using "located_in" for locations where the gene product is not active

[ValWood]

It seems this is allowed:
https://wiki.geneontology.org/Located_in

we just need to classify our annotations.

We could make everything located in except
i) annotations used in MF extensions -> ia_active in
ii) annotations used in GO-CAMs (although this should be a subset of i)
iii) annotations transferred by IBA, ISO, ISS should only be is_active_in

this would classify most correctly, and we could migrate the remaining slowly.

Issues, would need to enable in canto, and display on the website (and enable a filter)

[kimrutherford]

i) annotations used in MF extensions -> ia_active in
ii) annotations used in GO-CAMs (although this should be a subset of i)

Could you explain those?

[ValWood]

Need to think abiut this more. Follow up in New year

[kimrutherford]

Need to think abiut this more. Follow up in New year

Is this something to look at soon?

It seems this is allowed:
https://wiki.geneontology.org/Located_in

I don't understand that page:

What not to capture with the has_input relation

• 'Located in' is not used for ...
• 'Located in' should not be ...

Are they talking about has_input or "located in"?

[PCarme]

Are they talking about has_input or "located in"?

I'm pretty sure the "has_input" in the title is an error. The 2 examples of what shouldn't be captured definitely relate to the 'located_in' relationship (although, they should not be captured with 'has_input' either).
The same error is also present in the previous headline "What to capture with the has_input relation"

[kimrutherford]

Thanks Pascal. I'm less confused now. :-)

[ValWood]

I sent this to GO

No particular hurry for this. We can discuss when other stuff is done.