From: Floris Buelens (floris_buelens_at_yahoo.com)
Date: Mon Jun 06 2005 - 17:40:44 CDT
Many thanks for the detailed reply. After some
experimentation I believe I'm best off 'ignoring' the
issue and not including any ions, for reasons which
I'll quickly summarise.
I tried some simulations with an alchemically
disappearing sodium ion. In these simulations I ran
into a serious convergence problem. I initially placed
the ion randomly, and the beginning of the simulation
would give a nice stable dG, but at some point the ion
would drift close to the glutamate and be captured
next to it, giving a much larger dG. Obviously this is
not realistic sampling.
Interestingly, a regular sodium ion, not scaled with
lambda, does not get captured by the glutamate residue
in the same way - regular cations briefly associate
but don't get 'stuck'. If my understanding is correct,
this is because the alchemical Na+ exists within the
same "topology space" as the alchemical glutamate (in
NAMD FEP's "dual topology" paradigm) - whereas the
interactions of the Glu and the Na+ with the general
(non-alchemical) environment are scaled with the value
of lambda, interactions within the same alchemical
topology definition aren't scaled relative to each
other, so are relatively stronger - causing the
alchemical Na+ to sit next to the alchemical Glu.
Please do correct me if my understanding of this is
wrong.
My second justification for not adding ions is the
hope that my thermodynamic cycle should compensate for
any error introduced by the loss of a negative charge
- I'm completing the cycle by simulating the same
peptides in complex with a protein, which I've also
been doing without placing any ions. If there's some
kind of systematic error introduced, might it cancel
out as I'd be subtracting the same error later on
anyway? Any insight on this would be very welcome.
Kind regards,
Floris Buelens
--- Chris Chipot <Christophe.Chipot_at_edam.uhp-nancy.fr>
wrote:
> Dear Floris,
>
> unfortunately, there is no easy answer to your
> question. By and large,
> we haven't a sufficiently solid historical
> background to assert one
> way or another. In the scenario that you describe,
> there are pretty
> much two routes available: Either (a) you include a
> counter-ion that
> will appear as glutamine is mutated into glutamate,
> or (b) you ignore
> this change of the net charge in the box and rely on
> a possible
> compensation of the latter by a background charge.
> So far, I have
> failed to understand whether or not the NAMD
> implementation of PME
> will handle the variation of the net charge in the
> course of the
> simulation, and have, therefore, opted for the safer
> route (a). Yet,
> if you go for this choice, be aware that the
> counter-ion may come
> close to your perturbed residue and possibly bind to
> it, thereby
> modulating the resulting change in free energy as
> lambda goes from 0
> to 1. In fact, in a recent issue of J. Comput. Chem.
> (2005, 26,
> 115-122) Donnini et al. have advocated to avoid the
> use of
> counter-ions that can lead to serious sampling
> issues. One, unphysical
> way to circumvent this difficulty, would be to place
> the counter-ion
> far enough from the mutated site, and prevent it
> from approaching the
> latter by means of harmonic restraints or simply by
> keeping it fixed
> in space.
>
> Chris Chipot
>
>
> Floris Buelens wrote:
>
> >Hello,
> >
> >I'm running FEP where I'm trying initially to
> >calculate the free energy difference between two
> bound
> >peptides differing in one position, one glutamate,
> one
> >glutamine.
> >For my thermodynamic cycle, I'm simulating the
> 11-mer
> >peptide in solution, but I'm unsure how to handle
> the
> >change in charge between the two forms. When going
> >from the glutamine-containing peptide (no net
> charge)
> >to the glutamate one (net -1), should I introduce a
> >compensatory cation?
> >I've found some discussion of this in the
> literature
> >but I'd appreciate comments if anyone has
> experience,
> >Thanks for your time,
> >
> >Floris Buelens
> >
> >Crystallography, Birkbeck College, London
> >
> >
>
_______________________________________________________________________
>
> Chris Chipot, Ph.D.
> Equipe de dynamique des assemblages membranaires
> Unité mixte de recherche CNRS/UHP No 7565
> Université Henri Poincaré - Nancy 1 Phone:
> (33) 3-83-68-40-97
> B.P. 239 Fax:
> (33) 3-83-68-43-87
> 54506 Vandoeuvre-lès-Nancy Cedex
> E-mail:
> Christophe.Chipot_at_edam.uhp-nancy.fr
>
> http://www.edam.uhp-nancy.fr
>
> To sin by silence when we should protest makes
> cowards out of men
>
> Ella Wheeler Wilcox
>
_______________________________________________________________________
>
>
>
___________________________________________________________
Yahoo! Messenger - NEW crystal clear PC to PC calling worldwide with voicemail http://uk.messenger.yahoo.com
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:39:30 CST