pycolvars 0.1.0

Python tools to post-process and analyze metadynamics / collective variable data

pycolvars

Python tools to post-process and analyze collective variable (CV) data from metadynamics and umbrella sampling simulations: reading colvars/PLUMED-style trajectory files, reconstructing free energy surfaces (PMF), and finding minimum energy paths across them.

This package was originally developed alongside sea_urchin and later split out as a standalone tool.

Installation

pip install -e .

Minimum energy path algorithm

pycolvars.min_energy_path finds the path between two basins on an energy surface that minimises the maximum energy barrier crossed (the minimax / bottleneck path). It uses the Minimum Spanning Tree of the grid connectivity graph (edge weight = max endpoint energy), which provably yields the optimal minimax path. k-th alternative paths are found by iteratively removing the bottleneck edge of the previous best path and recomputing.

This is a native implementation using networkx. It replaces an earlier vendored copy of MEPSAnd by Marcos-Alcalde et al.