Re: AMBER: mm-pbsa:- atom charges and radii ?

From: Thomas E. Cheatham, III <>
Date: Tue, 11 May 2004 08:24:17 -0600 (Mountain Daylight Time)

> I have been trying to get my head around the atomic charges and
> radii used when performing a mm-pbsa calculation (I am using Amber7 at the
> present time). From published articles, most people seem to have used
> Amber charges, whether from the force field parameter file or fitted using
> RESP and have combined this with PARSE radii. However, on reading the
> article on the PARSE parameters (D.Sitkoff et al., J. Phys. Chem. 1994,
> 98, 1978-88) the following sentence seems to suggest this may not be a
> good idea :- "The PARSE charges lie within the range of charges used in
> the force field sets CVFF, CHARMM, and OPLS but differ significantly from
> the AMBER and ab initio sets." Given that this paper was published in
> 1994, will this still pose problems with the latest version of the AMBER
> force field ? Has there been any work published on how changes to charges
> and radii affect the mm_pbsa calculation ?

If you are using parm94.dat or parm99.dat, the same charges are present
and the issue could be apropos. There is a key dependence of the PB
results on the radii set employed, however the question you have to
address for yourself is, how important are these differences in your
particular case? For relative free energies of binding, there may be
sufficient cancellation; for absolute, clearly these will not be optimal.

There, to my knowledge, has not been an update of the continuum radii for
use with the AMBER force field; there is nice work by Banavali and Roux
how did extensive MD and FEP calculations to develop a more consistent set
of continuum radii for use with the all-atom CHARMM27 force field.

> radii and un-commenting the PARSE radii and re-compiling to get it to
> produce PARSE radii ?

Without looking at the code I would say probably yes; you can easily check
this by looking at the resultant PDB files.

Good luck.

The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to
Received on Tue May 11 2004 - 15:53:00 PDT
Custom Search