Hi,
Here are a few things I have noticed while using Antechamber to generate
AM1-BCC charges. The molecule will be minimised by DIVCON using a BFGS
minimisation algorithm and the charges will be derived from the optimised
structure. However, in the final output file (prepi,mpdb ...) the coordinates
will be those of the initial input file (non DIVCON optimised). Most of the
time that should not be a big issue since you are presumably going to run MD
on that compound and therefore the exact geometry used to generate the partial
atomic charges should not be an issue (if it is you have a problem !).
There is one exception though, I have read (but not checked personnaly) that
AM1-BCC will be inadequate if the molecule has intra molecular hydrogen
bonds (this would be due to errors in the AM1 wavefunction, if anybody knows
well semi empirical QM on this mailing list, could they confirm that AM1 has
trouble handling hydrogen bonds ?) . In this case, getting decent AM1-BCC
charges might be a pain because even if you modify the geometry of your input
file , you may still form the hydrogen bond during the minimisation by divcon.
If i have to face this issue in the near future, I think I would try to
modify the geometry of my ligand to break any intramolecular hbond
(whenever possible), use a crude minimisation to fix bad contact problems but
hopefully not cause large structural changes that would form an Hbond again.
Regards,
Julien Michel
University of Southampton
Department of Chemistry
Dr J. W. Essex group
> Dear ANTECHAMBER users:
> I have been using AM1-BCC charge models by invoking ANTECHAMBER...I have a
> question- Does DIVCON also geometry-optimize the supplied molecule before
> generating AM1-BCC charges, or it just performs an SCF calculation on the
> geometry.
>
> Thanks,
> Vineet
>
> "The future belongs to those who believe in the beauty of their dreams"
> -Eleanor Roosevelt
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Wed Jan 26 2005 - 11:53:00 PST
