AMBER: Explicit water in implicit enviornment?

From: Hayden Eastwood <>
Date: Sun, 9 Sep 2007 20:57:26 +0100

Dear AMBER users

I am calculating binding energies of 4 ligands with a target protein.
However, I wish to compare these binding energies with that of the
ligand-free form of the protein in which the binding pocket is known
experimentally to house 4 water molecules.

I have calculated the binding energy of each of the ligands using the
generalized born model (igb=2) from the crystal structure via an ordinary
minmization with sander. However, when considering the ligand-free form of
the protein, I have used explicit water (TIP3).

My questions is therefore:

Is it meaningful to obtain the energy of the explicit water in the binding
pocket when the system is immersed in a continuum solvent?

The reason I am doing this and not just examining an empty binding pocket in
an implicit environment is because I believe that hydrogen bonding plays a
major role in stabilising the pocket, and the GB model (as I understand it)
could never accurrately account for the specific hydrogen bonding framework.
Am I right in assuming that the source of the inaccuracy (in examning
explicit water in an implicit solvent) would arise largely from the GB model
attempting to compute the non-polar contribution of the solvation free
energy for each explicit water particle? I say this because my understanding
is that the the non-polar contribution is derived via a surface-area
weighted proportion of an empirically derived organic molecule (that does
not share such properties with a single water particle.)

Any ideas as to how best to model H-bonding in the pocket, whilst still
maintaining the energetics associated with immersion in bulk water?

Many thanks

Hayden Eastwood

The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to
Received on Wed Sep 12 2007 - 06:07:14 PDT
Custom Search