[Bioperl-l] Bio::Ontology support for several indentifiers per term

[Hilmar Lapp]

The GO term itself really has only one identifier (check out the 
GO.defs file). GO terms acquire secondary accessions when they are 
merged with other, subsequently obsoleted terms. There should be a 
document somewhere at geneontology.org that explains in detail what 
will happen to the identifiers upon merging or obsoleting terms.

You can get the secondary identifiers using $term->get_secondary_ids(). 
The existing ontology engine implementations do not index by secondary 
ids.

	-hilmar


On Wednesday, January 21, 2004, at 03:10  AM, Pär Engström wrote:

> Hi BioPerl developers,
>
> In the Gene Ontology, some terms have multiple identifiers. E.g. 
> "ATPase
> activity" has identifiers GO:0016887 and GO:0004002. However, it seems
> that the Bio::Ontology modules only support one identifier per term.
>
> Consider the following script (using bioperl 1.4):
>
> use Bio::OntologyIO;
> my $go_in = Bio::OntologyIO->new
>     ( -format => "go",
>       -defs_file => "GO.defs",
>       -files => ["component.ontology",
> 		"function.ontology",
> 		"process.ontology" ]);
> my $gene_ontology = $go_in->next_ontology();
> print scalar($gene_ontology->find_terms("GO:0004002")), " ";
> print scalar($gene_ontology->find_terms("GO:0016887")), "\n";
>
> The output is "0 1", i.e. the term can only be found from one of its
> identifiers.
>
> If this is a known issue in Bio::Ontology, is there some way to 
> retrieve
> terms by their "secondary" identifiers?
>
> Regards,
>
> Par Engstrom
> Graduate student, CGB, Karolinska
>
>
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at portal.open-bio.org
> http://portal.open-bio.org/mailman/listinfo/bioperl-l
>
-- 
-------------------------------------------------------------
Hilmar Lapp                            email: lapp at gnf.org
GNF, San Diego, Ca. 92121              phone: +1-858-812-1757
-------------------------------------------------------------