jamesaoverton
commented
5 years ago The meeting notes are here: https://docs.google.com/document/d/1P8uMYDqWxJ4Qh6QKD6ZeYqziNTbxsMHuN-s8jn7soEk
We renamed 'cell component' to 'subcellular structure'. My understanding is captured in this paragraph from the notes:
Darren: … having the ability to ‘house’ things - things can be there that aren’t an integral part of that component. There are things inside the nucleus that aren’t necessarily part of the nucleus itself. Chromosomes and ribosomes would be ‘subcellular structures’ but other complexes would not be. Protein complexes would not be because there is no way for something that is not part of the complex to be there. A channel could be a ‘subcellular structure’ because something can be in it that is not an integral part. Darren and Chris M. will work on this location-based definition.
cmungall
We discussed this at the F2F workshop but I don't see any issue created.
I don't have the meeting notes handy, but the general idea was to have biological material entities divided into disjoint chunks at different levels of granularity
atom < molecular entity < macromolecular complex < subcellular compartment < cell < gross anatomical entityIdeally core classes would be equivalent to existing OBO roots, but the challenge here is GO cellular component doesn't fit into the above scheme:
(hence the name 'subcellular compartment' rather than 'cell component' for the core class, since we want to avoid confusion with the GO class)
Note that these decisions may be more driven by annotation use case (lumping anything that can be used in a localization annotation into one ontology) rather than design. Within GO it may actually quite useful for us to have a 'subcellular compartment' along the lines of the core one: https://github.com/geneontology/go-ontology/issues/17696