> To add on to Soumendranath's comment, the only time you should _ever_
> add counterions is when you are running a system in explicit solvent.
> If this is your complex topology file for MM/PBSA calculations, salt
> effects are taken into account with a Debye-Huckel parameter for GB or
> explicitly in the Poisson-Boltzmann equation. Explicit counterions
> should not be used here.
I disagree with this, at least in the case of MM-PBSA/GBSA, since
sometimes explicit ions and/or water may be desirable, especially if they
form specific interactions with the solute. The implicit ion models do
not accurately capture these types of ion/water interactions. However I
do agree that including ions in implicit solvent MD simulations is not
recommended.
Two examples where explicit inclusion of ions and/or water improve the
results are:
(1) Energetic analysis of G-quadruplex structures which is effectively
misleading/impossible if the ions in the channel are not included.
R Stefl, TE Cheatham, III, N Spackova, E Fadrna, I Berger, J Koca,
and J Sponer. “Formation pathways of a guanine-quadruplex DNA revealed
by molecular dynamics and thermodynamical analysis of the substates.”
Biophys. J. 85, 1787-1804 (2003).
X Cang, J Sponer, and TE Cheatham, III. “Explaining the varied
glycosidic conformational, G-tract length and sequence preferences for
anti-parallel G-quadruplexes.” Nuc. Acids Res. 39, 4499-4512 (2011).
X Cang, J Sponer, and TE Cheatham, III. “Insight into G-DNA
structural polymorphism and folding from sequence and loop connectivity
through free energy analysis.” J. Amer. Chem. Soc. 133, 14270-14279
(2011).
(2) Study of drugs binding in the minor groove of DNA where inclusion of
bound water "helped".
N Spackova, TE Cheatham, III, F Ryjacek, F Lankas, L van Meervelt,
P Hobza, and J Sponer. “Molecular dynamics simulations and
thermodynamics analysis of DNA-drug complexes. Minor groove binding
between 4’-6-diamino-2-phenylindole and DNA duplexes in solution.” J.
Amer. Chem. Soc. 125, 1759-1769 (2003).
More recently, other groups have also played with post-processing MD
trajectory energy analysis including explicit water or ions.
Zhu YL, Beroza P, Artis DR. "Including explicit water molecules as part
of the protein structure in MM/PBSA calculations." J Chem Inf Model.
2014 Feb 24;54(2):462-9.
Wong S, Amaro RE, McCammon JA. "MM-PBSA Captures Key Role of
Intercalating Water Molecules at a Protein-Protein Interface."
J Chem Theory Comput. 2009 Feb 10;5(2):422-429.
However I will note that this approach is a complex/advanced use case that
requires strong knowledge of the underlying methods/code and requires
significant care. For example, you have to make sure your GB and/or PB
parameters for the ions give accurate free energies of solvation. Also,
you cannot compare systems with different net-charges.
--tec3.utah.edu
p.s. an alternative to MM-PBSA/GBSA is to look at raw potential energies
(assuming equivalent number of ions/solvent). Soon coming ASAP will be
our paper describing this in the context of highly charged inhibitors
interacting with RNA.
NM Henriksen, Hamed Hayatshahi, DR Davis, and TE Cheatham, III.
“Structural and energetic analysis of 2-aminobenzimidazole inhibitors in
complex with the hepatitis C virus IRES RNA using molecular dynamics
simulations.” J. Chem. Info. Model. [in press] (2014).
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu May 22 2014 - 15:30:02 PDT
