[gmx-users] fixed dihedrals

[Christoph Freudenberger]

Anton Feenstra wrote:
> Christoph Freudenberger wrote:
> [...]
> 
>> But is the residual freedom left by the constraints really larger
>> than that given by the improper?
> 
> 
> That would depend on a lot of details, I'd guess. If Shake convergence
> becomes the bottleneck (no experience with that myself, but I know it
> *can* sometimes be a problem), you might be better off with a tight
> improper and maybe even a shorter timestep. 
Maybe... setting up topologies (esp. for solvents) really isn't an easy
straitforward task... had to learn that during the last weeks while
optimizing my solvents;-)
What do you accually mean with "tight improper"...? It cannot be one
with a large force constant from your last mail.

> On the other hand, if you
> want it constrained (as in 'immobile'), you need a different construction
> to completely remove the degree of freedom from your molecule, for
> example a construction with three 'masses' and a set of dummy atoms,
> as was discussed on the list for some solvent molecule last week.
> 
> Otherwise, might we want to implement a planarity constraint?
> Is that at all feasible? Or sensible?
Yes, linearity can quite easily be accieved by a construction
similar to the one i proposed for acetonitrile. Planarity still
remains a real challenge, mainly because it is quite difficult to
redeistribute the masses in a way that doesn't disturb the
tensor of inertia too much. Ok, as an residue within a protein
this is sure not such a big problem, but for planar slovents
it is definitly important. Forunatly DMSO isn't planar ;-)

regards
-- 
Christoph Freudenberger
Abt. Organische Chemie I AK Siehl - Uni Ulm -Tel: ++49-731-502-2785