[gmx-users] topology defaults directive

[Berk Hess]

>From: "David L. Bostick" <dbostick at physics.unc.edu>
>Reply-To: Discussion list for GROMACS users <gmx-users at gromacs.org>
>To: <gmx-users at gromacs.org>
>Subject: [gmx-users] topology defaults directive
>Date: Wed, 13 Oct 2004 17:35:58 -0400 (EDT)
>
>
>Dear all,
>
>In ffgmx.itp, the defaults directive confuses me. This file has the
>following defaults:
>
>[ defaults ]
>; nbfunc        comb-rule       gen-pairs       fudgeLJ fudgeQQ
>   1             1               no              1.0     1.0
>
>This confuses me for 2 reasons:
>
>1) The manual says that fudgeLJ and fudgeQQ are not
>taken unless gen-pairs is yes. So, since in this case gen-pairs is no, does
>this mean that fudgeLJ=fudgeQQ=1.0 doesn't really do anything?

As I explained in my answer to your previous mail (and contrary to what
the current manual says) fudgeQQ is always used for all pair interactions.
In this case fudgeLJ is not used.

>
>2) I see that in ffgmxnb.itp [ pairtypes ] are listed for 1-4 LJ
>interactions. So if in a molecule.itp file, I have
>
>[ moleculetype ]
>; Name   nrexcl
>molecule     3
>
>the 1-4 intramolecular LJ interactions will be given by the [ pairtypes ],
>but what about the 1-4 coulomb interactions? With nrexcl = 3 and
>gen-pairs=no, does this mean that the 1-4 coulomb interaction will be
>completely excluded?

No.
nrexcl only pertains to the normal VdW and Coulomb interactions.
These are excluded up to the third neighbor.
Normally the moleculetype section of the topology will contain
a pairs section which lists all the pair interactions explicitly.
One can use LJ parameters for the pairs which are different from
the normal LJ parameters. The charges for the pairs are the same
as the normal ones, but fudgeQQ can be used to scale the interaction.
If such a pairs section is not present there are no pair interactions.

Berk.

_________________________________________________________________
MSN Search, for accurate results! http://search.msn.nl