Machine learning for molecular simulation

By Frank No'e, Alexandre Tkatchenko, K. Müller, Cecilia Clementi. Annual review of physical chemistry (Print), 2019

819 citations recorded by the source corpus.

Machine learning (ML) is transforming all areas of science. The complex and time-consuming calculations in molecular simulations are particularly suitable for an ML revolution and have already been profoundly affected by the application of existing ML methods. Here we review recent ML methods for molecular simulation, with particular focus on (deep) neural networks for the prediction of quantum-mechanical energies and forces, on coarse-grained molecular dynamics, on the extraction of free energy surfaces and kinetics, and on generative network approaches to sample molecular equilibrium structures and compute thermodynamics. To explain these methods and illustrate open methodological problems, we review some important principles of molecular physics and describe how they can be incorporated into ML structures. Finally, we identify and describe a list of open challenges for the interface between ML and molecular simulation. Expected final online publication date for the Annual Review of Physical Chemistry, Volume 71 is April 20, 2020. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

Subfields: Atomic and Molecular Physics, and Optics, Materials Chemistry

View Paper: https://www.semanticscholar.org/paper/b674a7aee72e9b9cc5390eca13f9c5c7812f2ba0

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

In Materials Chemistry

Citations in the corpus

Conditional Set Generation with Transformers (2020, background): “Some examples include predicting future states for a group of molecules in a simulation (Noé et al., 2020), object detection from images (Carion et al., 2020) and generating correlated samples for sequential Monte Carlo in object tracking (Zhu et al., 2020; Neiswanger et al., 2014).”
Emerging materials intelligence ecosystems propelled by machine learning (2020, background): “Likewise, other innovative material fingerprints and model architectures have also been proposed that account for invariance and symmetries present in the target property of interest (system energy, atomic forces) 222–225 .”
Integrating Physics-Based Modeling with Machine Learning: A Survey (2020, background): “This distinguishes our survey from other works that focus on more general knowledge integration into machine learning [1, 256] and other works covering physics integration into ML in specific domains (e.g., cyber-physical systems [204], chemistry [184]).”
Deep learning model to predict complex stress and strain fields in hierarchical composites (2021, background): “By incorporating field-based (density function theory), particle-based (MD), and continuum-based (FEM) modeling, ML sheds light on predictions of quantum interactions ( 33 – 35 ), molecular force fields ( 36 – 38 ), and material mechanics ( 19 , 20 , 39 – 41 ).”
Enhancing computational fluid dynamics with machine learning (2021, methodology): “Approaches to improve CFD with ML are aligned with the larger efforts to incorporate ML into scientific computing, for instance via physics-informed neural networks (PINNs) [16, 17] or to accelerate computational chemistry [8, 18].”