AMBER Archive (2009)

Subject: Re: [AMBER] problem in rmsd

From: Carlos Simmerling (carlos.simmerling_at_gmail.com)
Date: Tue Mar 10 2009 - 05:29:20 CDT


300 isn't reasonable for trpcage, even if you use the wrong prmtop.
the molecule just isn't that big. even if you scramble the atoms it
would only be 20 or so. it looks like you're not doing a best fit, or
something like that. you should visually inspect the mdcrd file (since
it's not that large). in order to get rmsd that large, you must have
coordinates in those frames that differ among atoms by 300 or so-
otherwise the rmsd can't be that large. if they look reasonable, it's
something in the fitting.

what would help us is if you
1) calculate the energies at each step (ntpr=1) in addition to writing
them. if the energies are ok, the rmsd can't be that large
2) show us the ptraj script that you used to calculate rmsd.

On Tue, Mar 10, 2009 at 4:30 AM, Hannes Loeffler
<hannes.loeffler_at_stfc.ac.uk> wrote:
> On Mon, 2009-03-09 at 23:36 -0400, Parimal Kar wrote:
>>     I output PDB files for each frame. interesting thing is when I
>> choose the option "new cartoon" in vmd, then the figure disappears.
>
> I think vmd computes the secondary structure (via the stride program)
> only once.  I am not sure on what frame it operates (first?, last?).  It
> happened to me occasionally that vmd gets totally confused and wasn't
> able to display parts or all of a structure anymore.  My advice would be
> not to rely 100% on vmd when you use "new cartoon", "tube" or similar
> representations.
>
>
> _______________________________________________
> AMBER mailing list
> AMBER_at_ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber