Re: Thiol on gold simulation in NAMD

From: Axel Kohlmeyer (
Date: Thu Feb 26 2009 - 10:00:05 CST

On Thu, 26 Feb 2009, Jorgen Simonsen wrote:

JS> Hi,
JS> Thanks for the answer although I hoped that it would be easy and not problem
JS> to parametrize the Cu2+/Cu+ for a classical force field.
JS> In my case I think the ion are pretty well integrated into the protein until
JS> now I have used the parameters for Cu2+ from the article Deng et al
JS> (doi:10.1529/biophysj.105.064444) where they have included it as 2+ and a
JS> vdw parameter. The problem is that the coordination of one methionine

hmmm... this is exactly the kind of study that makes me uncomfortable.
perhaps i'm blind, but i don't see any justification for using charge
plus covalent radius for cu2+ and on top of that, they use this cu2+
on a structural motif without experimental data confirming it.

this is the kind of system, where i would do a QM/MM study on the
Cu2+ binding first. the Cu2+ has a preference for a highly non-spherical
coordination (water is coordinated 4+2, with the two waters being
so weakly bound that Cu2+ is generally consider to be a "quadratic"

JS> residues is not kept during the simulation which I would expect it to be -
JS> in two of the other binding it is though so further investigation is needed.
JS> My main question is would you include bonding parameters for the metal ion
JS> and its interactions with its ligand like for example in the paper by Fuchs
JS> et al.(doi:10.1002/jcc.20392) or Stote and Karplus where they as Deng use

well, if i could predict the result of calculations, i wouldn't need
to do them. ;)

so in every system you have to validate your choices (hence my criticism
about studies that take parameters without justifying them). the fuchs
study clearly supports my assessment that a simple charge model is not
a good model for any of these elements with significant "back
distribution" of charges. they also limit the applicability to the
specific coordination scheme. i remember a former colleague using
a force matching scheme to reparameterize the metal ion and its
environment for similar purposes.

JS> the vdw along with its charge.
JS> I know that many problems araises due to the d9 of Cu20 so I am really
JS> puzzled how to proceed? Any suggestions or comments would be appreciated

i would start by searching the literature whether somebody actually
has already done some qm/mm studies or plain qm studies of small
analogues of the Cu containing motif of your system and then take
it from there. doing qm/mm is not that easy either, as different
qm methods have different strengths and weaknesses, and to get
transition-metals _and_ hydrogen bonds represented well, usually
requires a quite costly QM method (or a specific reparametrization
of a cheaper method). not sure if you are willing to dig into this?


JS> Best

Axel Kohlmeyer
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
If you make something idiot-proof, the universe creates a better idiot.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:52:24 CST