Re: AMBER: GB offset for proteins

From: David A. Case <>
Date: Wed, 24 Dec 2008 14:11:21 -0500

On Sun, Dec 07, 2008, Sally Pias wrote:
> Tutorial A4 points out that offset=0.13 with generalized Born (GB)
> implicit solvation gives results closest to Poisson-Boltzmann for
> nucleic acids (Tsui and Case, JACS 122, 2489, 2000). Has such an
> optimum offset value been determined for proteins? Is the default,
> offset=0.09, appropriate to use for NMR refinement (simulated
> annealing) of a protein?

Sorry to be so slow in replying to this. The default parameters should
be fine for NMR refinement. Note that the details of the force field
mainly become important when you don't have experimental data to define
the structure, and you should always be critical of parts of the
structure where this is the case.

Most recent results seem to indicate that igb=2 or 5 is better than
igb=1 for proteins, and there are some indications that igb=7 is also
better (among those flavors available in Amber). But if you have
enough restrints, you should see few differences that depend on the
choice. You should consider running some experiments yourself.


The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
Received on Fri Dec 26 2008 - 01:09:48 PST
Custom Search