Mappings to protein binding

[gocentral]

Hello,

I was wondering whether it made sense to have InterPro domains mapped to "protein binding" (GO:5515). It seems to me, all proteins bind something?

For example:

-InterPro:IPR000210 BTB/POZ -InterPro:IPR000488 Death -InterPro:IPR000693 Anenome neurotoxin -InterPro:IPR000724 IgG-binding B -InterPro:IPR000959 POLO box duplicated region -InterPro:IPR001219 Neurotoxin

etc. I counted 64 in the interpro2go file I have.

Thanks, Pascale

Reported by: pgaudet

Original Ticket: "geneontology/annotation-issues/184":https://sourceforge.net/p/geneontology/annotation-issues/184

[gocentral]

• milestone: --> InterPro
• assigned_to: nobody --> orchard

Original comment by: mah11

[gocentral]

Logged In: YES user_id=516865

Hi Pascale,

I would agree with this, I don't think they are particulary useful, although Interpro may have a reason for wanting to keep them. Ev may like them becasue they make the GOA stats look good ;). I filter out all Interpro mappings to 'protein binding' and just 'binding' (although I keep things like ATP/GTP binding, DNA/RNA binding.

Val

Original comment by: ValWood

[gocentral]

Logged In: YES user_id=1056932

I agree, in the majority of cases this GO mapping is not particularly useful.

I think it is okay, however, in some cases where the only known information about a functional domain is something like "protein-protein interactions".

Original comment by: kpilcher

[gocentral]

• labels: --> InterPro mapping

Original comment by: mah11

[gocentral]

Logged In: YES user_id=545117

Hi

I try to restrict this mapping to domains, where it is relevant information. I know it was stuck on families early on in the mapping process, but i have been removing these, as I find them.

Sandra

Original comment by: orchard

[gocentral]

• status: open --> closed

Original comment by: orchard