[Bioperl-l] Scoring an overlap

Chad Clark bioperl-list at superfrink.net
Sun Sep 23 16:21:44 UTC 2007


On Sat, 22 Sep 2007, Florent Angly wrote:

> So finally, here I am trying to score my alignments. For alignments of 2
> small sequences, it works, but as soon as the sequences get bigger than
> a few dozen nucleotides, it crashes:
> Segmentation fault (core dumped)

In my experience if a program runs on small sets of data and segfaults on 
larger sets it is likely running out of stack space.  You can try changing 
the allowed stack size with "ulimit" before running your program and see 
if it works with more data.

   [chad at water ~]$ ulimit -a | grep -i stack
   stack size              (kbytes, -s) 10240
   [chad at water ~]$ ulimit -s 40960
   [chad at water ~]$ ulimit -a | grep -i stack
   stack size              (kbytes, -s) 40960

I don't know the code / algorithm in question but it might require 
significantly more stack space as the data set grows in size so this 
change might not help enough.

Good luck,
Chad



More information about the Bioperl-l mailing list