[go] Requirement for all 'unknown' annotations to use ND code

[Chris Mungall]

On Sep 17, 2007, at 9:44 AM, Valerie Wood wrote:

>
> Would it be useful to know which other codes we have used for root  
> node annotations (in addition to the ND ones you mention here).
> i.e have people used TAS?

It's quite a motley collection, and yes TAS has been used a bit:

http://www.berkeleybop.org/goose?database=1&sql_query=SELECT+term.name 
%2C+evidence.code%2C+count%28DISTINCT+association.id%29%0D%0AFROM+++% 
0D%0A+association%0D%0A+INNER+JOIN+evidence+ON+%28association.id% 
3Devidence.association_id%29%0D%0A+INNER+JOIN+term+ON+% 
28association.term_id%3Dterm.id%29%0D%0AWHERE%0D%0A+term.name+IN+%28% 
27molecular_function%27%2C%27biological_process%27%2C% 
27cellular_component%27%29%0D%0AGROUP+BY+term.name%2C+evidence.code% 
3B&limit=1000&mirror=1

You can also see this broken down by ontology, code and DB:

http://www.berkeleybop.org/goose?database=1&sql_query=SELECT+term.name 
%2C+xref_dbname%2C+evidence.code%2C+count%28DISTINCT+association.id% 
29%0D%0AFROM+++gene_product%0D%0A+INNER+JOIN+dbxref+ON+% 
28gene_product.dbxref_id%3Ddbxref.id%29%0D%0A+INNER+JOIN+association 
+ON+%28gene_product.id%3Dassociation.gene_product_id%29%0D%0A+INNER 
+JOIN+evidence+ON+%28association.id%3Devidence.association_id%29%0D%0A 
+INNER+JOIN+term+ON+%28association.term_id%3Dterm.id%29%0D%0AWHERE%0D% 
0A+term.name+IN+%28%27molecular_function%27%2C%27biological_process% 
27%2C%27cellular_component%27%29%0D%0AGROUP+BY+term.name%2C 
+xref_dbname%2C+evidence.code%3B&limit=1000&mirror=1

If your email reader breaks up the above URLs you can go to

http://www.berkeleybop.org/goose

and try the following for the breakdown by ontology, code:

SELECT term.name, evidence.code, count(DISTINCT association.id)
FROM
association
INNER JOIN evidence ON (association.id=evidence.association_id)
INNER JOIN term ON (association.term_id=term.id)
WHERE
term.name IN  
('molecular_function','biological_process','cellular_component')
GROUP BY term.name, evidence.code;

and the following for breakdown by ontology, code and DB:

SELECT term.name, xref_dbname, evidence.code, count(DISTINCT  
association.id)
FROM   gene_product
INNER JOIN dbxref ON (gene_product.dbxref_id=dbxref.id)
INNER JOIN association ON (gene_product.id=association.gene_product_id)
INNER JOIN evidence ON (association.id=evidence.association_id)
INNER JOIN term ON (association.term_id=term.id)
WHERE
term.name IN  
('molecular_function','biological_process','cellular_component')
GROUP BY term.name, xref_dbname, evidence.code;


> We seem to have established that ND should not be used to retrieve  
> root node annotations.
>
> However, I'm not sure that we should use TAS and a paper as root  
> node annotations should reflect the fact that checks have been done  
> that no other paper has been published/ similarity identified/ from  
> which an inference/annotation could be made.
> If you were curating a paper which was one year old  and the author  
> stated there was no functional information, as a curator wouldn't  
> you check that there was no subsequent info?
>
>
>
>
> Mike Cherry wrote:
>
>> Duh, I was thinking no results for NR.
>>
>> Never mind...
>>
>> -Mike
>>
>> On Sep 15, 2007, at 8:46 PM, Karen Christie wrote:
>>
>>> Hi Mike,
>>>
>>> Thanks for the info, but the NR code is not the one that is used   
>>> for annotations to the root nodes. NR stands for Not Recorded  
>>> and  is obsolete for new usage.
>>>
>>> -Karen
>>>
>>>
>>>
>>
>
>
>
> -- 
> The Wellcome Trust Sanger Institute is operated by Genome Research  
> Limited, a charity registered in England with number 1021457 and a  
> company registered in England with number 2742969, whose registered  
> office is 215 Euston Road, London, NW1 2BE.