Hierarchical modeling of molecular energies using a deep neural network.
By Nicholas Lubbers, Justin S. Smith, Kipton Barros. 2017
312 citations recorded by the source corpus.
We introduce the Hierarchically Interacting Particle Neural Network (HIP-NN) to model molecular properties from datasets of quantum calculations. Inspired by a many-body expansion, HIP-NN decomposes properties, such as energy, as a sum over hierarchical terms. These terms are generated from a neural network-a composition of many nonlinear transformations-acting on a representation of the molecule. HIP-NN achieves the state-of-the-art performance on a dataset of 131k ground state organic molecules and predicts energies with 0.26 kcal/mol mean absolute error. With minimal tuning, our model is also competitive on a dataset of molecular dynamics trajectories. In addition to enabling accurate energy predictions, the hierarchical structure of HIP-NN helps to identify regions of model uncertainty.
Subfields: Atomic and Molecular Physics, and Optics, Materials Chemistry
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Atomic and Molecular Physics, and Optics
- Evgeny V. Podryabinkin
- Alexander V. Shapeev
- Gábor Csányi
- Konstantin Gubaev
- Ilyes Batatia
- D. Kov'acs
- Christoph Ortner
- Gregor N. C. Simm
In Materials Chemistry
- Alexander V. Shapeev
- Evgeny V. Podryabinkin
- Konstantin Gubaev
- Ivan S. Novikov
- Gábor Csányi
- Florian Hase
- Pascal Friederich
- Jonny Proppe
Citations in the corpus
- The MLIP package: moment tensor potentials with MPI and active learning (2020, background): “A related but slightly different class of potentials are the message-passing potentials that are based on neural networks, but go beyond the local descriptors of atomic environments [41,63].”
- Machine-learned potentials for next-generation matter simulations (2021, background): “…of ML potentials-based MD simulations can be achieved by lowering the dimensionality of the atomic representations of the local chemical environment and overcoming the limitations that arise due to the adverse scaling with the number of elements 40 described by the ML potentials 35,147 .”
- MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields (2022, methodology): “The DTNN [42], SchNet [41], HIP-NN [35], PhysNet [48], or DimeNet [20, 29] approaches could only come close to but not improve upon the atomic descriptor-based methods in terms of computational efficiency and accuracy on public benchmarks.”