Obsolete specific transcription factors binding terms

[pgaudet]

Hello,

According to our rules not to capture specific gene products, I think those terms should be obsoleted:

Number EXP annotations	Term
10	vitamin D receptor binding
39	thyroid hormone receptor binding
15	Tat protein binding
35	retinoid X receptor binding
26	retinoic acid receptor binding
14	progesterone receptor binding
7	NFAT protein binding
42	NF-kappaB binding
1	mineralocorticoid receptor binding
25	glucocorticoid receptor binding
65	estrogen receptor binding
11	aryl hydrocarbon receptor binding
38	androgen receptor binding

---

Also - NFAT protein binding has a strange LD:

binding and ('has input' some 'NFAT protein') and ('has input' some 'nuclear factor of activated T-cells, cytoplasmic 1') and ('has input' some 'nuclear factor of activated T-cells, cytoplasmic 2') and ('has input' some 'nuclear factor of activated T-cells, cytoplasmic 3') and ('has input' some 'nuclear factor of activated T-cells, cytoplasmic 4')

How can we spot those ?

Thanks, Pascale

[RLovering]

Not all of these terms are single gene products. NF-kappaB is a complex and it is helpful to have this term as the bound subunit is not always identified - while 35 proteins are associated with this term based on IPI, 10 are IDA, plus there are some IC and TAS proteins too.

Although probably the rest are single gene products, or small families

Best

Ruth

[deustp01]

As long as the small molecule that a particular protein factor binds can be fully captured in the "with" part of the annotation, no information is lost. But do GO rules and GO-CAM current functionality actually support such annotation? (Unclear to me from the GO-CAM jamboree discussion)

[pgaudet]

Right, I am also unsure about this one. What triggered me to look is the NFAT binding logical definition, that looks wrong.

And if we want to capture those we are missing many. As usual- either we get rid of them or we keep on building the ontology.