AMBER Archive (2007)

Subject: AMBER: implementation of GB model in AMBER

From: john chen (j11chen11_at_gmail.com)
Date: Sun Jun 17 2007 - 07:49:24 CDT


Dear amber community,

I have a question concerning the implementation of the GB model in AMBER.

I've read the papers by Hawkins, Cramer and Truhlar and those more recently
by Onufriev, Bashford and
Case. In these papers, the elecrostatic contribution to the free energy of
solvation (within the GB model) is
always described as a sum over pairs of atoms, represented as spheres, each
with an effective Born radius
which describes to what extent the atom is buried in the molecule, and an
effective point charge (located at
the centre of the sphere). However, in the AMBER force-field, the charge
density is represented using the
RESP method, which involves creating a grid of point charges which more
accurately represent the charge
density disrtibution of the molecule (or molecular fragment). Does this mean
that when implementing the GB
model, the RESP charges are disgarded and Mulliken-like point charges
centred at each atomic position are
used instead? In other words, is it correct to say that the RESP charges are
only used for calculating the
coulombic interaction term in the force-filed and play no role in the GB
model for calculating the solvation free energy?

Is this the same for the Possion-Boltzmann method? When the electrostatic
potential is calculated using the PB equation
for each charged atom (or each charge site), are the 'charge sites' the grid
of point charges formulated in the RESP method,
or has this been replaced by a more simple set of atomic 'single point
charges'?

with best regards,
jc.

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu