Skip to content

Analyze correlated motions in MD trajectories with only a few lines of Python.

License

Notifications You must be signed in to change notification settings

msmbuilder/mdentropy

Repository files navigation

Build Status Code Health PyPI version License Documentation DOI

MDEntropy

MDEntropy is a python library that allows users to perform information-theoretic analyses on molecular dynamics (MD) trajectories. It includes methods to calculate:

  • Bias-Corrected Entropy
  • Conditional Entropy
  • Mutual Information
  • Normalized Mutual Information
  • Conditional Mutual Information
  • Normalized Conditional Mutual Information

Documentation

Full documentation can be found at http://msmbuilder.org/mdentropy/. For information about installation, please refer to our Installation page.

We also have example notebooks with common use cases for MDEntropy. Please feel free to add your own as a pull-request!

Requirements

  • python>=3.4
  • numpy>=1.10.4
  • scipy>=0.17.0
  • scikit-learn>=0.17.0
  • msmbuilder>=3.5.0
  • nose (optional, for testing)

Citing

Please cite:

@article{mdentropy,
  author       = {Carlos X. Hern{\'{a}}ndez and Vijay S. Pande},
  title        = {{MDEntropy: Information-Theoretic Analyses for Molecular Dynamics}},
  month        = nov,
  year         = 2017,
  doi          = {10.21105/joss.00427},
  url          = {https://doi.org/10.21105/joss.00427}
}