[Go] 1 year limit on IEAs

[Daniel Barrell]

What Harold and Mike have described is correct. The mapping files 
(ec2go, hamap2go, interpro2go, etc) are sanity checked on a weekly basis 
and corrected/extended by curators if needs be. The IEA pipeline is run 
3 weekly after each UniProt release (so, realistically, monthly).

If you are parsing our files directly you should do it at least annually 
to conform to the GO 1 year rule, but ideally you'd be taking them far 
more often.

Cheers

Dan

Mike Cherry wrote:
> UNIPROTs IEA are updated with each of their monthly releases.  So if you 
> get their IEAs at least once a year you're fine.
> 
> The 1 year limit was set way back.  I think back in 2001 for the reasons 
> Suzanna gave.  If the IEAs cannot be updated once a year how actuate are 
> they?
> 
> -Mike
> 
> 
> On Jun 9, 2008, at 10:34 AM, Harold Drabkin wrote:
> 
>> Jim,  if I'm not mistaken (someone from UniProt chime in please) , the 
>> IEAs from UniProt are based on the translation files (ec2go, ip2go and 
>> spkw2go) based on information in a uniprot record (ec#, domains, 
>> keywords). These change  frequently; sometimes domains get added  or 
>> removed, or the translation files change (mappings that are suddently 
>> to recently obsoleted terms, eg.) and so the IEAs go stale, espeically 
>> after 1 year. We update daily. I would really worry if they were older 
>> than 1 month,l et  alone a year.
>>
>> hjd
>>
>> Jim Hu wrote:
>>> Hi All,
>>>
>>> As I recall, this was discussed at the SLC consortium meeting, but I 
>>> find that I can't reconstruct the rationale for the 1 year expiration 
>>> date on IEAs.  For E. coli, we're importing the IEAs from Uniprot, 
>>> but most of them will expire almost immediately.  I assume that there 
>>> is a problem in other genomes of too many IEAs.  But perhaps a limit 
>>> on how many per gene or some such might be more flexible.
>>> For our E. coli websites, I think it makes sense to display IEAs with 
>>> appropriate caveats, since we're trying to encourage the community to 
>>> do quality control on them.  But I'm not sure, and I imagine that 
>>> other mods have thought about this a lot.
>>>
>>> Jim
>>>
>>> =====================================
>>>
>>> Jim Hu
>>>
>>> Associate Professor
>>>
>>> Dept. of Biochemistry and Biophysics
>>>
>>> 2128 TAMU
>>>
>>> Texas A&M Univ.
>>>
>>> College Station, TX 77843-2128
>>>
>>> 979-862-4054
>>>
>>>
>>>
>>> ------------------------------------------------------------------------
>>>
>>> _______________________________________________
>>> Go mailing list
>>> Go at geneontology.org
>>> http://fafner.stanford.edu/mailman/listinfo/go
>>>
>>
>> _______________________________________________
>> Go mailing list
>> Go at geneontology.org
>> http://fafner.stanford.edu/mailman/listinfo/go
> 
> _______________________________________________
> Go mailing list
> Go at geneontology.org
> http://fafner.stanford.edu/mailman/listinfo/go

-- 

Daniel Barrell
EMBL - The EBI
Wellcome Trust Genome Campus
Hinxton, Cambridge CB10 1SD
Phone: +44 (0)1223 492551
Email: dbarrell at ebi.ac.uk