Re: An outlier appear on the curve of RMSD

From: JC Gumbart (gumbart_at_ks.uiuc.edu)
Date: Sun Jul 26 2009 - 21:14:22 CDT

I've seen this before; it's a failure of the RMSD algorithm, not your
simulation. You might try calculating it in VMD for JUST the
anomalous frame. Chances are it will be fine then.

On Jul 26, 2009, at 8:42 PM, wayj86 wayj86 wrote:

> Dear Leo,
>
> Thank you for your advice. I have visually inspected the
> trajectory, and found the frame in which an outlier appeared. The
> protein seemed to "expand" a bit compared with the previous frame,
> and was restored in the later frame. No obvious moving could be
> seen. So I was wondering whether it was caused by the random error
> or temporary system instability.
>
> Best wishes,
> Stanley
>
>
>
> 2009/7/25 Leonardo Trabuco <ltrabuco_at_ks.uiuc.edu>
> Dear Stanley,
>
> On Thu, Jul 23, 2009 at 10:37 PM, wayj86 wayj86<wayj86_at_gmail.com>
> wrote:
> > Dear NAMD users,
> >
> > I came across a strange problem after performing equilibration
> with NAMD,
> > that was, an outlier appeared on the curve of RMSD. It always
> appeared at
> > the rise period of the curve, much high than the other values.
> For example,
> > If the average of the RMSD was 2.2, that it could go to more than
> 3. After
> > the outlier the curve came back to the normal state and the curve
> contained
> > only one outlier. This might happen in the calculation with
> cluster or
> > standalone computer and the outlier made the curve ugly. So could
> you please
> > tell me the reason? I am confused because the outlier might
> disappear when I
> > redo the equilibration again with the same conditions.
>
> Have you visually inspected the trajectory to see what happened in the
> frame corresponding to your outlier?
>
> Cheers,
> Leo
>
> > Best wishes,
> > Stanley
> >
> > --
> > The future is now!
> >
>
>
>
> --
> Leonardo Trabuco, Ph.D. candidate
> Theoretical and Computational Biophysics Group
> University of Illinois at Urbana-Champaign
>
>
>
> --
> The future is now!

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:51:16 CST