[Bioperl-l] Bio::Tools::Run::Primer3 question

[Josh Burdick]

Sean Davis wrote:

>I am using Bio::Tools::Run::Primer3 to find internal oligos.  I feed in
>sequences and get results (code below), but the code reliably dies at the
>same place (after the same number (about 247) of primer3 runs) regardless of
>the sequence passed as input.  Can anyone shed some light on this behavior?
>
>  
>
   I ended up writing my own parser (which doesn't use all the 
inheritance stuff that Bioperl seems to, but doesn't run into this 
problem either.)  If anyone wants to use it, I'm happy to contribute it.
   I was getting the same problem.  It was fairly mystifying.
   My guess is that it's opening a new input file for each run, and not 
closing it, and so it runs out of file descriptors. I didn't follow this 
up, because I couldn't see how files were being opened (the actual 
opening of files was, I think, in Bio::Root::IO, which says it closes 
files automatically when an object is destroyed.  Perhaps these 
finalizers weren't being called?  This is strictly a guess...)  If you 
have /usr/sbin/lsof on your system, you perhaps could use that to see if 
the Perl process is using lots of file descriptors.  At any rate, I 
didn't understand the code well enough to fix the problem.

>Thanks,
>Sean
>
>  
>
    Josh

-- 
Josh Burdick
jburdick at gradient.cis.upenn.edu
http://www.cis.upenn.edu/~jburdick