[gmx-users] Problem - can't set dodecahedrons box

[Mark Abraham]

On 09/02/10 12:12, Maurício Menegatti Rigo wrote:
> Hi,
>
> I'm facing a problem with GROMACS that appears pretty simple, actually.
> I'm trying to set a dodecahedron box but all I can get is a triclinic
> box. For that, I'm using the following line:
> /
> editconf -f pMHC -bt dodecahedron -o -d 1 /

Since all box types are special cases of a triclinic box, they are all 
described in the same way as the latter. I expect the values are 
suitable for dodecahedron, but resemble triclinic.

> Am I doing something wrong?? Another problem, different from this one,
> is that my protein, during the simulation, is escaping/comming
> off/"running away" (sorry I'm butchering the english) from the water
> box. I tried to use periodic boundaries conditions, but didn't work (I
> added a line in my .mdp file - pbc = xyz). Someone could help me with
> that, please?

See 
http://www.gromacs.org/Documentation/Terminology/Periodic_Boundary_Conditions.

The .mdp file controls the algorithms of the simulation. What you do 
afterwards to visualise the results of a simulation is the business of 
your visualisation software, and/or your usage of trjconv to keep 
yourself happy with what you can see.

Mark