Re: AMBER: md and implicit solvation

From: David A. Case <case.scripps.edu>
Date: Tue, 24 Oct 2006 18:26:15 -0700

On Tue, Oct 24, 2006, Jardas sucuriba wrote:
>
> I ran 50 ps of an MD simulation (equilibration phase) using implicit
> solvation. However my calculation took 3 days to complete! Is it normal to
> take so long? I am really worried because I want to run a production phase
> of 1 ns. I wonder if someone could tell me if there is something wrong with
> my script or pehaps suggest me modifications to improve it. My system has
> around 15000 atoms.

This sounds like a very large system for GB (around 900 residues, if it's a
protein). Unfortunately, it is not surprising that things will be pretty
slow, since Amber is not well-optimized for large molecule gb simulations.
(Of course, as Adrian pointed out, it's hard to be sure without knowing more
about what sort of machine you are using.) Some ideas:

1. try setting rgbmax to something like 16 -- that will probably help some
   with a very small impact on accuracy;
2. use pmemd if you are not already doing that...it should be somewhat faster
   than sander.
2. see if you can find a parallel machine: gb simulations usually scale pretty
   well. (Even finding a two processor dual core machine would turn a
   two-month simulation into one month...)

...good luck...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 Oct 25 2006 - 06:07:30 PDT
Custom Search