[Biopython] Fwd: BlastParsing gives Value Error: Invalid header?

[Martin Mokrejs]

Hi Peter,

Peter Cock wrote:
> On Wed, May 9, 2012 at 7:01 PM, Martin Mokrejs
> <mmokrejs at fold.natur.cuni.cz> wrote:
>> Hi Brad,
>>  I just got bitten by this myself as well. Could be the legacy blast parser
>> improved to give clearer error message? E.g. that it failed to find the LOCUS
>> line or whatever was it looking for? With the legacy BLAST documentation being
>> gone from current Tutorial it is easy to pick the wrong parser. ;)
>>
>>  And BTW, please do not drop support for legacy BLAST. I just cannot make BLAST+
>> give me same alignment, no matter what arguments I use to adjust for the (it gives me
>> wider alignment than wanted and I can make it a look shorter, but shortening it just
>> a bit like legacy BLAST output .. is not doable).
> 
> Have you contacted the NCBI about this possible regression?

No, not yet.

>>  And, took me a while to find old biopython-1.52.tar.gz to lookup the old docs.
>> Could there be a hyperlink from Tutorial to these unpacked, browseable sources? ;)
>> I am speaking about http://biopython.org/DIST/docs/tutorial/Tutorial.html#htoc85 .
>> Thanks,
>> Martin
> 
> Could you clarify if you are talking about documentation for calling
> the 'legacy'
> BLAST command line tools (e.g. blastall), or documentation for parsing the

Yes, "blastall -p blastn ...", the default plaintext pairwise output (-m 0),
version 2.2.24.

> plain text human readable output (which still exists in BLAST+)?
> 
> On a related point, Bow's just done a bit of work updating our plain
> text parser to
> cope with BLAST+ (specifically changes in BLAST 2.2.25+ and/or 2.2.26+).
> 
> One of the aims of Bow's GSoC project will make dealing with the different
> BLAST formats a lot simpler.

Its great that we have GSoC students, would I have some spare time I would mentor
one. Good luck and thanks for your care, Peter!
Martin