FYI: BlastXML, Re: Bioperl: SeqAlign BLAST parsing

Wayne Parrott
Thu, 28 Oct 1999 13:55:29 -0500

FYI: I developed two Java frameworks for manipulating BLAST results. The
first is the Blast Parsing Framework (BPF), a non-XML, SAX-like
framework for parsing and converting a BLAST result into a tokenized
stream, or an in-memory object model. The output can be easily mapped
into other document formats, e.g., HTML, table delimited records, etc.
The second is BlastXML which sits on top of the BPF and converts a BLAST
result into an XML document. BlastXML includes a DTD, example XSL for
IE5, and XML parser (IBM's xml4j). 

The software is open-sourced and may be downloaded from the site, The DTD and
example BlastXML report are available at The site includes a sample report
that is formatted using XSL (IE5 support only). 

Note - the software has only been tested on Java 1.1.X and the jar files
are known to be problematic under Java 1.2. I plan to address this soon
(my day job keeps getting in the way).

I've tinkered with the idea of porting to Perl but just haven't had the
time. If the group addresses this with a Perl implementation I would
like to participate in discussions regarding the DTD design prior to

Please contact me if you desire more info or experience download or
operation problems.

 Wayne Parrott                   email:      |              voice: (972)491-3704                 |
 "Distributed Object Technology    fax: (972)491-7284                 |
   for Life Sciences"              web: |
 "The main thing, is to keep the main thing, the main thing" 
   lyrics by Scott Krippayne

Catherine Letondal wrote:
> "Andreas Matern" writes:
> > With a local installation of the NCBI toolbox, one has the option of
> > creating SeqAlign output for blast results.  Although obviously not meant to
> > be human readable, the output does have the benefit of being more easily
> > parsed than the conventional BLAST output.  Are any groups out there using
> > the SeqAlign option of blastall?  Just wondering if I'm going down a
> > dead-end road that someone else has already traveled.
> A very, very useful feature would just be to have an XML Blast output. XML documents
> are easy to parse, and benefit from a wide range of tools, __which allow to
> build custom parsers__ at a little cost. Several biologists here at Pasteur
> have already asked us (the computing center) about such a feature.
> Sincerely,
> --
> Catherine Letondal -- Pasteur Institute Computing Center
> =========== Bioperl Project Mailing List Message Footer =======
> Project URL:
> For info about how to (un)subscribe, where messages are archived, etc:
> ====================================================================
=========== Bioperl Project Mailing List Message Footer =======
Project URL:
For info about how to (un)subscribe, where messages are archived, etc: