AMBER Archive (2009)

Subject: Re: [AMBER] questions about RESP

From: FyD (
Date: Thu Jul 23 2009 - 12:49:23 CDT

Dear Jeffrey,
>    Many thanks. Since the fitted atomic charges are dependent on the
> conformation used for MEP calculation, we should use
> the conformation near the one bound in the receptor, which is
> usually not the minimum. Then why is it necessary to optimize the
> isolated ligind structure to the minimum to calculate the MEP? 
> After your detailed explanation, now my main confusion is about how
> it will affect the fitted atomic charge when we use a conformation
> not corresponding to the minimum at 6-31G**.

When you extract a ligand from its protein environment, a
monosaccharide from its water solvent or when you look at a
multiple-charged structure the conformation you are going to get by QM
using HF/6-31G* (or HF/6-31G**) in gas phase is generally different to
that in its original environment. Consequently, you need to find a
strategy so that the conformations you are going to select (to be
involved in MEP computation) make sense.

See Cieplak et al. J. Comput. Chem. 1995, 16, 1357-1377 for the
choices made for standard amino-acids & nucleotids used to construct
larger biopolymers.

> Any paper on this topic on how the non-minimum conformation affect
> the atomic charges?

No idea. I am not sure this makes sense. We could imagine to involve a
transition state structure in MEP computation. The question is here
how to define the structure(s) to be used in charge derivation.

regards, Francois

AMBER mailing list