[gmx-users] grompp error:Invalid order for directive defaults

[David Mobley]

Hi,

I don't think you can use two force fields at the same time, and I
think this would be a bad idea even if you could. That is, force
fields are parameterized to work well with themselves, NOT with other
force fields. That's why, for example, you don't model, say, the alpha
helical part of a protein using a force field that does particularly
well at modelling alpha helices and the beta sheet part using a force
field that does particularly well for beta sheets. The results you got
would probably make no sense.

What you need is to generate parameters for your small molecule that
are appropriate for the force field you are using. If you are using
OPLS, then you need OPLS parameters for your ligand. There are some
utilities, I think, that can automatically generate OPLS parameters
for many ligands. I haven't used them personally, but Schrodinger has
one, I think. Suggest using Google. If you can't turn anything up,
contact me again and I can probably dig up the name of it from
somewhere.

Otherwise, I would suggest using the AMBER FF, which is available as a
port to GROMACS from the Pande group. It is possible to generate
ligand parameters for that in a fairly automated way using antechamber
and gaff and a conversion script we have.

David Mobley
UCSF


On 6/12/06, Dongsheng Zhang <dong at pampas.chem.purdue.edu> wrote:
> Dear David,
>
> Thank you for your help. It makes sense now. However, it leads me to ask
> you another question:
>
> My poly.itp is got from PRODRG server. It has gromos type force field
> paramters. Do you think my cnt.itp has to be gromos type force field
> format as well (atomtype, bondtype etc)? In fact I used oplsaa to get my
> cnt.itp, that's the reason I need two defaults sections.
>
>
> Best Wishes!
>
>
> Dongsheng
>
>
> On Mon, 2006-06-12 at 17:20 +0200, David van der Spoel wrote:
> > Dongsheng Zhang wrote:
> > > Dear David,
> > >
> > > Thank you very much for your reply. I am sorry to tell you that I get
> > > the same error mesasge after I try your suggestion. Any other
> > > suggestion?
> >
> >
> > check the manual, check the archives.
> >
> > you can not have two defaults sections.
> >
> > >
> > > I think the question is if we can put two sets of force field parameters
> > > in two itp files which will be included in one topology file.  From the
> > > source code  (topio.c line 381 - 390):
> > >
> > >       if (DS_Check_Order (DS,newd)) {
> > >         DS_Push (&DS,newd);
> > >         d = newd;
> > >       }
> > >       else {
> > >         /* we should print here which directives should have
> > >            been present, and which actually are */
> > >         gmx_fatal(FARGS,"Invalid order for directive %s, file \"%s\", line
> > > %d",dirstr,curfile,curline);
> > >         /* d = d_invalid; */
> > >       }
> > >
> > > It seems gromacs does check if a same directive has been present. Is it
> > > correct?
> > >
> > >
> > > Best Wishes!
> > >
> > >
> > > Dongsheng
> > >
> > >
> > > On Mon, 2006-06-12 at 09:35 +0200, David van der Spoel wrote:
> > >> Dongsheng Zhang wrote:
> > >>> Dear GMX users,
> > >>>
> > >>> I'd like to study the interaction between carbon nanotube and a polymer
> > >>> chain. I have made cnt.itp for the nanotube and poly.itp. Both itp files
> > >>> contains force field parameters (such as [ defaults ], [ atomtypes],
> > >>> [ bondtypes ]), in this way, I don't need to include any force field in
> > >>> my topology file (but it will include *.itp) and it is much easier to
> > >>> modify the parameters. For use nanotube or polymer alone as my system, I
> > >>> can use grompp to process .gro, .itp, .top to get .tpr file. If I use
> > >>> nanotube and polymer together as my system, grompp give me the following
> > >>> error message:
> > >>>
> > >>> Invalid order for directive defaults, file ""poly.itp"", line 16
> > >> divide your itp files into two, force field and molecule description
> > >>
> > >> then in your .top file you include all four. chapter 5 in the manual
> > >>
> > >>> line 16 is [ defaults ], so I am thinking maybe I can't have
> > >>> [ defaults ] in each of two itp files. If I delete [ defaults ] in the
> > >>> second itp file. It will complain [ atomtypes] in the second itp file.
> > >>> It seems I can't put force field parameters in two itp files which will
> > >>> be included in one topology file. Could anyone tell me if my guessing is
> > >>> correct? If so, why does gromacs prevent it?
> > >>>
> > >>> Thank you in advance! If you need more information, I will be more than
> > >>> happy to provide it.
> > >>>
> > >>>
> > >>> Best wishes!
> > >>>
> > >>>
> > >>> Dongsheng
> > >>> _______________________________________________
> > >>> gmx-users mailing list    gmx-users at gromacs.org
> > >>> http://www.gromacs.org/mailman/listinfo/gmx-users
> > >>> 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
> > >>
> > > _______________________________________________
> > > gmx-users mailing list    gmx-users at gromacs.org
> > > http://www.gromacs.org/mailman/listinfo/gmx-users
> > > 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
> >
> >
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> 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
>