VMD-L Mailing List

From: Vermaas, Joshua (Joshua.Vermaas_at_nrel.gov)
Date: Thu Aug 31 2017 - 11:59:40 CDT

Hi Williams,

Alright, this is going to be a fun problem. For one thing, taking a perfectly straight channel is going to be wrong, since an actual ion will religiously follow any non-linearity in the potential. Which means you basically first need to compute the minimum free energy pathway between two endpoints, and use that non-straight pathway as your reaction coordinate. As far as I'm aware, there is no inbuilt tool within VMD to compute this straightaway. Other tools might exist (so googling is probably useful), but if I had to implement it myself, python and its libraries are your friend. MDAnalysis comes with a library to load dx files, (see https://github.com/MDAnalysis/GridDataFormats), and then setting up edgeweights so that Dijkstra's algorithm gives you the path between your end points. This would be setup either with networkX (https://networkx.readthedocs.io/en/stable/reference/algorithms.shortest_paths.html) or scipy.

-Josh

On 08/30/2017 04:51 PM, Williams Miranda wrote:

Hello

I have generated the electrostatic potential map for my system using PMEpot vmd plugin. But I would like to obtain the electrostatic potential profile along the transmembrane pore of the channel axis. How could I extract this information from the *dx file generated by PMEpot?

Thanks

Williams