[Bioperl-l] Question - Primary_id

Hilmar Lapp hlapp@gnf.org
Mon, 23 Dec 2002 12:32:33 -0800


On Monday, December 23, 2002, at 05:29  AM, Ewan Birney wrote:

> Out of these three, both display_id and accession_number have pretty
> consistent semantics and usage, whereas primary_id frankly should 
> probably
> be junked - it is too complex an idea to be enforced, and by 
> definition is
> implementation specific.

My $.02 are that it's the definition that should be junked, not the 
method itself. I think there should be a place to store a 
datastore-specific ID which may be different from an accession number. 
GenBank's GI number is a perfect example (and the genbank parser 
presently uses primary_id to store the GI number).

The semantics I'm using it with for bioperl-db, for example, is that it 
is an alternate key by itself, which is helpful as opposed to the 
accession/version/namespace triple you'd have to use otherwise. That 
might not hold though ...

	-hilmar

-- 
-------------------------------------------------------------
Hilmar Lapp                            email: lapp at gnf.org
GNF, San Diego, Ca. 92121              phone: +1-858-812-1757
-------------------------------------------------------------