[gmx-users] Regarding Position restraint and freezegroups
jalemkul at vt.edu
Thu Aug 29 22:59:57 CEST 2013
On 8/29/13 3:25 PM, rahul seth wrote:
> I have been performing simulations with nanotubes. I read several entries
> in the gromacs users forum regarding the correctness of the usage of the
> freeze groups in simulations.
> However, I am trying to perform NPT simulations at two different
> temperatures using berendsen and parrinello-rahman barostat respectively
> freezing the carbon nanotubes and position restraining them at the same
Why would this be necessary? Both position restraints and freezegrps restrict
movement; under what circumstances are both necessary? Hopefully, from a very
recent thread on this topic, it should be abundantly clear that the combination
of pressure coupling and frozen atoms is fundamentally unsound.
> I am using refcoord-scaling = com
> I do not have any information regarding the bonded interactions of the
> carbon nanotubes since I am freezing and position restraining them anyways.
Why not? CNT simulations are not novel; plenty of people have used reasonable
parameters for bonded and nonbonded interactions.
> Interestingly, although Berendsen barostat keeps the diameter constant at a
> fixed value the parrinello-rahman barostat changes the diameter of the
> nanotube quite significantly at the higher temperatures. Can anyone explain
> what is going on?
The Parrinello-Rahman barostat allows for wider oscillations in the pressure,
and higher temperatures exacerbate the outcome. I suspect that the combination
of a non-physical restraint/frozen combination and the natural behavior of the
barostat causes artifacts like this.
Justin A. Lemkul, Ph.D.
Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201
jalemkul at outerbanks.umaryland.edu | (410) 706-7441
More information about the gromacs.org_gmx-users