Re: AMBER: Building the parallel version in AMBER 8 and GIBBS

From: David A. Case <case.scripps.edu>
Date: Tue, 9 Nov 2004 21:07:05 -0800

On Tue, Nov 09, 2004, Ilyas Yildirim wrote:
>
> I have a dual processor PC (two P3-600Mhz CPU), linux installed on it.
> (RedHat WS Enterprise) I have installed AMBER 8 and when I tried to follow
> the parallel version installation, it failed. I used the following
> commands:
>
> # make clean
> # ./configure -mpi ifort

There is generally no "native" mpi on a Linux machine. You have to install
something like MPICH, and use the -mpich (not -mpi) as an argument to
configure.

>
> Another question I have is about GIBBS. I am going to do some free
> energy perturbations and am planning to use GIBBS for that. Can I do the
> free energy perturbation in sander too? If not, is it possible to
> explicitely install just GIBBS into AMBER 8? I have the previous ver of
> AMBER, too. (AMBER 7)

I sure I don't understand the question. You can use Amber7 gibbs just as
before. Just remember that free energy perurbation calculations in gibbs
puts some severe restrictions on the force field you can use; in particular,
it does not support the PME or GB options, nor will it support "extra point"
calculations. You are hence transporting yourself back about 10 years in time
in terms of the type of underlying simulation you can do. It is for this
reason that _most_ users are probably better off carrying out free enegy
simulations in sander.

.....dac

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Wed Nov 10 2004 - 05:53:00 PST
Custom Search