On Sun, Oct 24, 2010 at 8:01 PM, Cheng-I Lee <biocil.ccu.edu.tw> wrote:
>
> 1. Equilibration: how to judge if the duration of the equilibration is long
> enough? In the Amber tutorials (A7), 200 ps of equilibration was performed.
> Is 200 ps of simulation usually good enough?
The answer is really going to be system-dependent. Keep in mind that
reaching true equilibirum with your system can be really difficult to
do - what the tutorial refers to is really more of an extension of
minimization to get each replica to a target temperature. For your
simulation just keep an eye on the thermodynamic properties of the
system (at the very least the potential energy and temperature, and
for explicit solvent definitely keep track of the density as well).
Once the running averages and fluctuations start to look like they
aren't changing very much the system is probably OK for future runs,
but keep in mind this still might not be true equilibrium. With
replica exchange you can usually get away with running brief MD like
this to get your replicas to their target temperatures and then rely
on the exchanges themselves to drive you to equilibirum. In that case
just keep in mind that part of your "production run" is really just
equilibration. Note that the tutorial in question is for a quite small
(109 atom) system in implicit solvent; for larger systems more than
200 ps will certainly be needed.
>
> 2. In replica exchange, how many times of exchange should I expect?
Unfortunately there's no rule for this either; you should exchange for
as many times as it takes to see whatever you are hoping to see. You
may also want to read this article for a discussion about how often to
exchange ("Exchange frequency in replica exchange molecular dynamics."
D. Sindhikara et al. J. Chem. Phys. 128, 024103 (2008)).
Hope this helps.
-Dan
>
> Thanks for the help!
>
> Sophia
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Oct 25 2010 - 05:00:03 PDT
