[gmx-users] Pressure coupling with stochastic integrator for triclinic box

[Haiyang Zhang]

On Tue, Aug 26, 2014 at 5:15 PM, Mark Abraham <mark.j.abraham at gmail.com>
wrote:

> Hi,
>
> 1) GROMACS 4.5.5 is pretty old, so we aren't going to spend time trouble
> shooting things that might have been fixed since then. Please try out
> GROMACS 5, where there is a brand new SD integrator, and faster performance
> all round.
>
Will try the new version.

>
> 2) 80 +/- 3000 is not terribly significant. What gives you greater
> confidence about your Berendsen result?

Isotropic Berendsen coupling gives a pressure of 1 +/- 2711 bar, which
seems better than PR coupling (15 +/- 3516 bar), although both methods have
large fluctuations.

> It seems like you need either a
> larger system or a lot more sampling. See also
> http://www.gromacs.org/Documentation/Terminology/Pressure
>
> Mark
>
>
>
>
> On Tue, Aug 26, 2014 at 10:56 AM, Haiyang Zhang <
> haiyang.logan.zhang at gmail.com> wrote:
>
> > On Tue, Aug 26, 2014 at 4:34 PM, Mark Abraham <mark.j.abraham at gmail.com>
> > wrote:
> >
> > > Hi,
> > >
> > > It's hard to help without actual details of your P-R settings and
> GROMACS
> > > version!
> >
> > Pcoupl                     = Parrinello-rahman
> > Pcoupltype               = anisotropic
> > tau_p                       = 15
> > compressibility          = 4.5e-5 4.5e-5 4.5e-5 4.5e-5 0 0
> > ref_p                         = 1 1 1 1 0 0
> >
> > GROMACS version: 4.5.5
> >
> > Do things work with non-stochastic integrators?
> > >
> > For anisotropic, I have turned on/off off-diagonal pressure components.
> The
> > best setting is listed above. Using this setup with a normal md
> integrator,
> > I got a pressure of -1 +/- 2716 bar, which seems OK.
> >
> > However, for SD integrator, P is 80 +/- 3405 bar. When turning off all
> > off-diagonal components, P is 13 +/- 3520 bar. If using isotropic PR with
> > SD, P is 15 +/- 3516. All the simulations were done for 5 ns.
> >
> > Thank you very much.
> >
> > >
> > > Mark
> > >
> > >
> > > On Tue, Aug 26, 2014 at 10:00 AM, Haiyang Zhang <
> > > haiyang.logan.zhang at gmail.com> wrote:
> > >
> > > > Dear Gmx Users,
> > > >
> > > > Recently, I used Parrinello-rahman (PR) pressure coupling to
> simulate a
> > > > periodic metal-organic framework (MOF) system. The box is triclinic
> > with
> > > > angles of 90, 90, 120 degree.The system was simulated using a
> > stochastic
> > > > integrator. Isotropic and anisotropic pressure coupling have been
> > tested,
> > > > and the results show that PR method does not give correct pressures.
> > > >
> > > > However, a more robust berendsen (BD) method can give the reasonable
> > > > pressures for both isotropic and anisotropic coupling. PR is known to
> > > > perform better than BD, so I prefer to use PR to couple poressure.
> > > >
> > > > Did someone have any clue fixing this problem ? Thanks in advance.
> > > >
> > > > With my best wishes, Haiyang
> > > > --
> > > > Gromacs Users mailing list
> > > >
> > > > * Please search the archive at
> > > > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > > > posting!
> > > >
> > > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > > >
> > > > * For (un)subscribe requests visit
> > > > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users
> or
> > > > send a mail to gmx-users-request at gromacs.org.
> > > >
> > > --
> > > Gromacs Users mailing list
> > >
> > > * Please search the archive at
> > > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > > posting!
> > >
> > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > >
> > > * For (un)subscribe requests visit
> > > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> > > send a mail to gmx-users-request at gromacs.org.
> > >
> > --
> > Gromacs Users mailing list
> >
> > * Please search the archive at
> > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > posting!
> >
> > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >
> > * For (un)subscribe requests visit
> > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> > send a mail to gmx-users-request at gromacs.org.
> >
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>