David Case wrote:
> The TIP3 water model was defined to be a rigid model. If you allow
> the bond lengths to change it can find regions of conformational
> space where a hydrogen on one water molecule gets arbitrarily close
> to an oxygen on another, leading to negative infinities in
> energies.
Thanks, I should have recognized that point.
> Even if this doesn't happen, the properties of such a "non-rigid
> TIP3" will not be the same as "real" TIPE. So it is neither correct
> nor necessarily practical to run TIP3 simulations in Amber without
> using SHAKE.
But how can we really be so sure that __all__ the causes of solvation
on electrostatic proteins can be fully accounted for by a rigid water
model? For making sure a protein hangs together, it's fair enough to
apply SHAKE and finish the job as fast as possible, but for studying
somewhat subtle mutation effects and dynamics (particularly on
electrostatic proteins), I'm not so confident that the influence of
the hydrogen vibrations can be simply ignored. In all due humility
and respect and recognizing my own limitations here, doesn't water
tend to "rattle" things that stick out?
...WD
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Wed Jan 14 2004 - 15:53:12 PST
