Machine Learning Accelerated Molecular Dynamics

Neural Equivariant Interatomic Potentials

NequIP showed for the first time that geometric equivariance leads to interatomic potentials with remarkable accuracy and an unprecedented data efficiency [1]. This deep neural network model leverages E(3)-equivariant convolutions over geometric tensors instead of invariant convolutions over scalars. NequIP is able to outperform existing, invariant Machine Learning potentials with up to 1000x fewer reference data. 

Code: https://github.com/mir-group/nequip

Allegro is a new equivariant model that combines accuracy with scalability [2], with parallel performance reaching simulation sizes of tens of millions of atoms, at near first-principles accuracy. The Allegro team was selected as finalists for the 2023 Gordon Bell Prize [3], considered the "Nobel Prize" of Computing!

Code: https://github.com/mir-group/allegro

We leverage these potentials to study a wide variety of systems at a high level of accuracy, including surface reaction in heterogeneous catalysis, organic molecules, and complex ionic conductors.

[1] Nature Comm 13, 2453 (2022)https://arxiv.org/abs/2101.03164   

[2] Nature Comm14, 579 (2023)https://arxiv.org/abs/2204.05249 

[3] https://dl.acm.org/doi/10.1145/3581784.3627041 

Active Learning of Bayesian Force Fields

We are developing Bayesian inference and active learning methods to automate the training of many-body machine learned (ML) force fields for complex materials. Using uncertainties derived from Bayesian ML models, we draw structures "on the fly" from molecular dynamics simulations to automatically update the training set of the model. This process does not need human intervention and results in 


Code: https://github.com/mir-group/flare


Dimensionality reduction, enhanced sampling, coarse graining

Learning reaction coordinates. We developed a data-driven machine learning algorithm that uses a multitask encoder neural network model to automatically reduce dimensionality of molecular dynamics trajectory data and learn the latent space of collective variables. This approach is promising for the extraction of atomic-level mechanisms governing reactions in complex chemical systems and for estimation of free energy barriers and reaction rates, especially in cases where human intuition is limited.

Coarse graining. When studying complex materials, like biomolecules or electrocatalysts, simulations can be significantly accelerated when fast degrees of freedom are removed and only the slow motions are included. We developed an exact thermodynamically consistent differentiable framework for machine learning such effective coarse-grained interactions ("free energies") for describing dynamics and response of systems to applied external fields.

[1] https://doi.org/10.1021/acs.jctc.1c00143

[2] https://www.nature.com/articles/s41524-023-01183-5 

[3] https://arxiv.org/abs/2405.19386

https://youtu.be/80orUU59CXY 

Stanene 2D -> Tin 3D transformation

paper

SiC incongruent melting

Pt nanoparticle in H2 atmosphere

Pd-Au bimetallic particle restructuring

Dipoles in BaTiO3 under an electric field

Conformational dynamics in barocaloric materials