AMBER Archive (2004)

Subject: Re: AMBER: Solvation free energy of a charged molecule.

From: David A. Case (
Date: Fri Nov 12 2004 - 10:11:15 CST

On Thu, Nov 11, 2004, Chunhu Tan wrote:
> I want to know, if I can use the Sander TI to calculate the
> solvation free energy of a CHARGED molecule. For example, I performed
> Sander TI for an ARG analog ( +1 charged) in a water box and in gas phase,
> the integration of DV/DL of these two process are 130.4(gas) and
> 180.5(water), so, if the difference (-50.1) is the solvation free energy of
> this molecue?

It's not clear what "process" you were simulating, that is, what the lambda=0
and lambda=1 end states were for your two calculations. To get the total
solvation free energy, you would need to convert your model compound to
nothing in both the gas phase and in solvent, and then take the difference.
This would probably be done in two steps, one in which you remove all of the
charges, and a second in which you convert all atoms to dummy atoms.

  As I know, when the system has nonzero net charge, uniform
> neutralizing plasma will be added implicitly in order to perform the PME,
> my question is, if this have any effect on the solvation free energy? If
> yes, how I can find it out?

For reasonable size boxes, the effects are very small; see, e.g.

%A T. Darden
%A D. Pearlman
%A L.G. Pedersen
%T Ionic charging free energies: Spherical versus periodic boundary
%J J. Chem. Phys.
%V 109
%P 10921-10935
%D 1998

> By the way, in his J. Chem. Phys. 2003 paper, Dr. Darden said that
> "ischarged" option in Sander was used, but I can not find this word in the
> manual (7 and 8), is this defaulted?



The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to