Project ID: plumID:23.026
Name: Machine Learning Nucleation Collective Variables with Graph Neural Networks
Archive: https://github.com/mme-ucl/NNucleate/raw/main/examples/plumed_files.zip
Category: chemistry
Keywords: Nucleation, Machine Learning, Enhanced Sampling, Collective Variables, Graph Neural Networks
PLUMED version: 2.5.2
Contributor: Florian Dietrich
Submitted on: 29 Jun 2023
Publication: F. M. Dietrich, X. R. Advincula, G. Gobbo, M. A. Bellucci, M. Salvalaglio, Machine Learning Nucleation Collective Variables with Graph Neural Networks. Journal of Chemical Theory and Computation. 20, 1600–1611 (2023)

PLUMED input files

File Compatible with
plumed_analytical.dat tested on v2.9 tested on master
plumed_model_2D_WTD.dat tested on v2.9 tested on master with LOAD
plumed_model_pulling.dat tested on v2.9 tested on master with LOAD

Last tested: 28 Feb 2024, 06:46:35

Project description and instructions
The input file is designed to be executed using the driver to post-process LAMMPS simulations of the colloidal system described in the publication to obtain analytical reference values for the collective variables n and n(Q6). The model.py file contains a fully trained GNN model that can be used as a nucleation CV in this system using the PyCV PLUMED fork. The other PLUMED input demonstrate an example use of the trained CV in a pulling simulation. For code documentation and a tutorial visit here.

Submission history
[v1] 29 Jun 2023: original submission

Badge
Click on the image below and get the code to add the badge to your website!
plumeDnest:23.026