AMBER Archive (2005)

Subject: Re: AMBER: problem with solvateBox

From: David A. Case (case_at_scripps.edu)
Date: Fri Feb 25 2005 - 15:41:38 CST


On Fri, Feb 25, 2005, ReichertD_at_mir.wustl.edu wrote:
>
> I've seen a similar problem in the archives but no posted solutions. Using
> PGI compilers (version 5.2) on a RHEL3 linux box with dual AthlonXP's
> everything compiles and tests out Ok except solvateBox or solvateOct in
> xleap and tleap. Using solvateBox model TIP3PBOX 8, the waters are added
> but about 5 Ang from the protein. It's driving me nuts, different flags
> when compiling make no difference. On a PIII box and the Intel Fortran
> compiler, this works correctly using the same proteins and commands. Can
> anyone suggest a starting point ? I already shelled $$ for the PGI
> compilers and really don't want to pay for the Intel as well.

tleap and xleap are pure C codes, and run fine with gcc. You don't need
the Portland Group or the Intel compilers.

Or, prepare the systems on the PIII computer, and copy the resulting files
over to the Athlon's. The place you really need compute power is for the
simulation itself, not for the setup phase.

(Also, Nathan Baker (Dept. of Biochemistry and Molecular Biophysics) is an
local Amber/Opteron (and maybe Amber/Athlon?) expert...maybe he can give
some advice.)

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