[gmx-users] pdb2gmx -water-> grompp; grompp error

Paymon Pirzadeh ppirzade at ucalgary.ca
Tue Aug 25 20:12:43 CEST 2009


OK!
here is the .itp file which I generated. I might upload it because it
has passed the initial tests.
 
; This is an itp file to describe water's six-site model by H. Nada and
J.P. J. M. Van der Eerden, J. Chem. Phys. Vol.118, no.16, pp7401-7413
(2003)
; This model is a combination of TIP4P and TIP5P. It has three LJ sites
and 3 Coulomb sites
; O-H bond length is 0.980A, HOH angle is 108.00degrees, LOL angle is
111.00 degrees, O-M and O-L are about 0.230A and 0.8892A respectively

[ defaults ]
; non-bondedtype        combrule        genpairs        FudgeLJ
FudgeQQ         N
1                       2               NO

[ atomtypes ]
;name   mass    charge  ptype   c6      c12
OW      15.9994 0.0     A       0.3115  0.714845562
HW      1.00800 0.477   A       0.0673  0.11541
MW      0.000   -0.866  D       0.00    0.00
LW      0.00    -0.044  D       0.00    0.00

[ moleculetype ]
;molname    nrexcl
SOL            2

[ atoms ]
; nr    atomtype        resnr   residuename     atom    cgnr    charge
1       OW              1       SOL             OW      1       0.0
2       HW              1       SOL             HW1     1       0.477
3       HW              1       SOL             HW2     1       0.477
4       MW              1       SOL             MW      1       -0.866
5       LW              1       SOL             LP1     1       -0.044
6       LW              1       SOL             LP2     1       -0.044

[ settles ]
; OW    function        doh     dhh
1       1               0.0980  0.15856

[ dummies3 ]
; These set of parameters are for M site which can be easily calculated
using TIP4P calculations from tip4p.itp
; So, it will be described as dummy site 3: r(v)= r(i) + a*(r(i)-r(j)) +
b*(r(i)-r(k))
; const = |OH|/{|OH|*cos(HOH/2)} => Due to vector algebra a=b=const/2.
Remember that OM is in the same direction of OH bonds.
; Remember this site is in the same plane of OH bonds; so, its function
1
;
; site  from            function        a               b
4       1  2  3         1               0.199642536     0.199642536

; Now we define the position of L sites which can be obtained from
tip5p.itp
; So, it will be described as dummy site 3out: r(v) = r(i) +
a*(r(i)-r(j)) + b*(r(i)-r(k)) + c*(r(ij)Xr(ik))
; const1 = {|OL|*cos(LOL/2)}/{|OH|*cos(HOH/2)} => Due to vector algebra
|a|=|b|=const/2. since the lone pairs are in opposite direction of OH
bonds, a minus sign is added. This part is similar to M site.
; const2 = {|OL|*sin(LOL/2)}/{|OH|*|OH|*sin(HOH)} => The denominator is
the magnitude of vector product of OH bonds.
; This sites are tetrahedral sites; so, its function 4
;
; site  from            function        a               b
c
5       1  2  3         4               -0.437172388    -0.437172388
8.022961206
6       1  2  3         4               -0.437172388    -0.437172388
-8.022961206

[ exclusions ]
1       2       3       4       5       6
2       1       3       4       5       6
3       1       2       4       5       6
4       1       2       3       5       6
5       1       2       3       4       6
6       1       2       3       4       5

and here is the message I get from grompp:

Program grompp_d_mpi, VERSION 4.0.5
Source code file: topio.c, line: 415

Fatal error:
Syntax error - File sixsitewater.itp, line 12
Last line read:
'[ atomtypes ]'
Invalid order for directive atomtypes
-------------------------------------------------------

I believe the original files of GROMACS need admin permissions which I
do not have on the cluster that I am currently using and I am not sure
if that might not cause troubles for other users. So, I thought how I
can keep the changes local.

Payman


On Tue, 2009-08-25 at 13:53 -0400, Justin A. Lemkul wrote:
> 
> Paymon Pirzadeh wrote:
> > Hello,
> > Just a reminder that I had troubled with combination rules between my
> > water model and GROMACS forcefields. I fixed the problem by adjusting my
> > water model's parameters. BUT, now, I specified the new parameters in
> > my .itp file under [ atomtype ] category and grompp makes errors out of
> > the orders. Now I am using OPLSaa. I suspect again the problem comes
> > from the presence of the second [atomtype ]. How can I keep my new
> > parameters and use them in my water model's itp file? Or should I change
> > these parameters in the OPLSaa.itp file? 
> 
> You can't have a second [atomtypes] directive.  You can add new parameters to 
> ffoplsaanb.itp to account for the atom types you are implementing, though.
> 
> If you want anything more specific, you'll probably have to provide the topology 
> file you're working with, as well as any actual error messages you're seeing.
> 
> -Justin
> 
> > Regards,
> > 
> > Payman  
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > On Wed, 2009-08-19 at 10:20 +1000, Mark Abraham wrote:
> >> Paymon Pirzadeh wrote:
> >>> I checked all the force-fields available in the GROAMCS. None of them
> >>> have the #2 combination rule which matches my water model! Any
> >>> alternatives (rather than changing my water model)?
> >> A water model that isn't compatible with the model of the solutes isn't 
> >> of any value, so it looks like you have put the cart before the horse.
> >>
> >> Mark
> >>
> >>> On Tue, 2009-08-18 at 18:52 -0400, Justin A. Lemkul wrote:
> >>>> Paymon Pirzadeh wrote:
> >>>>> That was very helpful! But I realized a problem. The OPLSaa uses the
> >>>>> combination rule #3 in [ defaults ] while my water model uses #2. Now if
> >>>>> I comment out my water model's combination, I am not sure what I am
> >>>>> going to face! Any comments?
> >>>>>
> >>>> Then I would say that your water model's combination rule is inconsistent with 
> >>>> the method of nonbonded calculations in OPLS.  It appears that the only 
> >>>> difference between combination rules 2 and 3 comes when a specific interaction 
> >>>> is not defined in the [nonbond_params] section.  The specifics of these 
> >>>> combination rules are given in manual section 5.3.3.
> >>>>
> >>>> -Justin
> >>>>
> >>>>> Payman
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> On Tue, 2009-08-18 at 18:19 -0400, Justin A. Lemkul wrote:
> >>>>>> Paymon Pirzadeh wrote:
> >>>>>>> Well,
> >>>>>>> I changed the topology file of the system and manually typed the .itp of
> >>>>>>> my own water model. Just a reminder that I have used my own water
> >>>>>>> model's .itp file successfully in pure water systems. This .itp file
> >>>>>>> starts from [ defaults ] and contains all information needed. But when
> >>>>>>> it comes to grompp, the following error appears:
> >>>>>>>
> >>>>>>> Program grompp_d_mpi, VERSION 4.0.5
> >>>>>>> Source code file: topio.c, line: 415
> >>>>>>>
> >>>>>>> Fatal error:
> >>>>>>> Syntax error - File sixsitewater.itp, line 6
> >>>>>>> Last line read:
> >>>>>>> '[ defaults ]'
> >>>>>>> Invalid order for directive defaults
> >>>>>>>
> >>>>>> The only time that [defaults] should appear is in the force field that you are 
> >>>>>> calling (i.e., ffoplsaa.itp):
> >>>>>>
> >>>>>> http://oldwiki.gromacs.org/index.php/Errors#Invalid_order_for_directive_defaults
> >>>>>>
> >>>>>> -Justin
> >>>>>>
> >>>>>>> I am trying to use this water model with OPLSaa. what kind of directive
> >>>>>>> am I missing? 
> >>>>>>>
> >>>>>>> Payman
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>> On Tue, 2009-08-18 at 16:05 -0400, Justin A. Lemkul wrote:
> >>>>>>>> Paymon Pirzadeh wrote:
> >>>>>>>>> Hello,
> >>>>>>>>> I am trying to use a different water model for my proteins (using oplsaa
> >>>>>>>>> force field). I use the .itp file which I developed, but pdb2gmx does
> >>>>>>>>> not accept it. I am a bit confused on how I can make my water model work
> >>>>>>>>> with OPLSaa. Based on what I saw from spc.itp or other samples, do I
> >>>>>>>>> need to change the names of my particles in my water .itp file by
> >>>>>>>>> including _FF_OPLS?
> >>>>>>>> I don't know how pdb2gmx decides how to recognize water models; it may be 
> >>>>>>>> hard-coded.  In any case, your atom types have to be consistent with OPLS 
> >>>>>>>> standards; as for whether or not you need the #ifdef _FF_OPLS in your .itp file, 
> >>>>>>>> you may not need it if OPLS is the only force field with which it will be used. 
> >>>>>>>>   The #ifdef blocks just prevent a GROMOS system from trying to recognize OPLS, etc.
> >>>>>>>>
> >>>>>>>> -Justin
> >>>>>>>>
> >>>>>>>>> Regards,
> >>>>>>>>>
> >>>>>>>>> Payman 
> >>>>>>>>>
> >>>>>>>>> _______________________________________________
> >>>>>>>>> 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
> >>>>>>>>>
> >>> _______________________________________________
> >>> 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
> >>>
> >> _______________________________________________
> >> 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
> >>
> >>
> > 
> > _______________________________________________
> > 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
> > 
> 




More information about the gromacs.org_gmx-users mailing list