HELP!: GO Term Error

[Midori Harris]

On Thu, 15 Aug 2002, Pankaj Jaiswal wrote:

> Well I disagree with Midori !
> Reason at some point in an organism's life if one looks logically, each and
> every one of the gene product can be called as a nutrient reserve in some form
> or the other. All the proteins can be broken down in its components for example
> into  peptides ->amino acids->C/H/N/O/P/S/.

Certainly -- but this is not an _activity_ catalyzed by the proteins (or
lipids, carbohydtates, or nucleic acids) themselves. 'Nutrient reserve' is
itself far from an ideal GO term, largely for exactly this reason. The
term is being created to accommodate those proteins that are, as far as we
know, synthesized _solely_ for the purpose of stashing amino acids.

 This is what actually happens when
> an organism undergoes a situation where it becomes its necessity to survive by
> recycling its own organization units i.e. at the time of "aging". Proteins is
> just one example, same is true for nucleotides and carbohydrates and lipids.

This is covered by GO process terms such as 'aging', 'apoptosis',
'response to starvation', etc.

> Therefore I still feel that this term "storage protein of seed" should be
> brought back.

You have not addressed either of the problems that prompted us to remove
'seed storage protein'; 'storage protein of seed' is merely a rewording
of the same thing.

> Another reason that these proteins in "Monocot plants" also have
> an additional function of "serine protease inhibitor". 

This has nothing to do with the merits (or lack thereof) of the terms
'seed storage protein', 'storage protein of seed', or 'nutrient reserve'.
These proteins can and should be annotated to the existing term 'serine
protease inhibitor' (GO:0004867).

> There are several
> instances of such proteins as Tanya already mentioned about them in her first
> mail at the start of this debate.(cited below) 
> 
> http://www.geneontology.org/email/go-arc/go-2002/2028.html
> 
> "Midori: There are many different seed storage proteins: vicilin-type 
> globulin, albumin, glutenin/gliadin, C hordein, etc. "

These are all gene products. Their activities should be represented by GO
terms, but not the proteins or protein families.

Midori