search

geneontology / molecular_function_refactoring

A repository for the development of patterns and terms for axiomatisation and refactoring of the GO MF branch
1 stars 1 forks source link

Make design pattern and template for nuclear receptor #42

Open dosumis opened 7 years ago

dosumis commented 7 years ago

General classification for ligand activated transcription factors: https://en.wikipedia.org/wiki/Nuclear_receptor

Dependency: Needs pattern for (DNA binding) transcription factor activity first

thomaspd commented 7 years ago

A start:

has_part ligand binding occurs in cytosol activates has_part transcription factor activity occurs in nucleus

dosumis commented 7 years ago

Some notes on what we have

We have no single term for nuclear receptors. We do have:

"RNA polymerase II transcription factor activity, ligand-activated sequence-specific DNA binding" ; GO:0004879 'ligand-activated RNA polymerase II transcription factor binding transcription factor activity" ; GO:0038049'

These have related synonym: 'nuclear receptor activity'. Neither of these terms maps up to receptor, although all subclasses do.

The second term seems dubious. There is only a single annotation with experimental evidence in this branch: UniProtKB P04150 NR3C1 GO:0038051 PMID:17635946. As far as I can gather, all nuclear receptors bind DNA and most also bind co-activators etc. This term was probably intended to capture the latter activity, but that's certainly not how it's being used.

Proposal: Collapse this down into a single class

'nuclear receptor activity' EquivalentTo 'receptor activity' that has_effector some 'sequence-specific DNA binding transcription factor activity'

If we want to capture the cofactor/TF binding activity this can be done in LEGO at annotation time. We could also consider adding a specific term for the binding event in question.

dosumis commented 7 years ago

General policy issue: In defining molecular transducer (#31 ), we suggested a pattern in which the effector function was a biochemical activity. In this case we have a compound function as an effector. The molecular transducer pattern should therefore be broadened. We should also make it an explicit policy that we can nest compound functions. This needs to be taken into account when designing LEGO -> GPAD inference rules.

pgaudet commented 5 years ago

See http://noctua.geneontology.org/editor/graph/gomodel:5c4605cc00002111?barista_token=lx3p3kiu02bzr5fj94xb#