beyond the chemistry contained in the training set. Augmented model with quantum mechanical (QM) descriptors is anticipated to improve its generalizability. However, obtaining QM descriptors often requires CPU-intensive computational chemistry calculations. To identify when QM descriptors help graph neural networks predict chemical properties, we conduct a systematic investigation of the impact of atom, bond, and molecular QM descriptors on the performance of directed message passing neural networks (D-MPNNs) for predicting 16 molecular properties. The analysis surveys computational and experimental targets, classification and regression tasks, and varied dataset sizes from several hundred to hundreds of thousands of datapoints. Our results indicate that QM descriptors are mostly beneficial for D-MPNN performance on small datasets, provided that the descriptors correlate well with the targets and can be readily computed with high accuracy. Otherwise, using QM descriptors can add cost without benefit or even introduce unwanted noise that can degrade model performance. Strategic integration of QM descriptors with D-MPNN unlocks potential for physics-informed, data-efficient modeling with some interpretability that can streamline de novo drug and material designs. To facilitate the use of QM descriptors in machine learning workflows for chemistry, we provide a set of guidelines regarding when and how to best leverage QM descriptors, a high-throughput workflow to compute them, and an enhancement to Chemprop, a widely adopted open-source D-MPNN implementation for chemical property prediction.Keywords
All Related Versions
This publication has 64 references indexed in Scilit:
- Machine Learning‐Guided Computational Screening of New Candidate Reactions with High Bioorthogonal Click PotentialChemistry – A European Journal, 2023
- Generative models for molecular discovery: Recent advances and challengesWIREs Computational Molecular Science, 2022
- A gentle introduction to deep learning for graphsNeural Networks, 2020
- Predicting Regioselectivity in Radical C−H Functionalization of Heterocycles through Machine LearningAngewandte Chemie International Edition, 2020
- Prediction of Major Regio‐, Site‐, and Diastereoisomers in Diels–Alder Reactions by Using Machine‐Learning: The Importance of Physically Meaningful DescriptorsAngewandte Chemie International Edition, 2018
- The Development of Multidimensional Analysis Tools for Asymmetric Catalysis and BeyondAccounts of Chemical Research, 2016
- Learning from the Harvard Clean Energy Project: The Use of Neural Networks to Accelerate Materials DiscoveryAdvanced Functional Materials, 2015
- FreeSolv: a database of experimental and calculated hydration free energies, with input filesJournal of Computer-Aided Molecular Design, 2014
- First principles view on chemical compound space: Gaining rigorous atomistic control of molecular propertiesInternational Journal of Quantum Chemistry, 2013
- The M06 suite of density functionals for main group thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and transition elements: two new functionals and systematic testing of four M06-class functionals and 12 other functionalsTheoretical Chemistry Accounts, 2007