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).
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
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"
The documentation for AdaptivePELE can be found here
Daniel Lecina, Joan Francesc Gilabert, Oriol Gracia, Daniel Soler
Joan Francesc Gilabert ([email protected])
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}
}