[gmx-users] QM/MM simulations

Wed Sep 7 21:14:27 CEST 2016

I'm still working on the same problem I mentioned below. I'm trying to
verify that I'm getting the right information out of Gromacs to send over
to Gaussian (the input.com file). I might be misreading the "gau" script (
http://wwwuser.gwdg.de/~ggroenh/gau), but I think it's saying that there
should be two coordinate sections: the first section should contain the
cartesian coordinates for the QM atoms, while the second section should
contain the coordinates for the MM atoms as a field of point charges.

I was attempting to simulate octanol in water solvent. The octanol atoms
are to be treated by QM. I am pasting the body of the file in below. I can
see that the MM coordinates were not included in the file. Has anyone seen
this problem before?

%chk=input
%mem=1215752192
%subst l701 /fslhome/crk9/gromacs_src/modlinks/l701
%subst l301 /fslhome/crk9/gromacs_src/modlinks/l301
%subst l9999 /fslhome/crk9/gromacs_src/modlinks/l9999
#P  DFT/3-21G*
Nosymm units=bohr
FORCE Punch=(Derivatives) iop(3/33=1)

input-file generated by gromacs

 0 1
  1 62.2773850  18.0042321  12.8730289
  8 62.7275929  19.6615469  13.5481707
  1 60.8732094  22.8994305  14.4701237
  1 61.3726754  22.4880551  11.2309538
  6 60.8813102  21.4923398  12.9658582
  1 58.3447012  18.6825131  11.6416004
  1 56.8749667  21.5528031  12.3468668
  6 58.2902482  20.2147180  13.0171607
  1 55.7188751  17.8600687  15.1328293
  1 58.8560031  18.0276177  16.2996878
  6 57.3035799  19.1226107  15.5035981
  1 58.1924753  22.3120539  17.8026816
  1 55.1397300  22.2786189  16.7532522
  6 56.5899874  21.0450109  17.5392182
  1 54.3066933  18.3180812  19.3290486
  1 57.3206195  18.3784025  20.4499142
  6 55.7745716  19.6343339  19.9253595
  1 56.0312610  22.6056363  22.6202497
  1 52.9728344  22.2488265  21.3282668
  6 54.6149342  21.2279325  22.0382414
  1 55.1450509  17.9434118  24.4976697
  1 51.9915096  19.0228103  24.2290069
  6 53.9589936  19.6208130  24.3481110
  1 55.7838078  22.2669272  26.6621999
  1 54.5106599  19.8634027  28.4811908
  1 52.4125626  22.0430645  27.4308446
  6 54.1258397  21.0589621  26.8486065

--
Clinton King
Graduate Student
Chemistry Department
Brigham Young University

On Thu, Sep 1, 2016 at 7:03 AM, <gromacs.org_gmx-users-request
@maillist.sys.kth.se> wrote:

>
> Today's Topics:
>
>    1. Re: QM/MM simulations (Mark Abraham) (Groenhof, Gerrit)
>       (Clinton King) (Groenhof, Gerrit)
>     ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 1 Sep 2016 10:08:19 +0000
> From: "Groenhof, Gerrit" <ggroenh at gwdg.de>
> To: "gromacs.org_gmx-users at maillist.sys.kth.se"
>         <gromacs.org_gmx-users at maillist.sys.kth.se>
> Subject: Re: [gmx-users] QM/MM simulations (Mark Abraham) (Groenhof,
>         Gerrit) (Clinton King)
> Message-ID:
>         <858A7947BC0FE04DA05E1786A6D51D453307FF99 at um-excdag-a05.um.gwdg.de
> >
> Content-Type: text/plain; charset="us-ascii"
>
> No, it also works in gmx5.
>
> Don't use domain decomposition or the verlet scheme (use group based
> cut-off instead) and run a single thread (-nt 1)
>
>
> Hope it works of you too,
>
> Best,
>
> Gerrit
>
>
>
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Wed, 31 Aug 2016 13:28:56 -0600
> From: Clinton King <clintonking36 at chem.byu.edu>
> To: gromacs.org_gmx-users at maillist.sys.kth.se
> Subject: [gmx-users] QM/MM simulations (Mark Abraham) (Groenhof,
>         Gerrit)
> Message-ID:
>         <CABUMFLLgMSpeEX72t1nveb3tTTRWhyjgbuk=k4KHZF7Q0M97VQ at mail.gm
> ail.com>
> Content-Type: text/plain; charset=UTF-8
>
> Is is true that the gau script available at http://wwwuser.gwdg.de/~
> ggroenh/qmmm.html <http://wwwuser.gwdg.de/~ggroenh/qmmm.html> only works
> with gromacs versions < 5?
>
> --
> Clinton King
> Graduate Student
> Chemistry Department
> Brigham Young University
>
> >
> >
> >    5. Re: QM/MM simulations (Mark Abraham) (Groenhof, Gerrit)
> >
> > Hi,
> >
> > Unless there are no atoms in the QM region, the QM energy should not be
> > zero.
> >
> > Can you confirm that mdrun writes an input file (input.com)? and
> gaussian
> > is executed?
> >
> > If not, has gromacs been compiled with QMMM support
> > (GMX_QMMM_PROGRAM:STRING=gaussian)?
> >
> > Best,
> >
> > Gerrit
> >
> >
> > Hi,
> >
> > Sadly, most of the QM/MM interfaces have been lacking a maintainer on the
> > GROMACS side for quite a number of years. You should definitely be
> > following http://wwwuser.gwdg.de/~ggroenh/qmmm.html closely.
> >
> > Mark
> >
> > On Mon, Aug 29, 2016 at 9:36 PM Clinton King <clintonking36 at chem.byu.edu
> >
> > wrote:
> >
> > > I'm performing a QM/MM (using Gaussian 09 and Gromacs 5.1.2) simulation
> > of
> > > a single molecule of octanol in a box of water. In examining the
> standard
> > > output, it appears that call to Gaussian is proceeding as expected, but
> > > looking at the log file, it doesn't appear that quantum energy is being
> > > communicated correctly, ie the output looks like the following:
> > >
> > >
> > >
> > >            Step           Time         Lambda
> > >            7000        7.00000        0.00000
> > >
> > >    Energies (kJ/mol)
> > >         LJ (SR)          Disper. corr.       Coulomb (SR)   Coul.
> recip.
> > >      Quantum En.
> > >     2.45033e+04   -1.74106e+02   -1.86935e+05    6.68753e+01
> > 0.00000e+00
> > >       Potential          Kinetic En.         Total Energy
> Temperature
> > >      Pres. DC (bar)
> > >    -1.62539e+05    3.00966e+04   -1.32443e+05    2.97582e+02
> >  -2.34549e+01
> > >  Pressure (bar)   Constr. rmsd
> > >    -7.37300e+01    8.07450e-06
> > >
> > >
> > > Notice that the entry for Quantum En. is 0.000.
> > >
> > > Has anyone else seen this problem before? If so, what did you do about
> > it?
> > >
> > > --
> > > Clinton King
> > > Graduate Student
> > > Chemistry Department
> > > Brigham Young University
> > > --
>