Back to Search
Start Over
NewtonNet: a Newtonian message passing network for deep learning of interatomic potentials and forces.
- Source :
-
Digital discovery [Digit Discov] 2022 Apr 27; Vol. 1 (3), pp. 333-343. Date of Electronic Publication: 2022 Apr 27 (Print Publication: 2022). - Publication Year :
- 2022
-
Abstract
- We report a new deep learning message passing network that takes inspiration from Newton's equations of motion to learn interatomic potentials and forces. With the advantage of directional information from trainable force vectors, and physics-infused operators that are inspired by Newtonian physics, the entire model remains rotationally equivariant, and many-body interactions are inferred by more interpretable physical features. We test NewtonNet on the prediction of several reactive and non-reactive high quality ab initio data sets including single small molecules, a large set of chemically diverse molecules, and methane and hydrogen combustion reactions, achieving state-of-the-art test performance on energies and forces with far greater data and computational efficiency than other deep learning models.<br />Competing Interests: The authors declare no competing interests.<br /> (This journal is © The Royal Society of Chemistry.)
Details
- Language :
- English
- ISSN :
- 2635-098X
- Volume :
- 1
- Issue :
- 3
- Database :
- MEDLINE
- Journal :
- Digital discovery
- Publication Type :
- Academic Journal
- Accession number :
- 35769203
- Full Text :
- https://doi.org/10.1039/d2dd00008c