WepyAnalysis is a modular toolkit for analyzing data generated from Weighted Ensemble (WE) simulations using the Wepy framework.
The codebase is organized into five main components:
featurization/: Tools for extracting structural features from WE datadataset/: Code for generating datasets from WE datamsm/: Building Markov State Models (MSMs) and performing kinetic analysisexample/: Example scripts for running simulations with Wepy and building MSMs
This repository is under active development and intended for researchers working with WE data, especially those using the Wepy framework.
We recommend installing WepyAnalysis with a Python package manager such as conda or mamba. The package is tested and fully compatible with Python 3.12, and we strongly encourage using python>=3.10 for compatibility.
conda create -n wepy-analysis python=3.12
conda activate wepy-analysis
Once your python environment is ready, wepy-analysis can be installed with pip as follows:
pip install git+https://github.com/ADicksonLab/wepy-analysis
which will also install all dependencies. The installation procedure takes less than a minute to complete at a local desktop.
Wepy (https://github.com/ADicksonLab/wepy) >= 1.2
geomm (https://github.com/ADicksonLab/geomm) >= 0.3
csnanalysis (https://github.com/ADicksonLab/CSNAnalysis) >= 0.6.0
numpy >= 2.3.1
scipy >= 1.16.0
h5py >= 3.14.0
mdtraj >= 1.11.0
scikit-learn >= 1.7.0
deeptime >= 0.4.5
- Example dataset files can be found at current Zenodo DOI
- This repository is a part of the preprint "Determinants of Improved CGRP Peptide Binding Kinetics Revealed by Enhanced Molecular Simulations" and can be used to build MSMs explained in the paper.