[gmx-developers] mdrun -membed not functional with pieces > 1 ?

[Jochen Hub]

Am 30/07/15 um 14:07 schrieb Jochen Hub:
> Hi,
> 
> I am trying to put two molecules into a membrane with mdrun -membed.
> This should be possible with the option pieces = 2 in membed.dat.
> However, when mdrun checks if the energy group exclusions are consistent
> with the group chosen, one gets something like:
> 
> Energy exclusions "mol1" and  "mol1" do not match the group to embed "mol0"
> 
> The problem seems the check pasted below. It works only if there is an
> exclusion only between the atoms of *one* group (such as "Protein
> Protein"), but not with two groups (such as "mol0 mol0 mol1 mol1".
> However, if one makes one large group containing all pieces (here mol0
> and mol1), then a consistency check with the freezegroups fails.
> 
> Do I miss something here? Did anyone get mdrun -membed working with
> pieces > 1 ?

Sorry, I got it. One first has to select the complete group and later
the pieces - then the error does not appear anymore.

Jochen


> 
> Many thanks,
> Jochen
> 
> for (i = 0; i < ng; i++)
> {
>    for (j = 0; j < ng; j++)
>    {
>      if (inputrec->opts.egp_flags[ng*i+j] == EGP_EXCL)
>      {
>        bExcl = TRUE;
>        if ( (groups->grps[egcENER].nm_ind[i] != ins_grp_id) ||
>             (groups->grps[egcENER].nm_ind[j] != ins_grp_id) )
>        {
>           gmx_fatal(FARGS, "Energy exclusions \"%s\" and  \"%s\" do not
> match the group "
>                                   "to embed \"%s\"",
> *groups->grpname[groups->grps[egcENER].nm_ind[i]],
> *groups->grpname[groups->grps[egcENER].nm_ind[j]], ins);
>        }
>     }
>  }
> }
> 
> 

-- 
---------------------------------------------------
Dr. Jochen Hub
Computational Molecular Biophysics Group
Institute for Microbiology and Genetics
Georg-August-University of Göttingen
Justus-von-Liebig-Weg 11, 37077 Göttingen, Germany.
Phone: +49-551-39-14189
http://cmb.bio.uni-goettingen.de/
---------------------------------------------------