[Bioperl-l] UCSC database -> GFF

Paul Edlefsen pedlefsen at systemsbiology.org
Fri Jul 18 11:44:49 EDT 2003


Ewan Birney wrote:

>On Thu, 17 Jul 2003, Allen Day wrote:
>  
>
>I (*think*) they are SAX based...
>
>DAS is not a good way to download full datasets however... it was never
>designed for that and doesn't scale well.
>  
>
This was an in-a-pinch solution because the ucsc2gff script had some
issues.  The right thing to do (tm) is probably to fix the script.  We
ended up using a slick caching mechanism instead, that uses Storable
and gdbm to cache feature objects after they are retrieved, so we
didn't end up gffifying the DAS anyway.  FYI: Nat Goodman wrote that 
(caching) code
and it is CVSed on the freaky dev branch, but I think that he doesn't
feel that it is ready for non-hacky use yet.

:Paul



More information about the Bioperl-l mailing list