AMBER Archive (2009)

Subject: Re: [AMBER] Can I use only a single point calculation to derive RESP charges?

From: FyD (
Date: Sun May 10 2009 - 01:52:12 CDT

Dear Josmar,

>> I do not know which version you use.
> Thanks for the help. The version I'm using is III.2.
>> However, in the case of such cofactors - with
>> negative charge(s) - it is unlikely the structure after
>> optimization will be similar to the that taken from the
>> X-ray structure.
> That's why I thing it's even more reasonable to use that
> conformation. Don't you?

Yes, why not. Try & See... However, the X-ray resolution of the
structure containing the co-factor might become a limitation. This is
why we always use QM geometry optimization as a _general_ method to
try to generate "canonical" building blocks.

>> So you need to find a way to solve this problem...
> Maybe the ideal would be to use some method like ONIOM..?(what you
> be far from my wishes...rsrsrs!)

In Oniom or implicite solvent conditions you consider your co-factor
as a whole molecule; I see two main limitations (i) how efficient will
be the Oniom or the implicite solvent method in reproducing the
environment and (ii) the size of some co-factors.

This is why we use the building block approach described by Cieplak et
al. Application of the multimolecule and multiconformational RESP
methodology to biopolymers: Charge derivation for DNA, RNA, and
proteins. J. Comput. Chem. 1995, 16, 1357-1377.
The main limitation here is to find the correct connecting groups. The
main advantage is that you can handle large number of co-factors in
the same approach (and built the corresponding force field topology

Moreover, using R.E.D. derived charge values are always reproducible.

>> I can send you the corresponding force field libraries in the
>> Tripos mol2 file format to your personal email address if you tell me the
>> cofactor(s) you need, their total charge as well as the force field
>> you plan to use.
> If I understood your notation the (P"-2") means the charge on
> phosphate group(s) equal to -2 (not the total charge). So, I'd be
> grateful if you send me the following files with Hornack et al.
> force fields.

Hornak et al. Sorry for him. My mistake ;-)

> NAD (P"-2")
> NADH (P"-2")
> NADP (P"-2")
> NADPH (P"-2")
> FMN (P"-2")
> FMNH (P"-2")

I send you the 6 force field libraries in the Tripos mol2 file format
to your personal address. If you compare these 6 files and the
different residues present in the files, the building block approach
will become obvious.

regards, Francois

AMBER mailing list