[gmx-users] Shell dynamics with vdW terms.

Berk Hess gmx3 at hotmail.com
Wed Jun 10 10:15:32 CEST 2009 



> From: rafapa at us.es
> To: gmx-users at gromacs.org
> Subject: Re: [gmx-users] Shell dynamics with vdW terms.
> Date: Wed, 10 Jun 2009 08:42:26 +0200
> 
> On Tuesday 09 June 2009 18:27:16 Berk Hess wrote:
> > > From: rafapa at us.es
> > > To: gmx-users at gromacs.org
> > > Subject: Re: [gmx-users] Shell dynamics with vdW terms.
> > > Date: Tue, 9 Jun 2009 16:29:39 +0200
> > >
> > > On Tuesday 09 June 2009 10:02:47 Berk Hess wrote:
> > > > Hi,
> > > >
> > > > The equation you refer to is only an initial guess for the
> > > > optimization. The positions of the shells are optimized using the
> > > > forces from the full potential function.
> > > >
> > > > Berk
> > > >
> > > > > From: rafapa at us.es
> > > > > To: gmx-users at gromacs.org
> > > > > Date: Tue, 9 Jun 2009 09:51:13 +0200
> > > > > Subject: [gmx-users] Shell dynamics with vdW terms.
> > > > >
> > > > > Hello,
> > > > >
> > > > > I have some troubles understanding the way used by Gromacs in
> > > > > optimizing the position of shells. In section 3.5.1 the manual says
> > > > > that the force on a shell particle is decomposed into two terms. The
> > > > > first represents the polarization and the second includes the Coulomb
> > > > > and van der Waals interactions. It continues assuming that the second
> > > > > term is constant and derives an analytic expression to calculate the
> > > > > shell position.
> > > > >
> > > > > My troubles arise after reading the paper: J. Phys. Chem. B 2009,
> > > > > 113, 7270-7281 by A. Villa, B. Hess & H. Saint-Martin. In the paper
> > > > > they simulate Lanthanides in water. In Table 2 they assign a 1/r**12
> > > > > term to the ion-shell - water-shell interaction (the A_MIM
> > > > > parameter). I believe that this term is used in eq. 8 of the
> > > > > Supporting Information. Because is vdW like it seems to me that it
> > > > > breaks the previously mentioned assumption of vdW constancy.
> > > > >
> > > > > Could anybody help me?
> > > > >
> > > > > Thanks a lot.
> > > > >
> > > > > Rafael
> > > > >
> > > > > --
> > > > > Dr. Rafael R. Pappalardo
> > > > > Dept. of Physical Chemistry, Univ. de Sevilla (Spain)
> > > > > e-mail: rafapa at us.es
> > > > > _______________________________________________
> > > > > gmx-users mailing list    gmx-users at gromacs.org
> > > > > http://lists.gromacs.org/mailman/listinfo/gmx-users
> > > > > Please search the archive at http://www.gromacs.org/search before
> > > > > posting! Please don't post (un)subscribe requests to the list. Use
> > > > > the www interface or send it to gmx-users-request at gromacs.org.
> > > > > Can't post? Read http://www.gromacs.org/mailing_lists/users.php
> > > >
> > > > _________________________________________________________________
> > > > See all the ways you can stay connected to friends and family
> > > > http://www.microsoft.com/windows/windowslive/default.aspx
> > >
> > > Dear Berk,
> > >
> > > thanks for the quick answer.
> > >
> > > I am having lot of troubles trying to minimize a cation+1 water molecule
> > > using Gromacs 4.0.5. The minimization proceeds without warnings but at a
> > > certain point it seems that the shell minimizer does not work ok. The
> > > kind of structures I am obtaining are:
> > > Generated by trjconv : Cf in Water t=   9.00000
> > >     6
> > >     1Cf_s    Cf    1   1.49566   1.61167   1.73228
> > >     1Cf_s    Ms    2   1.49450   1.62107   1.72892
> > >     2SOL     OW    3   1.51913   1.41088   1.80113
> > >     2SOL    HW1    4   1.51596   1.36681   1.89843
> > >     2SOL    HW2    5   1.53786   1.31814   1.75093
> > >     2SOL     MW    6   1.51789   1.42144   1.79730
> > >    3.10740   3.10740   3.10740
> > > Generated by trjconv : Cf in Water t=  10.00000
> > >     6
> > >     1Cf_s    Cf    1   1.49838   1.58947   1.74003
> > >     1Cf_s    Ms    2   1.49587   1.60861   1.73299
> > >     2SOL     OW    3   1.51772   1.42189   1.79765
> > >     2SOL    HW1    4   1.51568   1.37036   1.89631
> > >     2SOL    HW2    5   1.53733   1.32238   1.75035
> > >     2SOL     MW    6   1.51602   1.43730   1.79166
> > >    3.10740   3.10740   3.10740
> > > Generated by trjconv : Cf in Water t=  11.00000
> > >     6
> > >     1Cf_s    Cf    1   1.50510   1.52690   1.75928
> > >     1Cf_s    Ms    2   1.50102   1.57015   1.74742
> > >     2SOL     OW    3   1.51647   1.43148   1.79236
> > >     2SOL    HW1    4   1.51461   1.38178   1.89116
> > >     2SOL    HW2    5   1.53582   1.33543   1.74711
> > >     2SOL     MW    6   1.50797   1.50425   1.77168
> > >    3.10740   3.10740   3.10740
> > >
> > > I see no warning or error in the md.log file albeit the Coulomb energy is
> > > rather large. The run continues at to t=50.000 with the message:
> > >
> > > Stepsize too small, or no change in energy.
> > > Converged to machine precision,
> > > but not to the requested precision Fmax < 0.1
> > >
> > > Could I sent you the input files in order to get some advice?
> > >
> > >
> > > Best regards,
> > >
> > > Rafael
> > >
> > > --
> > > Dr. Rafael R. Pappalardo
> > > Dept. of Physical Chemistry, Univ. de Sevilla (Spain)
> > > e-mail: rafapa at us.es
> >
> > At this point I have no clue if there might be something wrong at all.
> > There are many questions to ask:
> > What energies would you expect?
> > Have you checked if the minimized structure looks reasonable?
> 
> Yes, I have checked the structures. During the minimization the charges on the 
> cation nucleus and the mobile charge on the oxygen collapse.
> 
> > Is this a force field from the literature that it proven to work with this
> > water model? Polarizable models are tricky, if there is not enough
> > repulsion or the spring constant is too weak you can easily get a
> > polarization catastrophe.
> >
> > Berk
> >
> 
> The force field is an in house development. We have tried the minimization 
> using DL-POLY and a Monte Carlo code and the final result seems sensible. The 
> troubles are with Gromacs. We have checked that the initial energy is the same 
> between Gromacs and DL-POLY.
> 
> Thanks again.

But this is really pure energy minimization and not shell minimization during MD,
or did I misunderstand this?
During pure energy minimization (integrator steep or cg) the shells are treated
as any other particle, so there should be no particular shell issues at all.
But you could still have a polarization catastrophe. It could be that the initial
EM step (mdp parameter em_step) is too large.

Berk


_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE!
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20090610/e7a7b315/attachment.html>

More information about the gromacs.org_gmx-users mailing list