[gmx-users] how to remove an energy contribution in gromacs?

[Berk Hess]

>From: "Jones de Andrade" <johannesrs at gmail.com>
>Reply-To: Discussion list for GROMACS users <gmx-users at gromacs.org>
>To: "Discussion list for GROMACS users" <gmx-users at gromacs.org>
>Subject: [gmx-users] how to remove an energy contribution in gromacs?
>Date: Mon, 6 Aug 2007 09:49:26 -0300
>
>Hi all.
>
>Well, I'm trying to calculate an <E> energy from my simulations, where <E>
>here stands only for the sum of *intermolecular* vdw and coulombic terms,
>and LJ(1-4) and coul(1-4). This means that I should find a way to exclude
>any "1-5", "1-6", "1-7" and so on contributions for coulombic and vdw
>energies, which I guess are included in the SR terms if the rcut is long
>enough.
>
>I though on doing it by use of a modified topology (with an arbitrary
>increased "nexcl" term) after the simulation is completed just for a
>"-rerun" job, and then use the g_energy program to calculate these
>contributions.
>
>Unfortunatelly, for some reason I'm not getting the correct values by large
>differences. Properly saying, only doing the above described procedure is
>yelding me even modifications in the bond, angle and dihedral terms of my
>simulations, which should not vary in a rerun with new nexcl values.
>Moreover, all energy terms with no exception seem to be only divided by a
>factor around 10., in comparison with the simulations with the original
>nexcl.
>
>I'm pretty sure I'm missing something here, but I have no idea of what 
>could
>it be. Could someone please help me with this matter somehow?
>
>Thanks a lot in advance.

This seems the right approach to me.

I don't know what you are doing exactly.
But are you aware that the averages in g_energy and in the log file
are using every integration step, whereas with rerun you only get values
for every frame in your trajectory?
I think the averages of mdrun -rerun are still incorrect, you need to use
the average of the values in the energy.xvg file.

Also obtaining accurate potential energy averages for even small systems
requires a LOT of statistics.

Berk.

_________________________________________________________________
Play online games with your friends with Messenger 
http://www.join.msn.com/messenger/overview