Re: [AMBER] Simulation stops without errors

From: Jason Swails <jason.swails.gmail.com>
Date: Thu, 17 Sep 2015 07:28:56 -0400

On Thu, Sep 17, 2015 at 5:36 AM, Bala subramanian <bala.biophysics.gmail.com
> wrote:

> Thank you for the information.
> I visualized the trajectory but nothing seemed strange reg. the reported
> atoms 1307 and 1308, these atoms were Nz of LYS and one of the hydrogen
> connected to it. Actually this problem occurred not at the beginning, but
> after 100ns production run. I tried running a short 50ps simulation with
> 0.5fs time step to relax any severe clash and then continued the production
> and then switched to 2fs. Now it is running fine.
>

​It may have been a transient collapse of one atom into the center of
another. It happens on rare occasions due to some hydrogen atoms having a
L-J epsilon parameter equal to zero.

The "solution" is usually just to restart the simulation with the last good
restart and assign a new random seed. If it is running fine now, I would
guess it was an unfortunate rare event, and you did the correct thing to
recover.

HTH,
Jason

-- 
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Sep 17 2015 - 04:30:04 PDT
Custom Search