From: Luis Gracia (lug2002_at_med.cornell.edu)
Date: Thu Aug 24 2006 - 14:08:48 CDT

without looking at the trajectory it is quite difficult to tell what's
going on.
Can you add rmsd calculation to your script? A modified script is
appended (haven't check it thought)
Does your simulation come from a PBC simulation? it might have something
to do with the wrapping options in the simulation?

Luis

> proc align_frames {} {
>
> # use frame 0 as reference
> set reference [atomselect top all frame 0]
>>> set reference2 [atomselect top "protein and noh" frame 0]
>>> set fid [open "rmsd2.out" w]
>
> # the frame being compared
> set compare [atomselect top all]
>>> set compare2 [atomselect top "protein and noh"]
> 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
>>>
>>> $compare2 frame $i
>>> set rmsd [measure $compare2 $reference2]
>>> puts $fid "$i $rmsd"
> }
>>> close $fid
> }