[gmx-users] interfacing with Gaussian 09

[Groenhof, Gerrit]

Hi,

>From the gromacs side it seems to work fine. That the GMX_QM_MODIFIED_LINKS_DIR needs to be set is that gromacs still expect them, but when you use the script it does not need them.

If you inspect the script you'll see that it does nothing more than copy the input file created by gromacs into a temp.com, run gaussian with that file and extracts from the default temp.log file the energy and gradients on both QM and MM atoms into a fort.7 file, which gromacs reads in to add these informations to the force and energy arrays.

>From the stderr messages tells you that you have made modifications in the script that are faulty at line 18:

/fslhome/crk9/gromacs//gau: line 18: export: `/fslhome/crk9/gromacs/': not a valid identifier

my guess is you're mixing csh and bash languange. note that the gau script is a bash script

Best,

Gerrit




Gerrit,

Thanks for your suggestions. I am using GROMACS 5.1.2.

After reading through your message, I realized that I was using GAUSS_DIR
and GAUSS_EXE for the location and name of the gau script instead of
GMX_QM_GAUSS_DIR and GMX_QM_GAUSS_EXE. So I changed that, and then I got
the error message from line 227 of qm_gaussian.c : "no
$GMX_QM_MODIFIED_LINKS_DIR, this is were the modified links reside."

I thought that one could use the gau script successfully without modifying
any Gaussian links.

I decided to try setting the GMX_QM_MODIFIED_LINKS_DIR environmental
variable to an empty folder. I am attaching the resulting output (as links
at the end of this message); both what was sent to stdout as well as the
gmx generated log file.

Looking through qm_gaussian.c, as well as the output I have so far, it
appears that modified Gaussian links *are* required for gmx to interface
with Gaussian.

So, the bottom line is that in order for Gromacs to interface with Gaussian
to perform a quantum molecular dynamics computation, I must have access to
the Gaussian source code. Is that correct?

https://drive.google.com/file/d/0B-W4G39j_Icjc21TcTN1S2M0MVk/view?usp=sharing
https://drive.google.com/file/d/0B-W4G39j_IcjZlVPb1NySkUtRDQ/view?usp=sharing

(Please let me know if the links above don't work.)

Thanks for your help

--
Clinton King
Graduate Student
Chemistry Department
Brigham Young University


> Message: 2
> Date: Sun, 12 Jun 2016 05:16:13 +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] interfacing with Gaussian 09 (Mark Abraham)
> Message-ID:
>         <858A7947BC0FE04DA05E1786A6D51D45263D61B2 at um-excdag-a05.um.gwdg.de
> >
> Content-Type: text/plain; charset="us-ascii"
>
> Hi,
>
> Which version of Gmx are you using?
>
> I ask because of a recent bug fix. The bug was that I tried to print the
> value of an environment variable, without checking if that variable was
> set, causing set faulting.
>
> Make sure the GMX_QM_GAUSS_DIR (for Gmx < 5 GAUSS_DIR) are set and point
> to the location of the g09 directory.
>
> IN any case, for normal gaussian runs, the program should not write out
> the LJ.dat. IF the above does not help, you could perhaps send me the input
> (off list) and I can have a look at what goes wrong.
>
> Best,
>
> Gerrit
>
>
>
> Hi,
>
> Please don't subscribe to a digest if you're planning to reply to the
> digest and de-thread the discussion. Gerrit said the other week that gau
> works.
>
> Mark
>
> On Fri, 10 Jun 2016 18:18 Mark Abraham <mark.j.abraham at gmail.com> wrote:
>
> > Hi,
> >
> > Please check out http://wwwuser.gwdg.de/~ggroenh/qmmm.html and
> >
> http://www.gromacs.org/Downloads/Installation_Instructions/compiling_QMMM
> >
> > Mark
> >
> > On Fri, Jun 10, 2016 at 5:13 PM Clinton King <clintonking36 at chem.byu.edu
> >
> > wrote:
> >
> >> I'm having difficulty interfacing GROMACS with Gaussian 09. Is there
> >> anyone
> >> who has successfully done it who is willing to give some advice?
> >>
> >> --
> >> Clinton King
> >> Graduate Student
> >> Chemistry Department
> >> Brigham Young University
> >> --
> >> 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.
> >>
> >
>
>