[gmx-users] Error from residues added to rtp file

[Delmotte, Antoine]

Dear Gromacs users,

I am once again requesting your help for the editing of the opls force 
field .rtp and .hdb files.

I have inserted the parameters for a new residue, N-methyl methionine 
(MME). In this molecule, the methyl group is attached to the nitrogen 
atom from the peptide bond, which I think is the origin of my problem.

pdb2gmx works fine and generates the topology files without any error. 
Unfortunately, when I looked at the .gro file generated in PyMol (after 
conversion to pdb with g_editconf), I realized that 3 hydrogens have 
been added to the nitrogen atom, so I effectively get R2-N-H4:

H1 \   / CH3
H2 - N - CA - (rest of the aminoa acid)...
H3 /   \ H

instead of R2-N-H:

CH3 \
       N - CA - (rest of the aminoa acid)...
   H /

My guess is that Gromacs automatically adds those 3 hydrogens because 
MME is a terminal residue, but I don't know how to prevent pdb2gmx from 
doing so.

I have tried different modifications in the .rtp file, including using 
different atom types, removing the peptide bond (the line N -C ) in the 
list of bonds, changing the hdb file, but nothing seems to have had any 
effect.

I guess I could also simply remove these hydrogens from the .gro and 
.itp files after using pdb2gmx, but I would have a better confidence in 
what I changed in the force field files if pdb2gmx was giving me the 
right answer directly.

See below my .rtp and .hdb entries:

[ MME ]
  [ atoms ]
      N    opls_238   -0.500     1
     CA    opls_224B   0.140     1
     HA    opls_140    0.060     1
     CB    opls_136   -0.120     2
    HB1    opls_140    0.060     2
    HB2    opls_140    0.060     2
     CG    opls_210    0.048     3
    HG1    opls_140    0.060     3
    HG2    opls_140    0.060     3
     SD    opls_202   -0.335     4
     CE    opls_209   -0.013     5
    HE1    opls_140    0.060     5
    HE2    opls_140    0.060     5
    HE3    opls_140    0.060     5
      C    opls_235    0.500     6
      O    opls_236   -0.500     6
     CM    opls_244   -0.110     7
    HM1    opls_140    0.037     7
    HM2    opls_140    0.037     7
    HM3    opls_140    0.037     7
  [ bonds ]
      N    CA
      N    CM
     CM   HM1
     CM   HM2
     CM   HM3
     CA    HA
     CA    CB
     CA     C
     CB   HB1
     CB   HB2
     CB    CG
     CG   HG1
     CG   HG2
     CG    SD
     SD    CE
     CE   HE1
     CE   HE2
     CE   HE3
      C     O

MME     6
1       1       H       N       CM      CA
1       5       HA      CA      N       C       CB
2       6       HB      CB      CG      CA
2       6       HG      CG      SD      CB
3       4       HE      CE      SD      CG
3       4       HM      CM      N       CA

Once again, any idea about this problem would be greatly appreciated.

Thanks to all in advance.

Regards,

Antoine


On 08/26/2011 05:42 PM, Delmotte, Antoine wrote:
> Oh, thank you so much! That was indeed the error.
>
> It's amazing how these little things can sometimes drive you mad....
>
> Thanks a lot,
>
> Antoine
>
> On 08/26/2011 05:27 PM, Thomas Piggot wrote:
>> Hi,
>>
>> I think the problem is that you have a dash rather than a minus 
>> symbol for the sign of the charge on the OD atom.
>>
>> Cheers
>>
>> Tom
>>
>> Delmotte, Antoine wrote:
>>> Dear Gromacs users,
>>>
>>> I am currently trying to run an MD simulation with the OPLS-AA force 
>>> field on a protein having different non standard residues and a 
>>> ligand. I found the charges for the OPLS force field for these 
>>> residues in the literature and I am now trying to add them in the 
>>> OPLS force field parameter files.
>>>
>>> I have edited the aminoacids.rtp and the aminoacids.hdb files for 
>>> the OPLS-AA force field, as well as the residuetypes.dat file. Here 
>>> is an example for one of the amino acids, hydroxyproline:
>>>
>>> [ HYP ]
>>> [ atoms ]
>>> N opls_239 -0.140 1
>>> CA opls_246 0.010 1
>>> HA opls_140 0.060 1
>>> CB opls_136 -0.120 2
>>> HB1 opls_140 0.060 2
>>> HB2 opls_140 0.060 2
>>> CG opls_137 -0.120 3
>>> HG1 opls_140 0.060 3
>>> OD opls_167 −0.683 3
>>> HD opls_168 0.743 3
>>> CD opls_245 -0.050 4
>>> HD1 opls_140 0.060 4
>>> HD2 opls_140 0.060 4
>>> C opls_235 0.500 5
>>> O opls_236 -0.500 5
>>> [ bonds ]
>>> N CA
>>> CA HA
>>> CA CB
>>> CA C
>>> CB HB1
>>> CB HB2
>>> CB CG
>>> CG HG1
>>> CG OD
>>> OD HD
>>> CG CD
>>> CD HD1
>>> CD HD2
>>> CD N
>>> C O
>>> -C N
>>> [ impropers ]
>>> -C CA N CD improper_Z_N_X_Y
>>> CA +N C O improper_O_C_X_Y
>>>
>>>
>>> When I run pdb2gmx, I get the following error, which is not very 
>>> informative:
>>>
>>> All occupancies are one
>>> Opening force field file /usr/share/gromacs/top/oplsaa.ff/atomtypes.atp
>>> Atomtype 1
>>> Reading residue database... (oplsaa)
>>> Opening force field file 
>>> /usr/share/gromacs/top/oplsaa.ff/aminoacids.rtp
>>> Residue 58
>>> -------------------------------------------------------
>>> Program g_pdb2gmx, VERSION 4.5.3
>>> Source code file: 
>>> /builddir/build/BUILD/gromacs-4.5.3/src/kernel/resall.c, line: 389
>>>
>>> Fatal error:
>>> in .rtp file in residue HYP at line:
>>> OD opls_167 −0.683 3
>>>
>>>
>>> I would be grateful if anyone could shed some light on the origin of 
>>> this error, and on what I can do to correct it.
>>>
>>> I am using Gromacs 4.5.3.
>>>
>>> Thanks a lot in advance,
>>>
>>> Antoine
>>>
>>
>
>