Recommended by

Indexed by

Accurate and scalable multi-element graph neural network force field and molecular dynamics with direct force architecture

Cheol Woo Park1,2*, Mordechai Kornbluth1, Jonathan Vandermause3, Chris Wolverton2, Boris Kozinsky3*, Jonathan Mailoa1*

1 Robert Bosch Research and Technology Center, Cambridge, MA 02139, USA

2 Northwestern University, Evanston, IL 60208, USA

3 Harvard School of Engineering and Applied Sciences, Cambridge, MA 02138, USA

* Corresponding authors emails: cheolpark2016@u.northwestern.edu, bkoz@seas.harvard.edu, jpmailoa@alum.mit.edu
DOI10.24435/materialscloud:66-ec [version v1]

Publication date: Apr 06, 2021

How to cite this record

Cheol Woo Park, Mordechai Kornbluth, Jonathan Vandermause, Chris Wolverton, Boris Kozinsky, Jonathan Mailoa, Accurate and scalable multi-element graph neural network force field and molecular dynamics with direct force architecture, Materials Cloud Archive 2021.54 (2021), doi: 10.24435/materialscloud:66-ec.


Data includes the the ab initio molecular dynamic simulation of Li7P3S11 that was used to measure the performance of the GNNFF. The data is divided into training and testing sets. Brief descirption of the work: Recently, machine learning (ML) has been used to address the computational cost that has been limiting ab initio molecular dynamics (AIMD). Here, we present GNNFF, a graph neural network framework to directly predict atomic forces from automatically extracted features of the local atomic environment that are translationally-invariant, but rotationally-covariant to the coordinate of the atoms. We demonstrate that GNNFF not only achieves high performance in terms of force prediction accuracy and computational speed on various materials systems, but also accurately predicts the forces of a large MD system after being trained on forces obtained from a smaller system. Finally, we use our framework to perform an MD simulation of Li7P3S11, a superionic conductor, and show that resulting Li diffusion coefficient is within 14% of that obtained directly from AIMD. The high performance exhibited by GNNFF can be easily generalized to study atomistic level dynamics of other material systems.

Materials Cloud sections using this data

No Explore or Discover sections associated with this archive record.


File name Size Description
65.5 MiB train.pkl and test.pkl: pickled python dictionaries containing the training and testing snapshots taken from the MD trajectory of Li7P3S11. For more detailed description of how to access the data, please refer to the README.txt file that is included.


Files and data are licensed under the terms of the following license: Creative Commons Attribution Share Alike 4.0 International.
Metadata, except for email addresses, are licensed under the Creative Commons Attribution Share-Alike 4.0 International license.

External references

Journal reference
C. W. Park, M. Kornbluth, J. Vandermause, C. Wolverton, B. Kozinsky, J. Mailoa (in preparation)


machine learning molecular dynamics density-functional theory

Version history:

2021.54 (version v1) [This version] Apr 06, 2021 DOI10.24435/materialscloud:66-ec