Re: [gmx-users] electrostatic forces

[Mark Abraham]

On 01/22/10, Gianfranco Bocchinfuso <gianfranco.bocchinf...@uniroma2.it> wrote:

Hi,   I would like to calculate the electrostatic forces on the atoms of an Indole group (in a TRP residue). I have performed two reruns of my simulation with two different topologies : a normal topology and a topology with all the Indole charges set equal to 0.  I have also added energygrp_excl = Indole Indole (to delete electrostatic force inside the Indole) and also I have deleted in both topologies all the pairs between atoms belonging to the Indole group (to delete the electrostatic contribution coming from 1-4 interactions inside the Indole group). Furthermore I have used a very simple mdp file (with cut-off, without DispCorr etc.) to avoid contribution coming, for example, from ewald summation. In these conditions I expected that the differences between the forces computed in the two simulations were due only to the electrostatic interactions with the other atoms of protein and with solvent (SPC). Checking the obtained forces I see that atoms of the Indole (i.e. the CD2 and CE2), with charges equal to 0 in both topologies, show different forces in the two simulations. What could be the origin of this difference?

Your description does not sound self-consistent. Why do you have two topologies with indole atoms with zero charges, when your introduction described one with zero charges and one with normal charges?

Note that energy group exclusions exclude non-bonded interactions, not merely electrostatic ones.

A correct procedure for finding the electrostatic part of force components is to do a fully normal mdrun -rerun, and one with zeroed charges for the indole group, and subtract the force components so produced. Use gmxdump to compare your two .tpr files for what the actual differences are. Alternatively, alter a topology to have no bonded interactions, and make a version of the ffxxxnb.itp file with zero VDW parameters, grompp that, and do one mdrun -rerun with suitable energygroup-exclusions to make it run fast and correctly.

Whether that means anything in the context of an MM forcefield that wasn't parametrized to reproduce such components is unknown.

Mark

-- 
gmx-users mailing list    gmx-users@gromacs.org
http://lists.gromacs.org/mailman/listinfo/gmx-users
Please search the archive at http://www.gromacs.org/search before posting!
Please don't post (un)subscribe requests to the list. Use the 
www interface or send it to gmx-users-requ...@gromacs.org.
Can't post? Read http://www.gromacs.org/mailing_lists/users.php