SynD

Library for creating and using synthetic dynamics generators.

• Free software: MIT license
• Documentation: https://synd.readthedocs.io .

Installation

SynD can be installed with pip install SynD

Model Hosting

Serialized models can be stored/retrieved from an S3 object store using the functionality in synd.hosted module.

To use hosted models, set the MINIO_ACCESSKEY and MINIO_SECRETKEY environment variables, or create an .env from .env-template.

Usage

Some example scripts are provided in examples/.

Sample Scripts

• examples/data/simple_model.py parameterizes a SynD generator of a simple 3-state Markov system.
• generate_markov_trajectory.py Creates a MarkovGenerator, and generates some trajectories.
• create_model.py Creates a MarkovGenerator from simple_model.py, and serializes it.
• use_saved_model.py Loads the MarkovGenerator created by create_model.py and generates some short trajectories.
• examples/westpa Example of simulating Trp-cage unfolding, using WESTPA with a 10,500 state discrete Markov SynD propagator.

Features

• Trajectory generation from synthetic Markov models
• Model serialization/storage/retrieval via S3

Todo

• Add MD-specific Markov model, that backmaps MDTraj structures
• Make Markov propagation more efficient
• (?) Implement Markov models using deeptime
• Add progress bar to trajectory generation, using Rich

Credits

This package was created with Cookiecutter and the audreyr/cookiecutter-pypackage project template.