MORE on Bioperl:Bio:Tools:Blast-> parse - failing on the first report in stream

[Mauricio La Rota]

Hello everybody,

Michael B Thornton said a few weeks ago (see Aug 4th) he was having problem
parsing streams of blast reports, and that
his script was giving him "fatal exceptions" *only* for the first report.

It is very funny because I am having trouble with only the LAST report on
the stream (only 100 reports in the stream for now).  However,  it is not
consistent across different sets of report streams:  sometimes it works well
or sometimes it fails to parse the last report,  but never get an error
message such as the one displayed on Michael's email;  the parser simply
ends silently.     I am using parse_blast.pl script from the
./examples/blast directory with the following command:

cat  myfile.fasta  |  parse_blast.pl  -table 1  -nostats  -signif 1e-15  >
myfile.parsed
(or replace 'cat'  with a filter that eliminates the *** No hits found ***
reports from the stream before piping, to avoid exceptions from blast.pm)

In addition, because the script calls perl with the -w flag,  I get some
extra messages in STDERR that I don't know whether to interpret as errors or
light warnigs:

"Odd number of elements in hash list at
/u/user8/cmlarota/perl/bioperl-0.05/Bio/Root/Object.pm line 515"

Does anybody know how to solve these?

Also,  We are trying to get bioperl to work under NT.  It compiles, but it
breaks when IOManager makes calls to functions not supported by ActivePerl
such as "alarm()".   Is there a way around this?

All comments are greatly appreciated.

Mauricio La Rota
Ms/PhD student
Plant Breeding dept
Cornell University

=========== Bioperl Project Mailing List Message Footer =======
Project URL: http://bio.perl.org/
For info about how to (un)subscribe, where messages are archived, etc:
http://www.techfak.uni-bielefeld.de/bcd/Perl/Bio/vsns-bcd-perl.html
====================================================================