SourceForge meeting 2013-09-25

From GO Wiki
Jump to navigation Jump to search


What did we resolve for 'targeting' terms - do they really imply that the targeted substance actually makes it to the target?

Stems from the assert report # 47 and 48, summary copied from an email from Paola:

ADD GO:0036228 'protein targeting to nuclear inner membrane' GO:0051169 'nuclear transport'

ADD GO:0044744 'protein targeting to nucleus' GO:0051169 'nuclear transport'

GO:0044744 'protein targeting to nucleus' = The process of directing proteins *towards* the nucleus, usually using signals contained within the protein. But the proteins may not reach the nucleus after all...

GO:0051169 'nuclear transport' = The directed movement of substances *into*, out of, or within the nucleus. As is, the process assumes that the substance does reach the nucleus

has_target_end_location = [Typedef] id: has_target_end_location name: has_target_end_location alt_id: results_in_transport_or_localization_to xref: RO:0002339 is_a: results_in_transport_to_from_or_in ! results_in_transport_to_from_or_in

So I've probably missed something and we've resolved that 'targeting to CC' implies that the substance does reach the target.

I'm still a bit unsure if we may want/need to make this more explicit in the definitions of targeting terms.

We discussed whether targetting is a type of transport (as current graph suggests), or whether targetting encompasses addition of the tag + transport (in which case we should change our relationships to targetting HAS_PART transport).

The Comment in 'protein targeting' suggests the latter. The ontology structure suggests the former.

AI: David H to look at 10-15 MGI 'protein targeting' annotations and see how they're being used. AI: Becky to update Comment of 'protein targeting' to make it clear to annotators that it's for annotation of the machinery that does the tagging/recognizes the tag and NOT the protein that is being tagged and NOT the protein that contains the localization sequence.

Examples where 'protein targetting' would be used: - Member of docking complex that binds to the sequence motif - Protein that myristoylates a protein, to target it to a membrane.

Suggested tweak to textual definition pattern for detection of sensory stimulus terms

Relevant ticket:

"The series of events in which a sensory X stimulus is received by a cell and converted into a molecular signal."
"The series of events in which an X stimulus is received by a cell and converted into a molecular signal as part of sensory perception of X."

I think the current version sounds quite silted. My suggestion conforms more closely to the logical definition- and reflects that it is not the nature of the stimulus that makes it sensory, but the process that the detection is part of.

One other question: is it warranted to limit these terms to the cell level? e.g. detection of humidity may involve deformation of some part of a sense organ, with a cell then detecting the resulting mechanical stimulus.

The current definitions limit these terms to the cell level, but they don't have cellular parents, so David OS will go ahead and broaden the definitions.

Split nucleotide phosphatase activity ? Split GO:0019204 ; nucleotide phosphatase activity ?

Refer to ChEBI: they have 'polynucleotide' HAS_PART 'nucleotide residue'. (nucleotide in ChEBI is reserved for FREE nucleotides. Nucleotide residue is used when it's part of a bigger structure).

DNA packaging complex

Paola to assign SF item to Jane: this is a valid term: it's for any complex that plays a role in DNA packaging'. Defining complexes by what they do is good for cross-product formation

Transmembrane vs non-transmembrane amino acid import

David Hill: Binding of an amino acid to a receptor and that amino acid entering the cell by endocytosis is an example of amino acid import which isn't transmembrane. So it's correct not to have 'amino acid import' under TM transport.