AMBER Archive (2002)

Subject: Re: gaff and parm99

From: Pascal Bonnet (pascal.bonnet_at_man.ac.uk)
Date: Wed Jul 24 2002 - 10:51:30 CDT

"David A. Case" wrote:
>
> On Wed, Jul 24, 2002, Pascal Bonnet wrote:
> >
> > In GAFF (general amber force field) the carbon radius is 1.7 A and in
> > parm99 is 1.9080 A (and 1.5 and ~1.7 for oxygen radius)
> > I don't think it is coherent to use two different force fields in the
> > same system receptor-ligand?
> >
> > Why in GAFF, the hydrogen radius of TIP3P model is 1.2 and not 0 as in
> > parm9X?
> > RADIUS HW 1.2 (from GAFF)
> > HW 0.0000 (from pam99)
> >
>
> It looks like you are confusing the radius used for generalized Born
> calculations with the Lennard Jones r-sub-e values, (although I can't
> figure out what files you are getting your information from...)
>

I was looking in
/amber7/dat/antechamber/RADIUS.dat

As I understand, this file is only for GB calculations.

The gaff force field for antechamber is in
/amber7/dat/leap/gaff.dat

So I was confusing the two different directories.
It could be useful to have a 0README in the /amber7/dat/antechamber/
directory.

> In both parm99.dat and gaff.dat, carbons typically have an r-sub-e value
> of 1.908, and the tip3p water hydrogen (which is called "HW" in parm99.dat
> and "hw" in gaff.dat) is zero.
>
> The values of 1.7 for Carbon and 1.2 for Hydrogen look like the "bondi radii"
> that are used (in both force fields) for electrostatic radii in the
> generalized Born solvation model. Unlike some other implementations of
> continuum models, there is no requirement or expectation in Amber that the
> "dielectric radii" used in GB should be the same as the r-sub-e values from
> the Lennard-Jones terms.
>
> So, the force fields are equivalent in this sense. Note also that all
> atom types in gaff.dat are lower case, so the statement:
>
> > RADIUS HW 1.2 (from GAFF)
>
> must be wrong, since gaff doesn't have any atom type "HW".

This is also from /amber7/dat/antechamber/RADIUS.dat file.

Thanks for your help

pascal