Hi,
Yes, what I ran is not a periodic boundary case. Actually I have already set nscm = 1000 but it still does not work. The coarse grained model I used is Ca-sidechain model, basically every amino acid is represented by two beads. The force field comes from a statistical potential. The coarse grained model should have a large conformational change than all-atom simulation. So is changing the size of the virtual box a solution to this error?
Thanks.
Sincerely,
Qian
----- Original Message -----
From: Ross Walker <ross.rosswalker.co.uk>
Date: Monday, May 23, 2011 10:42 pm
Subject: Re: [AMBER] system extended beyond the virtual box
To: 'AMBER Mailing List' <amber.ambermd.org>
> Hi Qian,
>
> I assume you are running a gas phase or implicit solvent
> simulation here
> rather than a periodic one yes?
>
> What has happened is that your system is drifting, likely due to
> it having a
> net center of mass translational motion. Unchecked this would
> ultimatelyoverflow the output files so needs to be addressed.
> The simplest approach is
> to set nscm=1000 to periodically remove center of mass motion.
> This should
> hopefully stop your system drifting away from the starting point.
>
> Of course is you have modified the potential in some way and
> your gradients
> are incorrect then your system will no longer preserve angular
> momentum or
> linear momentum therefore you will have all sorts of problems.
> You do not
> specify exactly what you mean by coarse grain simulation, which
> method,force field etc so it is hard to know if anything was
> modified. I would
> begin by setting nscm=1000 and see if this helps.
>
> Good luck,
>
> Ross
>
> > -----Original Message-----
> > From: Qian Wang [mailto:qwang.mail.uh.edu]
> > Sent: Monday, May 23, 2011 8:29 PM
> > To: amber.ambermd.org
> > Subject: [AMBER] system extended beyond the virtual box
> >
> > Hi,
> >
> > When I used Amber10 to simulate protein folding, I obtained the
> > following error message:
> > system extended beyond the virtual box
> > All energys recorded in the last frame are normal. I checked the
> > mailing list and someone said this error message is related to
> the non-
> > bond list. So I set nsnb=1000000000, which is equal to my
> nstlim. But
> > it did not solve the problem. My simulation is a coarse-grained
> > simulation so the conformation should have a large fluctuaion.
> Is this
> > the problem to cause the error? Can I change the size of the virtual
> > box to a large value, like 100 Angstroms? Or there are other
> solutions.> Thanks a lot.
> >
> > Sincerely,
> > Qian
> > _______________________________________________
> > 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
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon May 23 2011 - 21:30:02 PDT
