[BioSQL-l] [Bioperl-l] Problem loading GO.

[Richard Cote]

Chris Mungall wrote:
>>> Could not store term GO:0006905, name 'vesicle transport':
>>> [...]
>>> There are duplicate terms, identical in the term table except for
>>> GOID:
>>> GO:0006905 and GO:0005480.  They are both "vesicle transport", and
>>> obsoleted:
>>
> I think that its actually wrong to include obsoletes and actual terms in 
> the same table - however, it's obviously astoundingly useful to be able 
> to do this, but it requires the hack to get ou of the uniqueness violation.
> 
> The EBI loads all of OBO into BioSQL regularly - I wonder how they 
> handle this?

I simply avoid the issue. There's no uniqueness constraint in term name. 
The only constraint is term ID, and even that is only unique in the 
context of an ontology namespace (i.e. it would be perfectly allowable 
to have FOO:1234 and BAR:1234). The only unique (and primary) key is 
generated by the ORM layer so I don't even have to deal with that.

We also have all the terms, obsoleted or not, in the same table because 
people are always querying on stuff that's been made obsolete but is 
still annotated with the old IDs.

Cheers,
Rc

-- 
Richard Cote
Software Engineer - PRIDE Project Team (Sequence Database Group)
European Bioinformatics Institute
Wellcome Trust Genome Campus                 rcote at ebi.ac.uk
Hinxton, Cambridge CB10 1SD                  Phone: (+44) 1223 492610
United Kingdom                               Fax  : (+44) 1223 494468