[gmx-users] Charmm36 in Gromacs: missing dihedrals

[Thomas Piggot]

You are doing something strange. The dihedrals section you posted is not 
the one which should have been produced by pdb2gmx if you selected the 
correct force field. When I run pdb2gmx then I get all of these 
dihedrals. Also in your posted section then you have the wrong function 
  for the dihedrals (it should be 9 not 1), which again suggests you are 
doing something very wrong. Are you sure you selected the correct force 
field and posted the correct topology?

The GROMOS force fields are another matter, where yes these dihedrals 
are deliberately excluded. If you search on the internet (and probably 
this mailing list) you will find more information about this.

Oh and by the way if you intend to run protein and lipid simulations 
(with CHARMM36) then make sure you use the -nochargegrp option for pdb2gmx.

Cheers

Tom

Li Jianguo wrote:
> Dear All,
>  
> I am trying to use Charmm36 force field in Gromacs. I downloaded the 
> ffcharm36 force field files created by Tom Piggot and Pär Bjelkmar from 
> the user contribution in Gromacs website. Then I used the pdb2gmx to get 
> the popc_c36.top using the following command:
> pdb2gmx -f popc.pdb -o popc.gro -p popc_c36.top
>  
> When I checked the popc topology file, I found that there are some 
> dihedral terms missing in the topology file. For example, in the 
> popc_c36.top file, atoms 3, 4, 5 are all methyl carbons, there should be 
> three dihedral terms 3 1 2 17; 4 1 2 17 and 5 1 2 17.
> Unfortunately, there are only 3 1 2 17, the last two are missing, part 
> of the popc_c36.top are shown below:
> [ atoms ]
> ;   nr       type  resnr residue  atom   cgnr     charge       mass  
> typeB    chargeB      massB
> ; residue   1 POPC rtp POPC q  0.0
>      1        NTL      1   POPC      N      1       -0.6     14.007   ; 
> qtot -0.6
>      2       CTL2      1   POPC    C12      2       -0.1     12.011   ; 
> qtot -0.7
>      3       CTL5      1   POPC    C13      3      -0.35     12.011   ; 
> qtot -1.05
>      4       CTL5      1   POPC    C14      4      -0.35     12.011   ; 
> qtot -1.4
>      5       CTL5      1   POPC    C15      5      -0.35     12.011   ; 
> qtot -1.75
>      6         HL      1   POPC   H12A      6       0.25      1.008   ; 
> qtot -1.5
>      7         HL      1   POPC   H12B      7       0.25      1.008   ; 
> qtot -1.25
>      8         HL      1   POPC   H13A      8       0.25      1.008   ; 
> qtot -1
>      9         HL      1   POPC   H13B      9       0.25      1.008   ; 
> qtot -0.75
>     10         HL      1   POPC   H13C     10       0.25      1.008   ; 
> qtot -0.5
>     11         HL      1   POPC   H14A     11       0.25      1.008   ; 
> qtot -0.25
>     12         HL      1   POPC   H14B     12       0.25      1.008   ; 
> qtot 0
>     13         HL      1   POPC   H14C     13       0.25      1.008   ; 
> qtot 0.25
>     14         HL      1   POPC   H15A     14       0.25      1.008   ; 
> qtot 0.5
>     15         HL      1   POPC   H15B     15       0.25      1.008   ; 
> qtot 0.75
>     16         HL      1   POPC   H15C     16       0.25      1.008   ; 
> qtot 1
>     17       CTL2      1   POPC    C11     17      -0.08     12.011   ; 
> qtot 0.92
>     18       HAL2      1   POPC   H11A     18       0.09      1.008   ; 
> qtot 1.01
>     19       HAL2      1   POPC   H11B     19       0.09      1.008   ; 
> qtot 1.1
> .......
> [ dihedrals ]
> ;  ai    aj    ak    al funct            c0            c1            
> c2            c3            c4            c5
>     3     1     2    17     1
>     2     1     3     8     1
>     2     1     4    11     1
>     2     1     5    14     1
>     1     2    17    23     1
>     2    17    23    20     1
>    21    20    23    17     1
> ........
>  
> I also checked the dpc.itp in gromos force field from Tielman's website, 
> which has been widely used by many people. There are also some missing 
> dihedral terms in the dpc.itp file. 
> I am wondering if those missing dihedrals are neglected intentionally 
> for gromos force field? When I use other force fields such as 
> charmm/amber, do I need to add the missing dihedrals? If so, is there 
> any way to do it instead of manually adding those missing dihedral terms?
> 
> Could anybody give some suggestions? Thanks very much!
>  
> Cheers,
> Jianguo
> 
> 
> 

-- 
Dr Thomas Piggot
University of Southampton, UK.