8000 GitHub - Jayashrita/GAMBES: Gaussian-Mixture-Based Enhanced Sampling
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

Jayashrita/GAMBES

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

76 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Gaussian-Mixture-Based Enhanced Sampling

GAMBES

In this enhanced sampling method the bias potential is estimated from a model probability distribution.

where the probability of the entire space is estimated as a mixture of probability densities of metastable islands . The probability density of each of these islands are then estimated using a Gaussian Mixture Model from the data generated using unbiased molecular dynamics trajectories.

The relative height of each island is estimated on-the-fly using:

More details about the method can be found at: https://arxiv.org/abs/1909.07773 or https://pubs.acs.org/doi/10.1021/acs.jpclett.0c01125

The code for this algorithm can be found at GAMBES.cpp , it is compatible with PLUMED v2.7-dev.

Running GAMBES for your system

  • First run unbiased molecular dynamics simulations from at least one known state and for all known states.
  • Using the GAMBES-module.ipynb code, fit a Gaussian Mixture Model to your state using any number of descriptors.
  • The final file generated using the notebook is an input for GAMBES in PLUMED. Generate similar files for every known state of interest and name them as < file-name >.0 , < file-name >.1, ... , < file-name >.m where 0,1,...,m are your states. This < file-name > is an input for GAMBES.
  • Copy GAMBES.cpp in a folder containing your plumed input, the states generated < file-name >.* and all other necessary files for running your system.
  • Run the simulation.

About

Gaussian-Mixture-Based Enhanced Sampling

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published
0