[Go] growth in manual GO annotation

Chris Mungall cjm at fruitfly.org
Thu Apr 10 12:15:25 PDT 2008


On Apr 10, 2008, at 10:08 AM, Doug howe wrote:

> Oopps...let me be more clear..I'm looking for:
>
> 1.  The number of distinct gene products (across all species)  
> annotated using NON-IEA, NON-ND evidence on 1/1 of each year from  
> 2002-2008.

SELECT count(DISTINCT gene_product_id) AS num_gps
FROM association INNER JOIN evidence ON  
(evidence.association_id=association.id)
WHERE code != 'IEA' AND code != 'ND';

go_old_20030101
num_gps
42746

go_old_20040101
num_gps
99116

go_old_20050101
num_gps
144635

go_old_20060101
num_gps
136734

go_old_20070101
num_gps
140370

go_old_20080101
num_gps
192535


> 2.  The number of distinct species with any NON-IEA, NON-ND GO  
> annotation on 1/1 of each year from 2002-2008.

SELECT count(DISTINCT species_id) AS num_species
FROM gene_product
  INNER JOIN association ON  
(gene_product.id=association.gene_product_id)
  INNER JOIN evidence ON (evidence.association_id=association.id)
WHERE code != 'IEA' AND code != 'ND';
go_old_20030101
num_species
207

go_old_20040101
num_species
375

go_old_20050101
num_species
533

go_old_20060101
num_species
638

go_old_20070101
num_species
884

go_old_20080101
num_species
930

(yep, these numbers are correct, there is a lot of non-MOD  
annotations to GO)

>
> Doug howe wrote:
>> Thanks Chris those are very useful numbers.  If you don't mind  
>> running two more queries, it won't be necessary to open the older  
>> stuff to Goose.
>> I'd be interested to see:
>> 1.  The number of distinct gene products (across all species)  
>> annotated on 1/1 of each year from 2002-2008.
>> 2.  The number of distinct species with any GO annotation on 1/1  
>> of each year from 2002-2008.
>>
>> -Thanks!
>> -Doug
>>
>>
>> Chris Mungall wrote:
>>
>>> On Apr 8, 2008, at 9:48 AM, Doug howe wrote:
>>>
>>>
>>>> Does anyone have, or know how to get, historical stats on the  
>>>> number of
>>>> GO annotations that have been contributed to the GOC over time?   
>>>> I'm
>>>> looking for the number of non-IEA, non-ND GO annotations that  
>>>> existed
>>>> for each year from 2002-2008.
>>>>
>>>> Midori provided me with the following numbers of GO terms for that
>>>> period if anyone is interested:
>>>> date            total       obsolete
>>>> 1/1/2002    10305    152
>>>> 1/1/2003    13339    383
>>>> 1/1/2004    16771    725
>>>> 1/1/2005    18219    969
>>>> 1/1/2006    20348    992
>>>> 1/1/2007    22928    1011
>>>> 1/1/2008    25758    1137
>>>>
>>> We have historical go dbs mirrored here - we can open these to  
>>> GOOSE if you like, or you can just request queries.
>>>
>>> This is what you're after:
>>>
>>> SELECT count(*) AS num_annots
>>> FROM association INNER JOIN evidence ON  
>>> (evidence.association_id=association.id)
>>> WHERE code != 'IEA' AND code != 'ND';
>>> go_old_20030101
>>> num_annots
>>> 133699
>>>
>>> go_old_20040101
>>> num_annots
>>> 386339
>>>
>>> go_old_20050101
>>> num_annots
>>> 416224
>>>
>>> go_old_20060101
>>> num_annots
>>> 469107
>>>
>>> go_old_20070101
>>> num_annots
>>> 489402
>>>
>>> go_old_20080101
>>> num_annots
>>> 580052
>>>
>>>
>>> This one may also be informative: the number of terms used  
>>> directly in annotations (all):
>>>
>>> SELECT count(DISTINCT term_id) AS num_terms_used_directly
>>> FROM association;
>>> go_old_20030101
>>> num_terms_used_directly
>>> 7116
>>>
>>> go_old_20040101
>>> num_terms_used_directly
>>> 9008
>>>
>>> go_old_20050101
>>> num_terms_used_directly
>>> 10134
>>>
>>> go_old_20060101
>>> num_terms_used_directly
>>> 11113
>>>
>>> go_old_20070101
>>> num_terms_used_directly
>>> 12340
>>>
>>> go_old_20080101
>>> num_terms_used_directly
>>> 13812
>>>
>>>
>>>
>>>
>>>> -Doug
>>>>
>>>> _______________________________________________
>>>> 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
>>
>>
>



More information about the Go mailing list