AMBER Archive (2004)

Subject: Re: AMBER: PBSA parameter question (Radii and Charges)

From: Ray Luo (
Date: Wed Oct 13 2004 - 11:38:38 CDT

> I have a question concerning the parameters used for Poisson-Boltzmann
>calculations with the pbsa program. Are those parameters suitable to
>reproduce eg protein solvation free energy when used with another "standalone"
>PB solver like Delphi or APBS, or are they somehow tied to the pbsa
>program ?
Given identical input and parameters and setting, pbsa should agree with
Delphi within 1%, but not to the last digit because pbsa uses a slightly
different algorithm for dielectric boundary. Nevertheless, delphi and
pbsa should correlate very well for a large number of conformations, for
example as in a mmpbsa calculation. If you find any discrepancy between
the two, please let me know.

However, apbs is very different, it uses a spline function to set up the
solvent excluded volume (both pbsa and delphi uses numerical procedures
to compute solvent excluded volume). When different solvent excluded
volumes are used in dielectric constant assignment, very different
solvation energies can be obtained. Of course, you won't see much
differences in small molecules, but only in large molecules.

>In the manual it is also mentioned that the atomic cavity radii for molecular
>surface calculation and dielectric assignment used by pbsa are optimized for
>model compounds of proteins. Which proteins were used, and which ?experimental?
>parameter was reproduced ?
These parameters are optimized with respect to experimental solvation
free energies for amino acid side chain analogs (i.e. without
mainchain). We're working to reoptimize them with respect to
electrostatic solvation free energies in explicit solvent/PME with free
energy perturbation.

>(Should these details be published already, then sorry for having missed the
> reference)
We will publish these radii when we finish the reoptimization.


Ray Luo, Ph.D.
Department of Molecular Biology and Biochemistry
University of California, Irvine, CA 92697-3900
Office: (949)824-9528         Lab: (949)824-9562
Fax: (949)824-8551          e-mail:
Home page:

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