Skip to content

AdaptivePELE is a Python package aimed at enhancing the sampling of molecular simulations

License

Notifications You must be signed in to change notification settings

BSC-CNS-EAPM/AdaptivePELE

Repository files navigation

AdaptivePELE

MIT license GitHub release PyPI release Conda release DOI

AdaptivePELE is a Python module to perform enhancing sampling of molecular simulation built around the Protein Energy Landscape Exploration method (PELE) developed in the Electronic and Atomic Protein Modelling grop (EAPM) at the Barcelona Supercomputing Center (BSC).

Usage

AdaptivePELE is called with a control file as input parameter. The control file is a json document that contains 4 sections: general parameters, simulation parameters, clustering parameters and spawning parameters. The first block refers to general parameters of the adaptive run, while the other three blocks configure the three steps of an adaptive sampling run, first run a propagation algorithm (simulation), then cluster the trajectories obtained (clustering) and finally select the best point to start the next iteration (spawning).

An example of usage:

python -m AdaptivePELE.adaptiveSampling controlFile.conf

Installation

There are two methods to install AdaptivePELE, from repositories, either PyPI or Conda (recommended), or directly from source.

To install from PyPI simply run:

pip install AdaptivePELE

To install from Conda simply run:

conda install -c nostrumbiodiscovery -c conda-forge adaptive_pele

To install from source, you need to install and compile cython files in the base folder with:

git clone https://github.com/AdaptivePELE/AdaptivePELE.git
cd AdaptivePELE
python setup.py build_ext --inplace

Also, if AdaptivePELE was not installed in a typical library directory, a common option is to add it to your local PYTHONPATH:

export PYTHONPATH="/location/of/AdaptivePELE:$PYTHONPATH"

Documentation

The documentation for AdaptivePELE can be found here

Contributors

Daniel Lecina, Joan Francesc Gilabert, Oriol Gracia, Daniel Soler

Mantainer

Joan Francesc Gilabert ([email protected])

Citation

AdaptivePELE is research software. If you make use of AdaptivePELE in scientific publications, please cite it. The BibTeX reference is:

@article{Lecina2017,
author = {Lecina, Daniel and Gilabert, Joan Francesc and Guallar, Victor},
doi = {10.1038/s41598-017-08445-5},
issn = {2045-2322},
journal = {Scientific Reports},
number = {1},
pages = {8466},
pmid = {28814780},
title = {{Adaptive simulations, towards interactive protein-ligand modeling}},
url = {http://www.nature.com/articles/s41598-017-08445-5},
volume = {7},
year = {2017}
}