On Wed, Feb 04, 2004, ML wrote:
> I am trying to calculate normal modes using nmode via mm-pbsa for a
> 3200-atom protein. With the help from the AMBER e-mail archive, I was
> able to compile AMBER7 on 64-bit AMD Opterons for this system.
>
> To hold down the memory needed for nonbond pairlist, I adjusted
> mm_pbsa.pl script to read in cutoff for nonbonded interactions. Sander
> minimization goes fine until gradient of 10E-4 (or even 10E-5) but nmode
> starts with a 10* or so worse gradient and fails! (How do sander and
> nmode calculate rms gradients? Do they differ in that? Why?)
Nmode re-creates the pairlist when it starts up; this is probably different
than the one that was being used by sander when it got to a low gradient.
For a variety of reasons, it is almost never a good idea to try to minimize
things with a pairlist in the Amber codes. Unfortunately, a 3200 atom system
is really pushing the capabilities of codes that were originally written in
the late 1970's to handle systems of 300 atoms. But, are you sure that you
cannot do a non-cutoff simulation? (You don't need the eigenvectors). For
what it is worth, Amber 8 will have a "smaller memory" option on nmode that
will reduce memeory usage, allowing somewhat larger systems to be run.
...good luck...dac
--
==================================================================
David A. Case | e-mail: case.scripps.edu
Dept. of Molecular Biology, TPC15 | fax: +1-858-784-8896
The Scripps Research Institute | phone: +1-858-784-9768
10550 N. Torrey Pines Rd. | home page:
La Jolla CA 92037 USA | http://www.scripps.edu/case
==================================================================
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Wed Feb 04 2004 - 16:53:00 PST
