AMBER Archive (2008)

Subject: Re: AMBER: Gbsa in NAB

From: David A. Case (
Date: Thu Apr 10 2008 - 12:24:17 CDT

On Thu, Apr 10, 2008, Hu, Shaowen (JSC-SK)[USRA] wrote:
> Thank you very much for your help. I am doing some tests on NAB with a
> section of DNA. I found, if I turned on gbsa, the structures of DNA are
> very distorted during the trajectories, but if I set gbsa=0, the
> trajectories are good comparing to Sander calculation. I have tried
> three GB models, two force fields(ff99SB and ff99bsc0), and various cuts
> and heat baths.

NAB and sander should give identical results for gbsa. The test case in
amber10/test/nab/xfin_sa.out.check is the same as the test case in
$AMBERHOME/test/gbsa_xfin, and both give the same results. I would be
very interested to hear of examples where the energies are different.

I would agree that one generally does not want to use gbsa for nucleic acids.
But that is different than saying that sander and NAB give different results.

Of course, there may well be a bug somewhere, which is why a test case would
be very helpful.


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