VMD-L Mailing List

From: John Stone (johns_at_ks.uiuc.edu)
Date: Wed Apr 30 2003 - 10:05:19 CDT

Hi Thomas,
  Which version of VMD are you running? We'll take a look and
see if there's any reason why compare move would give you trouble as
you describe below. I did notice that the ref2_xx atom selections
at the end of the script are not being deleted BTW, though I'll take
your word for it that the problem you're having is caused during the
atom selection move command. Let me take a look and get back to you on
this.

Thanks,
  John Stone
  vmd_at_ks.uiuc.edu

On Wed, Apr 30, 2003 at 04:28:01PM +0200, Thomas Hedegaard Pedersen wrote:
> I need help!!!
>
> My question probably relates very much to the earlier email posted to
> VMD-L by Joe Huang:
>
> # Memory problems with TCL scripts in VMD. Joe Huang (Thu Mar 20 2003 -
> 13:02:48 CST)
>
> I'm trying to do an RMSD calculation for each residue in a protein as
> function of time. I have produced a Tcl script for this purpose (see
> below) but when I run the script it gradually uses up all of my memory
> causing the computer to freeze. By successively commenting out lines in
> the script I could pin point the cause of the problem down to the line:
>
> "$compare2 move $trans_mat1"
>
> located inside the loop in the proc - if this line is commented out
> VMD's memory usage becomes relatively small and constant through out the
> run! Am I doing something completely wrong when using the "move" command
> or is it something else??????
>
> Thanks for your help
>
> Thomas H. Pedersen
> Tecnical University of Denmark (DTU)
> Kemitorvet build. 206 room 243
> DK-2800 Kgs. Lyngby, Denmark
>
>
> Tcl-script:
> ----------------------------------------------------------------------------
> # Calculates the RMSD of every residue as function of time.
> #Written by Thomas Hedegaard Pedersen
> #Technical University of Denmark
> #Kgs. Lyngby
>
> #REMEMBER TO SET BIGDCD-PATH, PSF and PDB FILENAMES, DCD FILENAMES,
> SELECTION#
> ##############################################################################
> source /home/Thomas/scripts/vmd/PCA/bigdcd.tcl
>
> #set sys
> set sys 1az8
>
> #set rmsd id
> set id RESID_CA
>
> #set sim id
> set sim SIM4_SIM4B
>
> #set paths
> set psf "/home/Thomas/${sys}/${sys}_psfgen.psf"
>
> set dcdpath1 "/home/Thomas/${sys}/SIM/dcd/${sys}_300K_SIM4.dcd"
> set dcdpath2 "/home/Thomas/${sys}/SIM/dcd/${sys}_300K_SIM4_B.dcd"
>
> set out_dir "/home/Thomas/${sys}/SIM/rmsd/Residues/"
>
> #set reference pdb
> set pdb "/home/Thomas/${sys}/${sys}_EQ_frame_1999.pdb"
> set ref EQ1999
>
> #set time between frames (picoseconds)
> set time_betw_frame 0.5
>
> #REMEMBER to do the following selections as less time demanding as
> possible as
> #this will have a profound impact on the performance
>
> #set fit selection
> set selection "protein and name CA"
>
> #set rmsd selection
> set rmsd_selec "protein and name CA"
> ##############################################################################
>
> #set ref_frame - only 0 (corresponding to pdb-file)
> set ref_frame 0
>
> set name1 "${out_dir}rmsd_${sys}_${sim}_${id}_ref${ref}.dat"
> if {[file exists $name1] == 1} {
> return "file $name1 already exists - remove it and run the program
> again"
> } else {
> set fid20 [open $name1 a]
> }
>
>
> proc rmsd_resid { frame } {
>
> global time_betw_frame selection rmsd_selec fid20 reference1 num_resid
>
> # the selection and frame used for fitting
> set compare1 [atomselect top $selection frame $frame]
>
> # compute the transformation
> set trans_mat1 [measure fit $compare1 $reference1]
>
> $compare1 delete
> unset compare1
>
> puts "Frame : $frame"
>
> puts -nonewline $fid20 "[expr {$frame * $time_betw_frame}] "
>
>
> # Calculate rmsd on the desired selection residue wise
>
> for {set n 0} {$n <= [expr {$num_resid-1}]} {incr n} {
>
> upvar #0 ref2_$n ref
>
> #the selection and frame used for rmsd
> set compare2 [atomselect top "($rmsd_selec) and (residue $n)"
> frame $frame]
>
> #####################################################################
> #THE FOLLOWING LINE SEEMS TO CAUSE THE PROBLEM REGARDING MEMORY USAGE
> #####################################################################
> # do the alignment
> $compare2 move $trans_mat1
> #####################################################################
>
> # compute the RMSD
> set rmsd1 [measure rmsd $compare2 $ref]
>
> puts -nonewline $fid20 "$rmsd1 "
>
> $compare2 delete
> unset compare2
> unset rmsd1
> }
>
> unset trans_mat1
>
> puts $fid20 ""
>
> flush $fid20
> }
>
> # read in psf-file
> mol load psf $psf
>
> #set reference for fit
> set reference1 [atomselect top $selection frame $ref_frame]
> puts "reference for fit set to selection: $selection"
> puts " at frame: $ref_frame\n"
>
> #set reference for rmsd
> set residues [atomselect top "(same residue as ($rmsd_selec)) and (name
> CA)"]
> set num_resid [$residues num]
>
> for {set n 0} {$n <= [expr {$num_resid-1}]} {incr n} {
> set ref2_$n [atomselect top "($rmsd_selec) and (residue $n)" frame
> $ref_frame]
> }
> puts "reference for residues set to selection: $rmsd_selec"
> puts " at frame: $ref_frame\n"
>
> #read in reference pdb-file
> animate read pdb $pdb
>
> # read in dcd files
> bigdcd rmsd_resid $dcdpath1 $dcdpath2 

-- 
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