VMD-L Mailing List

From: JC Gumbart (gumbart_at_ks.uiuc.edu)
Date: Fri Jan 04 2013 - 16:48:37 CST

Why are you opening and closing the file with each frame? Move those outside the innermost loop. Also your refsel/tempsel selections aren't doing what you think. You probably want this:

set tempsel [atomselect $traj "protein and ((resid $j and name N P OG) or (resid $j2 and name N CA CG))"]

Any selections that can be made outside of a loop (tempsel2) should be. Finally, delete them when you're done with them, e.g., "$tempsel2 delete".

All that being said, measure contacts is probably what's making it slow. Nothing you can do there.

On Jan 4, 2013, at 2:05 PM, Martin, Erik W wrote:

> For the record, since I first sent this a couple hours ago, I fixed most
> of the problems thusly:
>
>
>
> set filename "contactstest"
> puts $filename
> set cutoff 0.75
> set numfr [molinfo $traj get numframes]
> set firstframe 0
> set findP [atomselect $traj "protein and name P"]
> set Pres [$findP get resid]
>
> set wd40 [atomselect $reference "chain D"]
> set refsel [atomselect $reference "chain F and resid 4 and name N P OG1 or
> resid 5 and name N CA CG"]
> set lastframe [expr {$numfr-1}]
> puts "last frame: $lastframe"
> set mv [atomselect $traj all]
>
> set n 0
> foreach j $Pres {
> incr n
> set j2 [expr {$j+1}]
> set tempsel [atomselect $traj "protein and resid $j and name N P OG or
> resid $j2 and name N CA CG"]
>
> for {set i $firstframe} {$i < $numfr} {incr i} {
> puts "Measuring P site $j in frame $i"
> set outfile [open $filename$n.csv a]
> $mv frame $i
> $tempsel frame $i
> set M [measure fit $tempsel $refsel]
> $mv move $M
> set tempsel2 [atomselect $traj "protein and noh" frame $i]
> set tmp [measure contacts $cutoff $wd40 $tempsel2]
>
> set tmp2 [llength [lindex $tmp 0]]
> puts "$tmp2"
> puts $outfile "$i, $tmp2"
> close $outfile
> }
>
> }
>
>
>
>
> But It still runs really slowly, so any other tips are always appreciated!
>
> Thanks,
> Erik
>
>
>
> On 1/4/13 11:17 AM, "Martin, Erik W" <Erik.Martin_at_stjude.org> wrote:
>
>> Hi, I quickly wrote this script that I was hoping would perform the
>> following function: align each phosphorylated residue in a peptide with
>> a ligand in a crystal structure and loop through frames in a trajectory
>> measuring contacts between the simulated peptide and the protein its
>> "bound" to. This seemed simple enough, but I'm running into all kinds
>> of problems... The output file only contains the info from the last
>> frame, the "tempsel" atomselection does not seem to be updating (or at
>> least the alignment I'm doing doesn't work. Finally, it does not seem to
>> actually be closing the files that I label as "outfile" and I get this
>> error: couldn't open "contactstest4.csv": too many open files
>>
>> If anyone enjoys writing TCL scripts for trajectory analysis and would
>> like to help me with this troubleshooting I'd appreciate it! I thought
>> this was really straightforward...
>>
>> set filename "contactstest"
>> puts $filename
>> set cutoff 1
>> set numfr [molinfo $traj get numframes]
>> set firstframe 0
>> set findP [atomselect $traj "protein and name P"]
>> set Pres [$findP get resid]
>> set mv [atomselect $traj all]
>> set wd40 [atomselect $reference "chain D"]
>> set refsel [atomselect $reference "chain F and resid 4 and name P OG1 or
>> resid 5 and name N CA CG"]
>> set lastframe [expr {$numfr-1}]
>> puts "last frame: $lastframe"
>>
>> set n 1
>> foreach j $Pres {
>> incr n
>> set j2 [expr {$j+1}]
>> set tempsel [atomselect $traj "protein and resid $j and
>> name P OG or resid $j2 and name N CA CG"]
>> puts "test $j"
>> for {set i $firstframe} {$i < $numfr} {incr i} {
>> puts "testA $i"
>> set outfile [open $filename$n.csv w]
>> $tempsel frame $i
>> set M [measure fit $tempsel $refsel]
>> $mv move $M
>> set tempsel2 [atomselect $traj "protein and noh"]
>> set tmp [measure contact $cutoff $wd40 $tempsel2]
>> set tmp2 [llength [lindex $tmp 0]]
>>
>> puts $outfile "$i, $tmp2"
>>
>> }
>> close $outfile
>> }
>>
>> Email Disclaimer: www.stjude.org/emaildisclaimer
>> Consultation Disclaimer: www.stjude.org/consultationdisclaimer
>>
>>
>>
>
>