[Annotation] false positive IPR domain

[Judith Blake]

Doug,
submitting things and not having them in the tracker happened to Harold 
last week.  Perhaps there is a bug?  or someone else has open files?

judy

Doug howe wrote:
> I've submitted the request twice an don't see it in the tracker..not 
> sure what the problem is there.  Anyway, here is the body of it:
>
> I'm not entirely sure how one determines if an interpro domain called 
> for a specific protein is a false positive or not.
>
> The zebrafish proteins Q6NY23, A4JYG2 and B0R032 all have the interpro 
> domain IPR001014 "Ribosomal_L23_CS" associated.  The homologous Human 
> proteins do not have this domain.  The translation via interpro2go 
> produces a number of GO annotations for this gene product (kirrel) 
> that lead you to believe it is a ribosomal protein when in fact it 
> appears to be a transmembrane protein.
>
> Is this a case of false positive calls for IPR001014 on these proteins?
>
> -Doug
>
> edimmer at ebi.ac.uk wrote:
>> Yes - these methods have greatly benefited from external curator 
>> feedback.
>>
>> The InterPro2GO mapping resource in particular has been greatly improved
>> by information passed on from curators from lower eukaryotic/prokaryotic
>> databases, sometimes there can be an unintentional bias in the mapping
>> curators have made to domains; as the proteins in UniProt increases 
>> so the
>> validity of the GO mappings to InterPro domains needs to be reviewed.
>>
>> Thanks!
>> Emily
>>
>>  
>>> For anyone who can't find this  tracker it is here:
>>>
>>> https://sourceforge.net/tracker/?atid=605890&group_id=36855&func=browse
>>>
>>> If we all consistently submit  our false positives from the mappings 
>>> when
>>> spot them, it will rapidly improve the electronic mappings for all
>>> users..... most incorrect mappings affect many organisms.
>>>
>>> However, the mappings (Interpro and Uniprot keywords) are pretty good
>>> these days though, so perhaps we have almost reached the point where 
>>> most
>>> of the false positives have been weeded out already? This should be the
>>> case for any domains and families which are conserved in yeasts.
>>>
>>> Val
>>>
>>>
>>> edimmer at ebi.ac.uk wrote:
>>>    
>>>> Hi Doug,
>>>>
>>>> Exactly as Rama says, you can add it to the Annotation tracker - and
>>>> assign it to 'interhelp'.
>>>>
>>>> Thanks,
>>>> Emily
>>>>
>>>>      
>>>>> Doug,
>>>>>
>>>>> Annotation tracker on SF?
>>>>>
>>>>> Rama
>>>>>
>>>>> On Aug 18, 2008, at 2:12 PM, Doug howe wrote:
>>>>>
>>>>>        
>>>>>> To whom should we direct inquiries about false positive interpro
>>>>>> domain hits?  The alleged false positive is percolating through to
>>>>>> our GO annotations and causing spurious interpro2go translations I
>>>>>> think.
>>>>>>
>>>>>> -- 
>>>>>> Doug Howe, Ph.D.
>>>>>> ZFIN Scientific Curator
>>>>>> Zebrafish Nomenclature Coordinator
>>>>>>
>>>>>> _______________________________________________
>>>>>> Annotation mailing list
>>>>>> Annotation at geneontology.org
>>>>>> http://fafner.stanford.edu/mailman/listinfo/annotation
>>>>>>           
>>>>> _______________________________________________
>>>>> Annotation mailing list
>>>>> Annotation at geneontology.org
>>>>> http://fafner.stanford.edu/mailman/listinfo/annotation
>>>>>
>>>>>         
>>>> _______________________________________________
>>>> Annotation mailing list
>>>> Annotation at geneontology.org
>>>> http://fafner.stanford.edu/mailman/listinfo/annotation
>>>>
>>>>
>>>>       
>>> _______________________________________________
>>> Annotation mailing list
>>> Annotation at geneontology.org
>>> http://fafner.stanford.edu/mailman/listinfo/annotation
>>>
>>>     
>>
>>
>> _______________________________________________
>> Annotation mailing list
>> Annotation at geneontology.org
>> http://fafner.stanford.edu/mailman/listinfo/annotation
>>
>>   
>