[gmx-users] angular COM removal

[Tomas Kubar]

Hallo

Maybe try
comm-grps = protein all-but-protein
(all-but-protein being all atoms except those contained in protein)

I guess this problem is brought about by the way VCM groups are constructed:
src/kernel/readir.c, do_numbering(), line 1081 in 4.0.3
  if (grps->nr == 1) {
    groups->ngrpnr[gtype] = 0;
    groups->grpnr[gtype]  = NULL;
  } else {
    groups->ngrpnr[gtype] = natoms;
    snew(groups->grpnr[gtype],natoms);
    for(j=0; (j<natoms); j++) {
      groups->grpnr[gtype][j] = cbuf[j];
    }
  }

Indeed, grps->nr==1. If you add another VCM group, then grps->nr==2 and
the else branch is executed, which should do the trick, I believe. At
the same time, the removal of rotation of a group including almost all
atoms (solvent) should have a negligible effect on the velocities. I'm
testing that right now, because I've had the same problem, trying to
keep a DNA oligomer in the center of the box. (I'm aware of the possible
appearance of artifacts...) If this doesn't work, then I'm going to try
to hack do_numbering() a bit.

tomas


XAvier Periole wrote:
>
> Dears,
>
> I am asking for the angular removal of the COM motion of a protein
> assembly
> in gromacs-4.0.4.
>
> - grompp gives a strange warning:
>
> checking input for internal consistency...
>
> WARNING 3 [file md.mdp, line unknown]:
>   Removing the rotation around the center of mass in a periodic system
>   (this is not a problem when you have only one molecule).
>
> - the lines are:
>
> nstcomm                  = 1
> comm-mode                = angular
> comm-grps                = protein
>
>
> Then the nstcomm and comm-mode appear to be correctly defined in the
> tpr file
> generated. However during the run the angular removal of the COM is
> clearly not
> done. The protein assembly is rotating ...
>
> Any reason for this?
>
> XAvier.
>
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://www.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 thewww
> interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>