bioperl blast parsing problem

[rhinoceros]

Maybe your problem is related to an outdated XML::SAX? (first note). Just an idea..

[Mark]

Yea, I know but I don't do the blast updates for our system and so can't take that route to a fix. Thanks for trying to help though.

[rhinoceros]

You mean stitle/salltitles? But yeah, 2.2.28+ only.

[Mark]

Subject description. Yes I know you could grab the oridinal ID and the go back to the blastdb and pull out the record and parse it's description from there but that's a bit of a pain so I would rather find out if there is a solution to the parsing of the xml output.

[rhinoceros]

Unfortunately, you can't get the hit description via that method, a serious shortcoming.

What do you mean by hit description? There are like 40 fields to choose from..

[Mark]

Unfortunately, you can't get the hit description via that method, a serious shortcoming.

[rhinoceros]

Can't really help you there but have you considered other output formats (e.g. outfmt 6) and simply parsing your files with unix commands?

[Mark]

version 1.006901

[kmcarr]

Which release of BioPerl are you using?

[Mark]

We recently upgraded our installation of blast from v2.2.19 to blast+ 2.2.26 and I have found that my bioperl blast xml parsers no longer work properly (typically they return some form of truncated result). Has anyone else encountered this and if this is a known problem have either the NCBI or BioPerl folks provided a solution.

Thanks

Mark