From: Ching Wong (wongcw_at_TITAN.SFASU.EDU)
Date: Mon Jul 10 2006 - 14:49:57 CDT
I am studying a glycosyl hydrolase enzyme and it works via a retaining
mechansim during which a glycosyl-enzyme intermediate is
formed.(nucleophile 282 attacks the anomeric carbon), and the
nucleophile never seem to be in the state of COOH. (it is either COO
before and after the reaction of as an intermediate). May I ask if MD
simulation should start with a (x-ray)structure protonated according to
pKa? Any suggested ways to do prediction? I tried with propKa and UMBD
and it seems to be the case. But I am doubtful about the conformation
of the structure, would that affect the pKa prediction?
jas
-------------------------
These are hard questions, and I don't think they're easy to answer in
general. I don't have any direct expertise, but others my research group
are involved in computational studies of enzymne reaction mechanisms,
so I
have picked up a few things.
X-ray structures are a good starting point, but you need to consider
carefully whether that structure is close to the one in the active
enzyme
(simulate without substrate to see what happens), whether partial
(de)protonation (though hydrogen-bonding) of pH-sensitive sites is
relevant. All of this is extremely hard to solve definitively in the
absence of strong experimental evidence. Implicit solvation calculations
for determining pKas have defects, and explicit solvation calculations
aren't large enough for non-artificial proton association/dissociation
to
occur (never mind whether the model physics can handle it). All I can
suggest is to read thoroughly the experimental literature for the
system,
and read about the techniques employed in other computational approaches
to similar systems. There is no prescribed method, because nothing is
perfect and the field is all too new.
Because they're of interest to other people on lists like namd-l, it's
usually a good idea to keep this sort of discussion on the list if they
started there.
Mark
-------
Jasmine
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:43:49 CST