AMBER Archive (2005)

Subject: Re: AMBER: Mg2+ issues of using MM-PB/GBSA. .

From: Don.Bashford_at_stjude.org
Date: Mon Dec 05 2005 - 12:09:20 CST


>>>>> "mingche" == mingche Pan <panmingche_at_yahoo.com> writes:

    mingche> I wonder whether I can use MM-PB/GBSA method to calculate
    mingche> the binding free energy in the case of Mg2+ directly
    mingche> involving in the ligand binding....

Dear Mingche,

This idea has some real difficulties. Almost 10 years ago Lou
Noodleman's group, in collaboration with my group, published a study
of QM/PB calculations for Mn(2+), Mn(3+), Fe(2+) and Fe(3+) [Li et
al., Inorg. Chem. 35 4694-4702 (1996)]. We found that for getting the
correct solvation energies (and thus, redox potentials) it was
necessary to include the first and second solvation shells explicitly,
and use continuum representation of solvent only beyond those shells.
The reason is that there is significant charge transfer between water
molecules and the metal ion, or in other words, a partially covalent
character in the metal-water interaction. This is really beyond what
any of these continuum-based models are meant to represent. I would
think that the change from a metal-water ligand interaction to a
ligand interaction with some protein residue(s) would contain
substantial energy terms that purely classical electrostatic
calclulations could not reproduce.

So as Dave said, all this would "represent a real research problem",
which, I might add, needs methods beyond the MD/PB/GB scope.

Don Bashford
St. Jude
Memphis

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu