[gmx-users] Inconsistent shifts using multiple bonds type 6

[Berk Hess]

>From: TJ Piggot <t.piggot at bristol.ac.uk>
>Reply-To: Discussion list for GROMACS users <gmx-users at gromacs.org>
>To: Discussion list for GROMACS users <gmx-users at gromacs.org>
>Subject: Re: [gmx-users] Inconsistent shifts using multiple bonds type 6
>Date: Tue, 06 Mar 2007 12:44:16 +0000
>
>Hi Berk,
>
>Thanks for your reply. Yes the latter is the case in my system so i will 
>use pbc=full or remove these 'bonds' that i know will be greater than half 
>of a box dimension

You should remove those bonds, or make the box larger.
pbc=full will work, but the distance vector will change to different
periodic images when an "inter"-molecular periodic image distance
becomes shorter than the true intramolecular distance.

I assume your molecule is properly connected by normal bonds.
A real fix would then be adding
i!=F_HARMONIC &&
to the if statement on line 272 of src/gmxlib/mshift.c
and recompiling Gromacs, such that the harmonic potentials
are not put into the graph.

Berk.

_________________________________________________________________
Hotmail en Messenger on the move http://www.msn.nl/services