VMD-L Mailing List

From: Sara baretller (sarabiocomputation_at_gmail.com)
Date: Fri Jul 08 2011 - 09:52:16 CDT

Thank you so much for your respond, the changes to the scripts seems to be
very helpful. i tried for 8 frames and i got different angles per each frame
( 8 frames so far) did not want to run it for teh whole trajectory until i
figure out the open leak. for now i am just adding the
# $sel1 delete
# $sel2 delete

 some scripts are using the animate command and at the end of the script
they write animate delete all. i am not using the animate , but the for
loop. o i have to delete the frames too or no need?

Thank you

On Wed, Jul 6, 2011 at 11:26 AM, Axel Kohlmeyer <akohlmey_at_gmail.com> wrote:

> On Wed, Jul 6, 2011 at 10:38 AM, Sara baretller
> <sarabiocomputation_at_gmail.com> wrote:
> >
> > I have been having a problem with this script , I will really appreciate
> > your help. so the script does get me the angle between two vectors but
> then
> > it will not go to the next loop. it stops in the first. by giving the
> angle
> > between the first vector and the vector {0 1 0 }.
> >
> >
> >
> > Script:
> >
> > for {set j 0} {$j < 6} {incr j} {
> >
> > set outfile [open "out.dat" "w"]
>
> you open the file _inside_ a loop. i don't think
> that is intended.
>
> that will reopen the file in every iteration,
> _or_ if you are on an operating system like
> windows, will may even refuse to open it for
> writing a second time.
>
> >
> > set mol [molinfo top]
>
> this line can also be hoisted outside the loop.
>
> > set sel1 [atomselect $mol "residue $j and resname LEU and
> > name BBc"]
> >
> > set sel2 [atomselect $mol "residue $j and resname GLU and
> > name BBc"]
>
>
>
> > $sel1 frame $j
> > $sel2 frame $j
>
> hmmm.... you have j as variable for residue number
> what does this have to do with the frame number?
>
> you most likely want to write a nested loop, where the
> "outside" loop would loop over another variable, say 'i'
> and then over the number of frames.
>
> > $sel1 update
> > $sel2 update
>
> you selection text is so that the selected atoms won't change over
> the trajectory. no need to update the selection again. better use
> the frame key word to the atomselect keyword right away.
>
> (even faster would be to make a pair of selection for each of the
> 'j' residues and store them in an array for the loop over the trajectory
> frames and then delete them only at the end. this way you avoid the
> time consuming selection update process entirely).
>
> > puts "residue $j of 6 peptides"
> > puts "$sel1 $sel2"
> >
> > set nLEU [llength [$sel1 list]]
> > set nGLU [llength [$sel2 list]]
> >
> > set coord1 [$sel1 get { x y z }]
> > set coord2 [$sel2 get { x y z }]
> >
> >
> > set vec1 [vecsub [lindex $coord1 0] [lindex $coord2 0]]
> >
> > set vec2 {0 1 0}
> >
> > set vec1l [veclength $vec1]
> > set vec2l [veclength $vec2]
> >
> > if {$vec1l != 0} {
> > set vec1 [vecnorm $vec1]
> > }
> > if {$vec2l != 0} {
> > set vec2 [vecnorm $vec2]
> > }
> >
> > set dotprod [vecdot $vec1 $vec2]
> >
> > set angle [expr acos($dotprod)]
> >
> > set angle [expr ($angle * 180/3.14159)]
> > puts "Final angle : $angle"
> > puts $angle
>
> you are at the end of a nested loop in which you
> generated atom selections with the atomselect command,
> but didn't "delete" them. this will cause a memory leak
> and depending on the number of trajectory frames you may
> run out of memory and/or VMD will crash or even crash
> your machine. this is probably the most common mistake
> VMD users make when writing analysis scripts. have a closer
> look at the VMD user's guide section on writing trajectory
> analysis scripts and also search the mailing list archives
> for previous discussions on the subject of memory leaks.
> there are plenty and there is a lot to learn from them.
>
> cheers,
> axel.
>
>
> > }
> >
> > close $outfile
> >
> > tcl councel:
> >
> > angle 1.5284741404344042
> > Final angle : 87.57519131337723
> > 87.57519131337723
> >
> >
> > Thank you
> >
> > Sara
> >
> >
>
>
>
> --
> Dr. Axel Kohlmeyer
> akohlmey_at_gmail.com http://goo.gl/1wk0
>
> Institute for Computational Molecular Science
> Temple University, Philadelphia PA, USA.
>