[Bioperl-l] bioperl-db: It's aaaaaaliiiiiive

Ewan Birney birney@ebi.ac.uk
Mon, 9 Apr 2001 22:30:06 +0100 (BST)

I have been bringing bioperl-db up to full storage of Bio::SeqI objects.
It handles:

   Bio::SeqI and the following components:

Bio::PrimarySeqI, Bio::SeqFeature::Generic (and in general
Bio::SeqFeatureI compliant features - but of course they will be stored
"as if" they are Bio::SeqFeature::Generic), Bio::Location::Split/Simple
(no Fuzzies, but should be easish to add in. hohohoho. No remote features.
I want to see how the EMBL schema handles local/remote features before I
take that on), Bio::Annotation::DBLink, Bio::Annotation::Comment, no
Bio::Annotation::Reference and Bio::Species yet.

Bio::Species and some other db info is kept as singletons. Bio::SeqI
implementing objects is implemented in a "lazy" fashion of delaying object
gets until it needs it, but then cach'ing the get. This works only really
as a read-only concept.

I have done two GenBank round-trips and it looks pretty good. Mainly
whitespace problems in the diff. The lack of full fuzzy support is felt
for those pesky '<' and '>' on CDS lines.

But - hey - it works and the code is super-clean, conforming to the
Ensembl adaptor scheme and well tested with a total of 102 tests already.

So - does anyone want to take it for a drive? I think you'll like
it.... ;)


Ewan Birney. Mobile: +44 (0)7970 151230, Work: +44 1223 494420