gamda: GPU-Accelerated Molecular Dynamics Analysis

gamda, a python library which utilizes the CUDA-enable GPU to accelerate the analysis of molecular dynamics (MD)

Dependency

1. cupy

cupy is the backend of the CUDA kernels for analysis

2. MDAnalysis

MDAnalysis is used to handle the reading of MD trajectories and the selection of atoms

Installation

from pypi

pip install gamda

from gitee

git clone https://gitee.com/gao_hyp_xyj_admin/gamda.git
cd gamda
pip install .

Unittest

git clone https://gitee.com/gao_hyp_xyj_admin/gamda.git
cd gamda
cd unittest
python -m unittest

Usage

A simple exampe:

# Here, we create a xyz file as an example
with open("test.xyz", "w") as f:
    f.write("3\n3\nO 1.11 2.22 3.33\nH 3.33 2.22 1.11\nH 2.22 2.22 2.22")

# Import the package
import gamda
import numpy as np

# Import the desired observable
from gamda.observable import PositionZ

# Import the disired analyzer
from gamda.analyzer import Histogrammer

# gamda.Universe is a subclass of MDAnalysis.Universe
u = gamda.Universe("test.xyz")

# Get your AtomGroup in host (CPU)
ag = u.select_atoms("element H")

# Get your AtomGroup in device (GPU)
dag = u.get_dag("atom", ag)

# Initialize your observable
z = PositionZ("z", dag, gamda.Argument("z", np.float32(0)))

# Initialize your analyzer and add it to the universe
zdf = Histogrammer("zdf", 0, 4, 4, z)
u.add_analyzer(zdf)

# Print the source code
print(u.source_code)

# Run
u.run()

# Free the memory
u.free()
del u

# Normalize the result and save
zdf.normalize()
zdf.save("test_zdf.txt")

Reference

1. cupy
   • https://github.com/cupy/cupy
2. MDAnalysis
   • N. Michaud-Agrawal, E. J. Denning, T. B. Woolf, and O. Beckstein. MDAnalysis: A Toolkit for the Analysis of Molecular Dynamics Simulations. J. Comput. Chem. 32 (2011), 2319–2327. doi:10.1002/jcc.21787
   • R. J. Gowers, M. Linke, J. Barnoud, T. J. E. Reddy, M. N. Melo, S. L. Seyler, D. L. Dotson, J. Domanski, S. Buchoux, I. M. Kenney, and O. Beckstein. MDAnalysis: A Python package for the rapid analysis of molecular dynamics simulations. In S. Benthall and S. Rostrup, editors, Proceedings of the 15th Python in Science Conference, pages 98-105, Austin, TX, 2016. SciPy. doi:10.25080/Majora-629e541a-00e