PaCS-MD

The scripts to performe PaCS-MD

DEMO

Features

The current implementation is for local workstation. The parallel version for super computer is under construction.

Requirement

• Gromacs (2018 or later)
• AmberTools
• Python 3.x
• MDAnalysis
• numpy
• matplotlib (for analysis)

conda install -c conda-forge mdanalysis

Installation

Copy the lib directory on the your each working directory.

mkdir protein_X
cd ./protein_X
git clone https://github.com/riquri/PaCS-MD.git

Place the initial and target structure

cp {your_initial_structure.pdb} ./input/initial.pdb
cp {your_target_structure.pdb} ./input/target.pdb

Modify the leap input file (./lib/run_leap.sh)

set mol box {50 50 50 } # Size of the PBC box

Modify the preference file (./lib/preference.sh)

The GMX and GMX_MPI is set for the full path for the Gromacs binary.

You can specify scoring script as SCORING_SCRIPT in ./lib directory. If you need your original scoring script, place it in the ./lib directory.

Usage

nohup ./lib/run_jobs.sh {TRIAL_NAME} &

To restart PaCS-MD,

cd {TRIAL_directory}
nohup ./lib/run_pacs.sh {START_CYCLE(not include the exist cycle)} {LAST_CYCLE} {NUMBER_OF_CANDIDATE=10}

To visualize the progress with score,

cd {TRIAL_directory}
python3 ./lib/plot_score.py

This script draw a plot and save a pdf file.

Note

Please cite this paper.

Ryuhei Harada and Akio Kitao. Parallel cascade selection molecular dynamics (PaCS-MD) to generate conformational transition pathway. J. Chem. Phys. (2021)DOI:[https://doi.org/10.1063/1.4813023]

Contributers

• Rikuri Morita*, Ryuhei Harada*.
• Center for Computational Sciences, University of Tsukuba
• morita@ccs.tsukuba.ac.jp, ryuhei@ccs.tsukuba.ac.jp

License

This scripts are under MIT license.