[Bioperl-l] blastall & StandAloneBlast

Matthew Laird biolist at brinkman.mbb.sfu.ca
Wed Mar 2 13:36:43 EST 2005

Hi all,

I'm yet again being faced with a mysterious crash in blastall and Bioperl 
that has been occuring for the past year.  I'm receiving more reports from 
people around the world using our software also experiencing this problem, 
and the only answer I once received about the problem was, "That shouldn't 
be possible."

Anyhow, the error occurs when blastall is called from StandAloneBlast.pm, 
Blast returns with a -1 error code which causes the Bioperl module to 
throw an exception.  We've had reports of this occurring on multiple Linux 
distributions as well as on Solaris and OS X.  But it doesn't happen on 
all machine even if they're running the same distribution.

The crash output is below....

Fatal error:
------------- EXCEPTION  -------------
MSG: blastall call crashed: -1 /usr/local/blast/blastall -p  blastp  -d
"/usr/local/psort/conf/analysis/sclblast/gramneg/sclblast"  -i
/var/tmp/6m0QxSirC3  -e  1e-09  -o  /var/tmp/AKCDNMCTyo  -F  F

STACK Bio::Tools::Run::StandAloneBlast::_runblast
STACK Bio::Tools::Run::StandAloneBlast::_generic_local_blast
STACK Bio::Tools::Run::StandAloneBlast::blastall
STACK Bio::Tools::Run::SCLBlast::blast
STACK Bio::Tools::PSort::Module::SCLBlast::run
STACK Bio::Tools::PSort::Pathway::__ANON__
STACK Bio::Tools::PSort::Pathway::traverse
STACK Bio::Tools::PSort::classify
STACK (eval) /usr/local/psort/bin/psort:320
STACK toplevel /usr/local/psort/bin/psort:320


The line in StandAloneBlast.pm we track the problem back to is:
$self->throw("$executable call crashed: $? $commandstring\n")unless 
($status==0) ;

Odd thing is, Blast DOES run.  If one comments out this line in 
StandAloneBlast.pm, the execution succeeds perfectly fine.  When I've 
editted the error message being thrown to give more details, perl says the 
error is related to a process not being able to be created, which is even 

So, for some odd reason either blastall is passing back this -1 or perl is 
giving it back to bioperl for whatever reason.  The only advice we have 
for people is to comment out this line in StandAloneBlast.pm.

Anyone have any thoughts of advice on where this problem is coming from?


Matthew Laird
SysAdmin/Developer, Brinkman Laboratory, MBB Dept.
Simon Fraser University

More information about the Bioperl-l mailing list