[Dock-fans] Docking with metal ions
jir322 at gmail.com
Thu Jan 19 09:05:51 PST 2012
My suggestion is to make the metal part of the receptor, even if you
don't think of it that way. Two reasons: 1) the residues on the
protein that bind the metal are not limitless. Enumerate the limited
possibilities. 2) forcefields for ligands assume a limited range of
atoms, usually C, N, O, F, P, S, Cl, Br, I, H. For instance, B and
even Si are often poorly supported. Metals are in my experience not
well modeled by the force fields usually used in drug discovery
On Thu, Jan 19, 2012 at 2:12 AM, Jose R. Valverde
<jrvalverde at cnb.csic.es> wrote:
> Dear all,
> besides the problems in my previous message, I am also trying
> to dock a molecule that is normally associated with a metal ion. My
> problem is how to get Zou GB/SA scores or estimate binding affinity.
> The problem is that the ion should be placed wherever the ligand
> docks, i. e. always relative to the ligand pose, as it is not normally
> coupled to the enzyme, but the ligand. So docking should be done with
> the ion included in the ligand.
> I do get the primary conformers using a grid score, but when
> they are minimized and rescored with Zou GB/SA I keep getting weird
> results (likely due to the presence of Mg++ in the ligand). Although
> I do not get these errors with Hawkins GB/SA or PB/SA, I'm worried
> the results may be wrong as well.
> Something similar happens with Amber score: I cannot generate
> a topology for the ligand+ion (I tried SwissParam, ATB,..) and
> so the Amber method refuses to work.
> Can anybody suggest any good alternative approach?
> Scientific Computing Service
> Solving all your computer needs for Scientific
> Dock-fans mailing list
> Dock-fans at docking.org
More information about the Dock-fans