[Bioperl-l] 0.7 release: Blast

Paul Gordon gordonp@niji.imb.nrc.ca
Tue, 14 Nov 2000 16:21:32 -0400 (AST)

> My humble suggestion is to wait a little longer until XML output becomes
> stable, and then write a new Blast parser which uses the DTD to parse the
> file and build the respective necessary data storage object(s).  Then
> deprecate the use of the old Blast parsing (I'd guess that those who use
> BioPerl for large scale blast parsing would also be able to control their
> blast output format option), in favor of XML-parsing.
Yes, it will take a while for that to stabalize.  There are still quite a
few issues with the lossiness of the XML from Blast that I've been meaning
to send them...

Anyway, although that would certainly be a nice way to do it (XML queries
on the document for each method instead of parsers), I don't think that
such a module would replace the current one because there are lots of
people who won't want to produce XML reports for one reason or another,
especially people using command line Blast and looking at the text

<snip /> 

> P.S. I'll reiterate our desire for someone to present us with a Grand
> Unified DTD for sequence comparison/alignment, usable by both Blast and
> FASTA and others ... and then tell us how after meeting the spec of the
> DTD, we can extend it for our own internal purposes (data types that the
> rest of the world probably won't care about and therefore shouldn't be in
> the grand unified dtd).
Could a basis for this come from the parent class of Blast and FastA?