Does minimizeEnergy work across the PBC?

The functionality of OpenMM will (eventually) include everything that one would need to run modern molecular simulation.
POST REPLY
User avatar
lewis martin
Posts: 63
Joined: Tue Mar 06, 2018 8:56 pm

Does minimizeEnergy work across the PBC?

Post by lewis martin » Wed May 09, 2018 6:57 pm

Hi all,
I have a pre-equilibrated cell, to which I added a ligand that dips over the edge of the cell, forming bad contacts with water.

minimizeEnergy() is unable to remove the contacts - does this function work across the boundary? It can be fixed by increasing the z direction of the PBC (and recovering with a barostat if necessary), just wondering how it works.

Code: Select all

initial energy =  2135329600067.072 kJ/mol
after energy =  2298638365251.072 kJ/mol

User avatar
Peter Eastman
Posts: 2544
Joined: Thu Aug 09, 2007 1:25 pm

Re: Does minimizeEnergy work across the PBC?

Post by Peter Eastman » Wed May 09, 2018 7:02 pm

If your initial conformation is too bad, the minimizer won't be able to fix it. Minimization is for making small changes to the atom positions. If you have two molecules right on top of each other, it's unlikely to be able to fix that. In this case, your initial energy is over 2 trillion kJ/mol, which is way beyond what you can reasonably expect it to fix.

You need to identify which water molecules the ligand is clashing with and remove them. Once that's done, things should work a lot better.

POST REPLY