AMBER Archive (2003)

Subject: Re: question about partial charge

From: sychen (yuann_at_bioinfo.ndhu.edu.tw)
Date: Tue Feb 18 2003 - 21:52:32 CST

Dear Lepsa and everyone,

Thanks for your suggestion. Because the two subunits are 'identical',
I guessed this strategy will be okay.
About the charge of this metal ion,Mg++ , it is between the two subunits,
and I've no idea if its charge to be assigned as formal charge
or fitted by RESP after Gaussian98' calculation? Especially each subunit
has one ionized O enolate anion that are coordinated to the divalen
cation, Mg++, in the complex.

Hope that the AMBER authors or someone could give me some ideas
I'm appreciated in advance your responses!

Best Regards,

sychen.

On Tue, 18 Feb 2003 08:56:59 +0100
"Lepsa" <lepsik_at_marilyn.uochb.cas.cz> wrote:

> Dear Sychen,
> for metal ions I don't know. As for organic dimer, if the two subunits are
> "reasonably similar"
> you could use just one. (Provided optimizing monomer you don't get too far
> from its structure in the dimer).
> If the two subunits are different you should derive charges on both of them
> and then do multiple conformation fitting.
> (for details see, Cieplak et al, J COMPUT CHEM 16: (11) 1357-1377 NOV 1995)
>
> Alternatively, you could make methyl- or hydrogen-capped fragments of your
> molecule, optimize them, calculate ESP, fit it
> and go thru antechamber and prepgen to fit the charges on open-valence
> fragments.
>
> Hope this helps.
>
> Best regards,
>
> Martin
>
> ----------------------------------------------------------------------------
> ----------------------------------------------------------------------------
> ----
> Martin Lepsik, PhD student Phone: +420/2/20183-540, Fax:
> +420/2/20183-292
> Dept. of Theor Chem & Center for Complex molecular Systems and Biomolecules
> Institute of Organic Chemistry and Biochemistry (IOCB)
> Flemingovo nam 2,
> Czech Academy of Sciences,
> 166 10, Prague 6,
> Czech Rep.
> URL: www.uochb.cas.cz/~teochem
>
> > I also want to do a optimization calculation with HF/6-31g* function on
> > an organic molecule with two identical structure to obtain its partial
> > charge. The initial structure is extracted from a complex structure
> > stored in PDB.
> > Unluckily, this organic molecule is too large to do optimization
> > job in Gaussian98 even though I sent this job to my server, IBMP690,
> > with 16CPU. So, is it reasonable to simply do optimization job on only
> > one structure of this dimer or just do general job without optimizaion
> > on a whole dimer in order to get its ESP for RESP fitting?
> >
> > Thanks for your suggestion!
> >
> > Best Regards,
> >
> > sychen.