AMBER Archive (2007)Subject: Re: AMBER: sander and pmemd
From: Robert Duke (rduke_at_email.unc.edu)
Date: Mon Dec 03 2007 - 08:19:24 CST
Nadia -
There should be no problems with mixing; the calcs done for any given task
you wish to perform should be equivalent in both pmemd and sander for any
task that pmemd supports.
Regards - Bob Duke
----- Original Message -----
From: "Nadia Vahdati" <nadiav_at_soton.ac.uk>
To: <amber_at_scripps.edu>
Sent: Monday, December 03, 2007 8:20 AM
Subject: AMBER: sander and pmemd
> Dear Amber users,
>
> I am simulating a solvated protein system ~65000 atoms.
>
> I usually minimise with a serial version of sander-9. In the equilibration
> stage and production, I then switch to a parallel version and use around
> 16
> processors.
>
> I now have access to a parallel version of PMEMD-9 with large number of
> processors to use.
>
> - I understand that pmemd is a reimplementation and should replicate
> sander
> but with much better speed of performance. Is it ok to mix sander and
> pmemd? ie. use sander for the minimisation and
> heating and then pmemd for production?
>
> -or should I only use pmemd? and just reduce the number processors used
> during the minimisaiton and heating?
>
> Thank you in advance for any advice,
>
> Nadia
>
> --
> Nadia Vahdati
> Postgraduate Research Student
> Prof. J.W. Essex Group
> University of Southampton
> UK.
>
>
>
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber_at_scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
|