Abnormal terms in component

[gocentral]

This came out of a discussion we had on a PRO curation call concerning protein aggregates. There are at least two terms that were agreed to not be normal occurrences GO:0097481 neurofibrillary tangle and GO:0097413 Lewy body (and childrenb)

Each has in their def a sentence to the effect that they contain abnormally phosphorylated proteins Others to perhaps remove:

GO:0097419 Pick body GO:0097416 Lewy body-like hyaline inclusions

Perhaps these should be removed

Reported by: hdrabkin

Original Ticket: geneontology/ontology-requests/10612

[gocentral]

See http://www.jbiomedsem.com/content/4/1/20/abstract

Original comment by: cmungall

[gocentral]

Chris are you saying that Lewy body should be kept as thought went behind it's inclusion? Ruth

Original comment by: RLovering

[gocentral]

Hi,

Including those CC terms in GO was not done without debating it first. I gave a brief summary of the thoughts behind that at the GOC meeting in Bar Harbor but I'm happy to mention it again here as the PRO people were probably not aware of it:

• in the process of integrating terms in GO from the SAO (Subcellular Anatomy Ontology), a process detailed in the paper that Chris links to, we ran into the issue of several 'abnormal' CC components
• are these really and always 'abnormal'? Because of the way they are usually found (in post-mortem sections of people with neurological diseases), we tend to think so. But there is a good likelihood that several of those inclusions are normally present in the brains of aging people - we may not know for sure because post-mortem searches for them are not frequently done in 'healthy' elderly people. I don't have references handy but I'm quite sure of this likelyhood, at least in some measure, at least for some inclusions. So those CC terms may not always and necessarily be detected in pathological conditions. Aging is an ok process in GO.
• processes that involve/act upon those inclusions may well be physiological (removal of them, proteolytic degradation...). This is or should be object of study of research on neurodegeneration. To annotate such processes, we'd need the GO terms to link to the CC terms, in our modern way of expressing things in GO. Therefore we concluded it's ok to have those CC terms in GO. I've done my best to make them as general as possible in their definitions, but I couldn't ignore a bit of references to pathological situations. I'm happy to edit them further if people can suggest changes.
• Note that with the merging-in of SAO into GO, at present there is no other ontology than GO as far as I'm aware that could do a proper job of hosting those CC terms and allowing us to link to them.
• A point we make in the paper is that GO terms are/can be used for annotation of images as well as papers. We'd deprive the scientific community of this potentiality if we were too strict on allowing not-always-physiological components. Note that I've not included terms from SAO if there was any ground for them to be experimental artefacts only. (Or at least I've tried my best - Lewy bodies are most definitely real...)

All of this was discussed with authors of the JBMS paper that Chris quotes, and with Ruth in relation to her neurodegeneration/aging-related work (present and future). I hope this summary helps clarify it all, otherwise I'm happy to discuss it further.

Thanks for reading :-)

Paola

Original comment by: paolaroncaglia

[gocentral]

Harold agrees that this ticket can be closed.

Original comment by: paolaroncaglia

[gocentral]

• labels: --> Other term-related request, neurobiology
• status: open --> closed-works-for-me
• assigned_to: Paola Roncaglia

Original comment by: paolaroncaglia