Deep learning the slow modes for rare events sampling

Luigi Bonati, GiovanniMaria Piccini, and Michele Parrinello, arXiv preprint arXiv:2107.03943 (2021).

Important

This repository is kept as supporting material for the manuscript, but it is no longer updated. Check out the mlcolvar library for data-driven CVs, where you can find up-to-date tutorials and examples.

This repository contains input data and code related to the manuscript:

• data --> input files for the simulations and the CVs training
• mlcvs --> python package to train the Deep-TICA CVs
• plumed-libtorch-interface --> interface to load Pytorch models in PLUMED2 for enhanced sampling
• tutorial --> jupyter notebook with tutorial to train the CVs

Due to size limits the outputs trajectories of Chignolin and Silicon simulations are deposited in the Materials Cloud repository.