[gmx-users] CHarmm and popc membrane (response to "Missing Residues are list ind pdb file")

[chris.neale at utoronto.ca]

Hi Stefan and Pär,

I believe that the latest response had the wrong subject, so I am  
changing it back.

Pär, two comments:
First, does that option really exist?

$ /scratch/cneale/GPC/exe/intel/gromacs-4.0.7/exec/bin/pdb2gmx -h > z  
2>&1; cat z|grep nochargegrp|wc -l


Second, I am not entirely convinced that one can discount the need for  
gromacs-style charge groups. When CHARMM includes an atom in the  
cutoff, it includes the entire charge group, but that doesn't mean the  
algorithms of CHARMM and gromacs are entirely the same, especially  
with respect to:

"Since atoms only see each other when the centers of geometry of the charge
   groups they belong to are within the cut-off distance, too large charge
   groups can lead to serious cut-off artifacts."

I admit that it has been 3 years since I used CHARMM, but as I recall  
the requirement for pair interactions to count within the cutoff is  
only that one atom within the charge group is close enough, whereas  
the gromacs warning indicates that in gromacs the interaction is only  
counted when the center of mass of the charge groups are also within  
the cutoff.

Any charmmers out there that can clear this up? I did some searching  
on their archives but I was blocked after 3 searches unless I logged  
in (ugh!).

I gather that the large charge group listed below is the one that is  
spurring the message:

GROUP                  !                 H31
ATOM N    NTL    -0.60 !                 |
ATOM C11  CTL2   -0.10 !             H33-C13-H32
ATOM C12  CTL5   -0.35 !                 |
ATOM C13  CTL5   -0.35 !           H21   |   H43
ATOM C14  CTL5   -0.35 !            |    |   |
ATOM H11  HL      0.25 !       H22-C12---N---C14-H42   (+)
ATOM H12  HL      0.25 !            |    |   |
ATOM H21  HL      0.25 !            H23  |   H41
ATOM H22  HL      0.25 !                 |
ATOM H23  HL      0.25 !                 |
ATOM H31  HL      0.25 !                 |
ATOM H32  HL      0.25 !             H11-C11-H12
ATOM H33  HL      0.25 !                 |
ATOM H41  HL      0.25 !                 |
ATOM H42  HL      0.25 !                 |
ATOM H43  HL      0.25 !                 |



-- original message --

Hi,

Note, that this is indeed the size of the charge groups in the CHARMM  
force field. You could run pdb2gmx -nochargegrp if you know what  
you're doing.

Regards,
Pär Bjelkmar

> I think the notes from grompp are pretty clear. You need to redivide  
>  your charge groups so that they contain fewer atoms. You can do  
> this  in the .top or .itp file and you can find out how to do that  
> in the  manual. Feel free to post your original and  
> charge-group-modified  topologies back here if you want people to  
> take a look at them.
>
> Chris.
>
> -- original message --
>
> I am trying to equilibrate a POPC membrane with 128 POPC molecules, but
> grompp gives out the following notes and minimization with mdrun does not
> work. The output on screen of mdrun comes out like this:
> Steepest Descents converged to machine precision in 15 steps,
> but did not reach the requested Fmax < 10.
> Potential Energy  =  1.4422483e+17
> Maximum force     =            inf on atom 4082
> Norm of force     =            inf
>
> And grompp notes are:
> NOTE 2 [file topol.top, line unknown]:
>   The largest charge group contains 12 atoms.