[gmx-users] Freezing a molecule

[Mark Abraham]

Christian Seifert wrote:
> Hi!
> 
> You are using a "frozen group" but you are not using constraints?!?
> 
> As far as I know, a frozen group IS a constrained group. Did you mean 
> something different?

No, frozen groups, constraints and restraints are distinct concepts in 
GROMACS - they imply the use of different algorithms. One is asking for 
trouble when one is treating any atom with more than one of these.

> Am Dienstag, 27. Oktober 2009 um 17:23 schrieb Jennifer Williams:
> Hello,
> 
> One quick questions...
> 
> I have a structure for which I now want to freeze a portion.
> I already have a .top file where the entire structure is flexible  
> (bond angles, stretches and torions defined).
> When freezing, do I need to delete all those bond stretches, angles  
> and torsions associated with the frozen part from my .top file?
> (I am not using constraints on the frozen portion as I have seen this  
> can cause problems).
> Will the bonding parameters all be set to zero when the these atoms  
> are frozen or will their contribution be calculated if I leave these  
> terms in.

grompp probably lists the number of the various sorts of interactions - 
or at least gmxcheck will do so when comparing .tpr generated with and 
without frozen groups.

Mark