From: SourceForge.net <no...@so...> - 2004-06-15 01:03:19
|
GO website item #952651, was opened at 2004-05-12 15:25 Message generated for change (Settings changed) made by jl242 You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=600519&aid=952651&group_id=36855 Category: request feature Group: None Status: Open Resolution: None Priority: 5 Submitted By: Jennifer I Clark (jenclark) >Assigned to: Jane Lomax (jl242) Summary: moving GO:ids between ontologies Initial Comment: This is so we can put something in the documenation to say that we shouldn't move GO:ids between ontologies. Rather we should obsolete in one and make a new term in the other. I include an e-mail from Midori to explain how the plan came about. Jen Hi, >> What's the policy on moving terms from one ontology to another? I have >> a term 'microtubule severing activity', which I think might be a >> process; it's not defined. Would it be OK to move it into the process >> ontology, or should it be obsoleted and then moved? I'm not sure I've >> ever come across this situation before! It is pretty rare, which probably explains why there's no 'official' documentation and precious little by way of precedent. I think it would be safest to obsolete the function term and create a new ID for the process term, not because the concept is really different, but to avoid confusion and the attendant acrimony. [Tho I do remember from way back in the mists of time, probably a year BDE (before DAG-Edit), that one of the reasons for not encoding any meaning into GO IDs was Joel Richardson saying that we could not put F/P/C into the IDs unless we could guarantee that no term would ever ever be moved from one ontology to another ... but as it happens there are other reasons, such as paths, for IDs to be nothing more or less than pure unique identifiers. Cheers, m ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=600519&aid=952651&group_id=36855 |