[annotation] RNA trimethylguanosine cap binding query

[Harold Drabkin]

I do think that is the more sensible thing to do.
hd


Valerie Wood wrote:
> Why don't we obsolete GPI binding  and create a new term (it only has 31 manual annotations which appear to be mainly incorrect), plus the IEA mappings, which we have already established are wrong.
>
> I can't think of any other binding term which may have been used inconsistently in this way. I don't really know why we did this, but as now see  I originally requested the term I will take the blame for establishing this bad practice. 
>
> I have an alternative way to curate PTMs (and I guess most others do) so I'll just reannotate using this method. It would be useful to have a way to collect PTMs consistently with and between orgs, but as Karen C said, they don't belong in GO.
>
> We could then add a comment to all of the other binding terms to say that they are ONLY intended  for non-covalent modifications.
>
> Would everyone be happy with this?
>
> Val
>
>
>
>
> Harold Drabkin <hjd at informatics.jax.org> wrote: 
>   
>> To use binding for the creation of a covalent bond is just absolutely 
>> incorrect.  This is just basic undergraduate chemistry. 
>>
>> Modifying the definition is fine; I would add both statements.  
>> However,  if anyone has used this in this way it is wrong. Those 
>> annotations should stricken when found.  I cannot see any "but".  
>> Although we do not obviously have the time to revisit each one, in the 
>> cases Val found,  they should be removed; they were incorrect use of the 
>> term binding.
>>
>>
>> hjd
>>
>> Alexander Diehl wrote:
>>     
>>> While I agree that common annotation practice has been to exclude 
>>> covalent bonds between gene products and other entities (including 
>>> other proteins), the fact that this has not be specifically excluded 
>>> by the definition of 1135 MF binding terms suggests to me the there 
>>> could be in fact be many annotations using these binding terms that 
>>> include covalent bonds.  Per standard GO practice, we should obsolete 
>>> all 1135 binding terms, provide equivalent, but redefined new binding 
>>> terms, to force people to reexamine their annotations.
>>>
>>> Clearly, we are not going to do this.  So in fact, I recommend that we 
>>> do not change the definitions at all, although perhaps we could add a 
>>> comment like, "Note that binding terms are not intended for annotation 
>>> of covalent bonds."
>>>
>>> -- Alex
>>>
>>>
>>> Midori Harris wrote:
>>>       
>>>> Has this gone into SF? (hint, hint)
>>>>
>>>> should be easy to fix ...
>>>>
>>>> On Wed, 16 Jan 2008, Harold Drabkin wrote:
>>>>
>>>>         
>>>>> Hopefully a script can be designed to add this as a sentence to the 
>>>>> end of each definition!
>>>>>
>>>>> hjd
>>>>>
>>>>> Karen Christie wrote:
>>>>>           
>>>>>> while you're right that terms inherit from their parents, it's 
>>>>>> still probably better to fix it everywhere, so that people who only 
>>>>>> look at a specific term and not all its parents get the 
>>>>>> non-covalent idea clearly.
>>>>>>
>>>>>> -Karen
>>>>>>
>>>>>>
>>>>>> On Wed, 16 Jan 2008, Harold Drabkin wrote:
>>>>>>
>>>>>>             
>>>>>>> I actually noticed that the core term, binding, with def of :    
>>>>>>> The selective, often stoichiometric, interaction of a molecule 
>>>>>>> with one or more specific sites on another molecule.
>>>>>>> is missing the non-covalent idea; fixing it in this term should 
>>>>>>> suffice for it's children
>>>>>>>
>>>>>>>
>>>>>>> The Oxford bible: binding: the act or process by which one 
>>>>>>> molecule attaches to another by noncovalent forces.
>>>>>>>
>>>>>>>
>>>>>>> Valerie Wood wrote:
>>>>>>>               
>>>>>>>> yes thanks it seems obvious now, although I think I have some 
>>>>>>>> rogue annotations to fix for GPI binding to fix.! This might be 
>>>>>>>> obvious to everyone, but it might be worth adding 'non covalent' 
>>>>>>>> to all of the binding term defs just to emphasise. Its quite 
>>>>>>>> clear when you read the 'RNA trimethylguanosine cap binding def, 
>>>>>>>> but I'm not sure that it is for some of the others. thanks
>>>>>>>>
>>>>>>>> val
>>>>>>>>
>>>>>>>>                 
>>>       
>>
>>     
>
>