From: Johannes Müllegger (hannes_at_zymeworks.com)
Date: Thu Aug 24 2006 - 13:50:20 CDT

I've had a closer look at the problem. When I align the frames with
the RMSD Trajectory Tool I always get the same spikes. When I use my
script (which I attach at the end of the mail) I get different
spikes, but the same again if I repeat the procedure (i.e. load the
trajectory, align the frames, measure rmsd).
I attach pdb files of the rmsd graphs:

spikes.pdf -> aligned, calculated with the RMSD Trajectory tool
spikes_myscript.pdf -> aligned with the attached script, calculated
with the RMSD Traj. Tool
no_spikes.pdf -> ran align twice with the RMSD Traj. tool, then
calculate

The trajectory I load consists of 5 files. The first containing only
one frame being the crystal structure I start of with, and 4 files
which contains 1000 frames each. When I load only the last 1000
frames I don't see the behavior, also I had a look at the coordinates
of the frames - that cause trouble - in the pdb file but couldn't
detect anything wrong with them.

How can I find out what build I am using?

Thanks for your help
-hannes-


proc align_frames {} {

     # use frame 0 as reference
     set reference [atomselect top all frame 0]

     # the frame being compared
     set compare [atomselect top all]
     set num_fr [molinfo top get numframes]

     for {set i 0} {$i < $num_fr} {incr i 1} {
         # get the current frame
         $compare frame $i

         # compute the transformation
         set trans_mat [measure fit $compare $reference]

         # do the alignment
         $compare move $trans_mat
        }
}

On Aug 23, 2006, at 8:27 PM, John Stone wrote:

>
> Hi,
> Is the behavior you're observing deterministic, or is it random?
> If you load the same trajectory and perform the same sequence of
> RMSD and alignment operations in two successive VMD sessions, do
> you get precisely the same results? I've just recently run the Linux
> version of VMD through valgrind while running Luis' RMSDTT plugin
> to measure RMSD and align trajectory frames, and didn't see any
> signs of problems, nor did I see such spikes in my testing. Which
> version of VMD are you using, the 32-bit or 64-bit build?
> Depending on whether the behavior you're seeing is deterministic or
> random, I may have some further experiments for you to try on your
> machine.
>
> John Stone
> vmd_at_ks.uiuc.edu
>
> On Wed, Aug 23, 2006 at 05:25:58PM -0700, Johannes Müllegger wrote:
>> Dear VMD users,
>>
>> I am observing strange spikes in my rmsd calculations on
>> trajectories using the 'RMSD Trajectory Tool' (VMD 1.8.4 on Gentoo
>> Linux). The spikes seem to appear at different timepoints and only
>> one of the frames at a time is affected. I have aligned the frames
>> with either the function that is included in the RMSD Trajectory plug
>> in (Reference molecule top, protein and noh) or with a script that I
>> wrote (using the selection [atomselect top all]) with the same
>> outcome. When I redo the alignment the spikes disappear. The height
>> of the spikes is about 0.1 to 0.3 above the other values which are ~1
>> and there are from 1 to 4 of these spikes in a trajectory of 4000
>> frames.
>> Has anybody observed similar behavior or is that a specific problem
>> of my trajectory files? Anything I can do to avoid that?
>>
>> Thanks for your help
>> -hannes-
>>
>>
>> _/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
>>
>> Johannes Müllegger
>> Enzyme Engineer
>> Zymeworks Inc.
>> 201 - 1401 West Broadway
>> Vancouver, BC, Canada
>> V6H 1H6
>> Phone: 604 678 1388 -29
>> Fax: 604 737 7077
>> www.zymeworks.com
>>
>> _/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
>>
>>
>>
>>
>
> --
> NIH Resource for Macromolecular Modeling and Bioinformatics
> Beckman Institute for Advanced Science and Technology
> University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
> Email: johns_at_ks.uiuc.edu Phone: 217-244-3349
> WWW: http://www.ks.uiuc.edu/~johns/ Fax: 217-244-6078
>