AMBER Archive (2003)

Subject: Re: AMBER: Basic potential question:

From: Wayne Dawson (dawson_at_nih.go.jp)
Date: Thu Dec 25 2003 - 21:04:37 CST

Piotr Cieplak writes:
> I would not use amber ff in discover - this is a very ancient version
> of parametrization, not very useful for any purpose now.

One can extend the life-time of discover using Prof. Dupradeau's
AmberFFC, at least keeping in mind the caveats he has stated himself.

In my humble opinion, I can see that the newer Amber ffs are
remarkably better in their predictive capacity, but even the older
Amber is not all that bad. If there is an old SGI sitting around with
a license for discover, I see no problem using it for comparison
purposes or _assisting_ in a straight-forward, well-defined model
building exercise.

..WD

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