whampy

whampy is a python3 script that performs a WHAM calculation on umbrella sampling datasets in order to obtain a potential of mean force.

Getting Started

Prerequisites

Disclaimer: program in development

• python3.x must be installed.
• The following python packages must be installed:
  • numpy
  • scipy
  • matplotlib
• These can be installed with pip

whampy instructions

The whampy program computes the potential of mean force of an umbrella sampling simulation using a minimization of a log-likelihood function of the probability distribution in 1D.

The execution of the program is as follows:

python3 wham.py [-h] [-s] [-i INPUT] [-o OUTPUT]

where the INPUT file is a plain text file with the format as specified in wham.in. The trajectory files sourced from the paths found in the input file are assumed to be in the two-column NAMD .traj format as shown examplarily in the traj/window*.traj files.

For more information about the options, the [-h] optional flag brings up the help text.

Example: Using WHAM in constant potential simulations

I've adapted the original whampy code of enfo14 (https://github.com/enfo14/whampy) (details are found in the subfolder wham/) to account for a linear bias potential activated by setting the linear bias flag in the metafile.

• Use the examplary data provided in traj/window*.traj stemming from a constant potential simulation of graphite electrodes at various applied electric potentials in contact with 1.5M BMIPF₆ in ACN.
• adapt simulation parameters in wham.in to account for specific needs and your simulation setup. a summary of all available settings is found in /wham/symdata.py
• use flag #linear = True in the metafiles to activate a linear bias potential. This interprets the second column in the metafile paths definitions as an applied electrostatic potential.
• run wham.py -i wham.in -o wham_output to do the WHAM on the provided data.
• output is at 0.0V. however, one can make use of the fact that generally, the distribution, P(σ), shifts according to applied voltage P(σ|ΔΨ) ∝ P(σ|0V) exp(−σSΔΨ/(kB*T)) (see eq. (3) Merlet, C. et al. The Electric Double Layer Has a Life of Its Own. J. Phys. Chem. C 118, 18291–18298 (2014). This is done in the Jupyter notebook.
• if MC is activated with #num_mc_runs = 2000 to estimate the confidence, a file containing the bootsrapped PMFs for post processing is stored under bootstraps.pmf. Details of this are found in the accompanying Jupyter notebook.