[gmx-users] grompp and nonbonded interactions
Berk Hess
gmx3 at hotmail.com
Tue Oct 12 09:59:17 CEST 2004
>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] grompp and nonbonded interactions
>Date: Fri, 8 Oct 2004 16:59:52 -0400 (EDT)
>
>
>Hello, I have some questions on how grompp builds nonbonded interactions.
>
>1) when the charge on an atom type is specified by a given ff.itp file,
>will the charge placed on an atom given in a molecule.itp file take
>precedence? for example, with
>
>#include ffgmx.itp
>#include molecule.itp
>
>if the charge on a particular atom in the molecule.itp file is different
>from the ffgmx.itp file, which file will the atom's charge be taken from.
>
The charge is taken from the atoms section in the moleculetype definition.
If no charge is present there (empty column) then the charge is taken
from the atomtypes section.
>2) if gen-pairs is set to yes (and fudgeQQ=fudgeLJ=0.5 for example) under
>[defaults] and a molecule.itp file specifies [pairtypes] and [pairs], will
>fudgeQQ be taken as 0.5 for 1-4 electrostatics while the fudgeLJ=0.5 is
>overridden by the 1-4 LJ interactions specified in [pairtypes] and [pairs]?
>In other words, can fudgeLJ be overridden by [pairtypes] even when
>gen-pairs is set to yes?
There have been several mails on this subject the past months.
I have clarified this for the next release of the manual.
fudgeQQ is always used, irrespective of the gen-pairs setting.
For atoms in the pairs section first the parameters in the pairs
section are checked. If non are present, the ones in pairtypes
are used. If the atom pair is not present there and gen-pairs=yes
the pairs are generated from the non-bonded parameters using fudgeLJ,
if gen-pairs=no a warning is generated and zeros are used.
Berk.
_________________________________________________________________
MSN Search, for accurate results! http://search.msn.nl
More information about the gromacs.org_gmx-users
mailing list