[Gofriends] Redundancy in go_XXXXXX-assocdb-tables/dbxref.txt

[Fiona McCarthy]

Hi Gabriel,

I am not sure about rat but I did notice earlier in the year that Ensembl
was reporting GO in a very strange way - it was directly attributing GO
annotations from human orthologs to the chicken & cow genes. This was
causing a lot of apparent redundancies when I scanned the GO annotations
but when I checked them more carefully, many of the Ensembl GO annotations
made no sense.

This happened in version 49 and was supposed to be fixed in version 50.
Does this sound like what you are seeing?

Fiona


Gabriel Berriz <gberriz at hms.harvard.edu> on Monday, September 08, 2008 at
4:49 PM +0000 wrote:
>Dear GO friends,
>
>
>For some species, the info given in the dbxref table includes IDs from
>multiple databases, which raises the possibility of "cryptic
>redundancies", i.e. associations distinct because they are assigned IDs
>from different databases that in fact refer to the same underlying gene
>product.  For example, if I compare the sets of rat dbxref's that have
>database names RGD and ENSEMBL respectiely, I find that the overlap
>(redundancy) of these two sets of IDs consists of about 1000 IDs, which
>is over 5% all the possible rat gene products.  (To compute this overlap,
>I first mapped the RGD IDs to ENSEMBL IDs using the mappings provided by
>Ensembl version 49.)
>
>
>Would it help to avoid these "cryptic redundancies" if a single database
>(Ensembl, RGD, whatever) was used for each species?
>
>
>Thanks for your comments,
>
>
>Gabriel Berriz



The AgBase Databases
Department of Basic Sciences
Box 6100
MS 39762-6100 
Mississippi State University
USA
Tel:   (+ 1) 662 325 5859
Fax:  (+ 1) 662 325 1031

http://www.agbase.msstate.edu/