[Biojava-dev] Blast SAX parser update

[Keith James]

>>>>> "Simon" == Simon Brocklehurst <simon.brocklehurst@CambridgeAntibody.com> writes:

[...]

    Simon> Also, I'm sure that you've captured all the changes between
    Simon> formats, but just in case... I have a dim and distant
    Simon> memory that NCBI made a minor change to the summary ouput
    Simon> lines for one or more versions of Blast between versions
    Simon> 2.0.* and 2.2.*.  Something like adding a frame number for
    Simon> translations maybe, or number of HSPs? Kind of like some of
    Simon> the difference in summary line output between Wu-Blast and
    Simon> NCBI-Blast?  Did you get that one?

I made diffs of (t)blast(npx) between 2.0.11/2.2.3 and didn't spot
that difference. I've just eyeballed the raw output again with the
same result. Maybe it was a transient change between 2.0.*, 2.1.* and
some 2.2.* version?

Before I started I googled to see whether anyone had a rogues gallery
of old blast outputs, but no luck. Previously I've raided the
BioPython test dataset, but theirs was less useful this time.

    Simon> I could easily be wrong of course...  memorizing the
    Simon> precise formats of different versions of Blast really isn't
    Simon> one of my major hobbies ;-)

Thanks for the warning - I'm keeping my eyes open for strange
behaviour. I'll be running more parallel searches + diffs between
2.0.11 and 2.2.3 to see if I can smoke out any more differences. The
tests could be made more extensive too.

Keith

-- 

- Keith James <kdj@sanger.ac.uk> bioinformatics programming support -
- Pathogen Sequencing Unit, The Wellcome Trust Sanger Institute, UK -