SPKW problem mappings to kinase

[gocentral]

Some sp entries are kinase inhibitors and have the keyword 'kinase'

kinase maps to kinase activity, but this is incorrect for these entries

see O74960 O14128

either the keyword, or the mapping needs to be removed

Thanks

val

Reported by: ValWood

Original Ticket: "geneontology/annotation-issues/468":https://sourceforge.net/p/geneontology/annotation-issues/468

[gocentral]

• milestone: --> GOA
• assigned_to: nobody --> edimmer

Original comment by: mah11

[gocentral]

Logged In: YES user_id=1647919 Originator: NO

Dear Val,

thanks for the feedback

The problem appears to be the electronic assignment of keywords to TrEMBL entries. So the GO mapping is correct and does not need to be changed. I have asked Kati to fix the two entries you mentioned (as I think you already had) and I also sent a a query to the UniProt Help desk to see if any other TrEMBL entries could be cleaned up (see copy of e-mail below).

Evelyn

UniProtKB EMAIL TRACK # 246742

Tracking No.: 070917-493293 Server: EBI

Contact Info:

Name: Evelyn Camon (UniProt Help Desk query 070917-493293) E-mail: (UniProt Help Desk query 070917-493293) camon@ebi.ac.uk Institution: EBI, GOA

Query:

subject: Kinase Keywords in TrEMBL Message:

At the moment the keyword 'kinase' is inherited by TrEMBL entries with DE line 'xxxx kinase inhibitor'. If these proteins are indeed kinase inhibitors then by the definition of the Swiss-Prot Keyword in expasy the Keyword to UniProtKB/TrEMBL assignment is incorrect. As the keyword 'kinase' is mapped to the GO term 'kinase activity', this incorrect keyword assignment creates an incorrect GO term assignment.

I appreciate that these are un-curated TrEMBL entries however it is always nice to make the most accurate electronic annotations possible. Is there any way that we can fix these in TrEMBL? I'm not sure how many there is?

See examples: O74960, O14128 (although Kati says that she will fix these 2 for me)

cheers

Evelyn

Original comment by: camon

[gocentral]

Logged In: YES user_id=516865 Originator: YES

OK, thanks Ev, wasn't sure whether this one was a curation problem or a keyword problem, so covered both bases.!

Cheers

Val

Original comment by: ValWood

[gocentral]

Logged In: YES user_id=1647919 Originator: NO

RESPONSE FROM UNIPROTKB:(via Claire)

This is due to the UniProtKB/TrEMBL creation program which does mappings based on the names provided by the submitter. It is a simple mapping and in some cases throws up errors like this. It also predates our automatic annotation system. We will be reviewing its actions in the near future and these issues should be resolved.

evelyn

Original comment by: camon

[gocentral]

• status: open --> pending

Original comment by: edimmer

[gocentral]

Logged In: YES user_id=1312539 Originator: NO

This Tracker item was closed automatically by the system. It was previously set to a Pending status, and the original submitter did not respond within 90 days (the time period specified by the administrator of this Tracker).

Original comment by: sf-robot

[gocentral]

• status: pending --> closed

Original comment by: sf-robot

[gocentral]

• status: closed --> pending

Original comment by: mah11

[gocentral]

Logged In: YES user_id=837342 Originator: NO

Hi Val,

The kinase keyword was removed from both of these entries on the 23-OCT-2007, so this sourceforge item should be resolved now. Cheers, Emily

Original comment by: edimmer

[gocentral]

• status: pending --> closed

Original comment by: edimmer