AMBER Archive (2005)

Subject: Re: AMBER: Cu++ and RED

From: FyD (fyd_at_u-picardie.fr)
Date: Fri Feb 11 2005 - 11:23:53 CST

> I have been trying to compute RESP charges for an
> aminoacid derivative in complex with Cu(+2). I use
> Gamess and RED to compute partial atomic charges.

Which version of R.E.D. do you use: version I or version II ?
RESP charges for metal complexes are automatically derived _only_ with R.E.D.-II

> However, I noticed that Gamess somewhat assigns (+1)
> charge to the AA derivative - Cu complex but not (+2)
> charge. The A.A. derivative itself has zero charge.

Here, I am not sure I understand you. What is the total charge for your complex:
Let's say it is +2. We can imagine some charge delocalisation making the partial
charge of your copper quite strange. I got data from Sebastian F.A. about an
iron-complex and we observed together important charge delocalisation for his
complex. We even run some charge calculations with Piotr and we clearly
observed that RESP charges are quite weird in some cases...

> Is there anyway that I could set/modify RED parameters
> so that (+2) charge is assigned to the AA
> derivative-Copper complex ?

I would first compare ESP and Mulliken charges computed by the QM soft, with the
RESP charges obtained with R.E.D. Then, if you see some strange effects, you
could add intramolecular restraints (IMR) to 'affect' the charge value of your
copper. IMR are not automatically handled in R.E.D.-II. This means you have to
proceed in two steps (i) run R.E.D.-II without IMR (ii) get the espot,
input1/input2 files - add the IMR in input1, freeze the corresponding charges
in the input2 and re-run resp manually using the 2 aliases (resp1; resp2)
defined @
http://www.u-picardie.fr/labo/lbpd/RED/FAQ-I.htm#27

Adding IMR is directly related to the automatic generation of LEaP libraries for
unusual aminoacids/sugar/nucleotide. This will be implemented in R.E.D.-III
(using the 'tripos.mol2' format instead of the 'prep.in' one). Mulliken, ESP
and RESP charges will be also compared (with statistics) to detect possible
problems... We will start this very soon, now. This is quite straightforward to
implement in R.E.D.-II. The 1st beta version of R.E.D.-III (without manual)
should be available before next June...

I hope this helps, regards,

  F.-Y. Dupradeau
DMAG EA 3901 & Faculte de Pharmacie, Amiens, France
The Scripps Research Institute, La Jolla, CA, USA
       --
http://www.u-picardie.fr/labo/lbpd/FyD.htm
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu