[gmx-users] gromacs versus amber boxsize -correction-

[Xavier Periole]

On Tue, 29 Jan 2008 13:14:21 +0100
  "servaas michielssens" <servaas.michielssens at student.kuleuven.be> wrote:
> I made a mistake in the numbers, but the question remains:
> 
> I ran the following command in gromacs:
> editconf -bt dodecahedron -f p61_vac_gro.gro -o p61_vac_dod.gro -d 0.8
> 
> The equivalent command in amber:
> solvateoct p61_vac.pdb TIP3PBOX 8.0

is there any reorientation in volved?
  
> As far as I understand this, both command created a box were the edges are 
>0.8 nm away from the solute. But the resulting image distance (d in the 
>manual) 

amber or gromacs manual?

> in gromacs is:
> 121.3092
> and in amber:
> 110.5103716

Are the d the same in both programs?
The definitions of dodecahedron are different in gromacs and gromos.
could be a similar problem in with amber!
Note however that the gromacs definition decreased the volume as
compared to gromos.
> 
> This is makes a hugde difference in the number of water atoms to add.
> Does anyone has an explanation for this paradox?
> 
> 
> kind regards,
> 
> servaas

-----------------------------------------------------
XAvier Periole - PhD

NMR & Molecular Dynamics Group
University of Groningen
The Netherlands
http://md.chem.rug.nl/~periole
-----------------------------------------------------