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

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!

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

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}
}