Table triage

[dosumis]

Organ	Graph	Unvalidated Edges	New Cell Types	Notes	Priority
Bone-Marrow	Bone-Marrow graph	Bone-Marrow unvalidated edges	Bone-Marrow new cell types	Very little of this validates. Most edges should probably be converted to has_part	Low
Brain	Brain graph	Brain unvalidated edges	Brain new cell types	Some issue with partonomy. Need to check who experts were on this. Could potentially replace with ABA IDs and directly re-use ABA partonomy. This came directly from Allen Brain Institute.	Low
Blood	Blood graph	Blood unvalidated edges	Brain new cell types	Very little of this validates. Most edges should probably be converted to has_part	Low
Eye	Eye graph	Eye unvalidated edges	Eye new cell types	Table looks reasonably high quality. Plenty to discuss and fix. Some obvious errors in table; A few issues on anatomical partonomy. Bigger issues around partonomy for cell types. Lots of new cell types.	Medium/High
Fallopian_tube	Fallopian_tube graph	Fallopian_tube unvalidated edges	Fallopian _tube new cell types
Heart	Heart graph	Heart unvalidated edges	Heart new cell types	Graph looks scary, but very many cases might be fixed by making general terms in subject position specific. Look at adding using table. Also some interesting potential improvements to Uberon (e.g. sinoatrial node)	Medium: Tackle the X part_of Y terms and revisit graph later.
Kidney	Kidney graph	Kidney unvalidated edges	Kidney new cell types	Not so much left to fix.	High
Knee	Knee graph	Knee unvalidated edges	Knee new cell types	Requires general discussion on modelling of joints. Include Meghan.	Low/Medium - revisit once we have better agreement on how to model joints.
Large_intestine	Large_intestine graph	Large_intestine unvalidated edges	Large_intestine new cell types	Graph looks terrifying. We may be able to tame it by adding lots of x part_of y terms. Add as many of these as possible & then revisit.	Low/medium.
Liver	Liver graph	Liver unvalidated edges	Liver new cell types
Lung	Lung graph	Lung unvalidated edges	Lung new cell types
Lymph_node	Lymph_node graph	Lymph_node unvalidated edges	Lymph_node new cell types	Anatomy looks OK and might be easy(ish) to fix. Cell types are mostly general, so we probably should not bother to fix.	Medium (This table is a priority for the ASCT+B project)
Lymph_vasculature	Lymph_vasculature graph	Lymph_vasculature unvalidated edges	Lymph_vasculature new cell types
Ovary	Ovary graph	Ovary unvalidated edges	Ovary new cell types
Pancreas	Pancreas graph	Pancreas unvalidated edges	Pancreas new cell types
Peripheral_nervous_system	Peripheral_nervous_system graph	Peripheral_nervous_system unvalidated edges	Peripheral_nervous_system new cell types	Huge project, multi-stage in collaboration with SPARC consortium	Low/Medium
Prostate	Prostate graph	Prostate unvalidated edges	Prostate new cell types
Skin	Skin graph	Skin unvalidated edges	Skin new cell types	Collaboration and multiple manuscripts with GE for data/ASCT+B connection	Medium/High
Small_intestine	Small_intestine graph	Small_intestine unvalidated edges	Small_intestine new cell types
Spleen	Spleen graph	Spleen unvalidated edges	Spleen new cell types
Thymus	Thymus graph	Thymus unvalidated edges	Thymus new cell types
Ureter	Ureter graph	Ureter unvalidated edges	Ureter new cell types
Urinary_bladder	Urinary_bladder graph	Urinary_bladder unvalidated edges	Urinary_bladder new cell types
Uterus	Uterus graph	Uterus unvalidated edges	Uterus new cell types
Blood_vasculature	Blood_vasculature graph	Blood_vasculature unvalidated edges	Blood_vasculature new cell types	High value, Griffin Weber Vascular CCF paper, coordinate with manuscript submission?	High

[emquardokus]

@dosumis Example represents common issue to all "invalid" edges scenario.

Brain "invalid edge" example cingulate gyrus ccf part_of limbic lobe is new information, but correct from anatomical part of relationship perspective. Currently cingulate gyrus in uberon is listed with a functional part_of limbic system relationship. Both are true, but only the latter is currently in uberon. See citation provided below as evidence for both relationships.

Vogt BA. Cingulate cortex in the three limbic subsystems. Handb Clin Neurol. 2019;166:39-51. doi: 10.1016/B978-0-444-64196-0.00003-0. PMID: 31731924.

This type of issue applies to more than just the Brain ASCT+B table and represents a more global question and opportunity for ADDING relationships to Uberon. The report interpretations should be provided with context to what "invalid" means. Highest level interpretation is "not found in uberon"--it should not be "invalid" as that suggests WRONG.

Would it be useful to provide some sort of feedback system of options with checkboxes for expert review? Options: 1) Incorrect relationship 2) Correct functional relationship/missing in uberon 3)Correct anatomical relationship/missing in Uberon?

[dosumis]

Please don't get too hung up on the word I chose here. To avoid confusion I've change to 'unvalidated' Here's the breakdown we started on in the meeting last week:

https://github.com/obophenotype/asct-b_validation/blob/main/doc/harmomization_SOP.md. (needs fleshing out with examples - please extend as needed).

Plenty of examples of this breakdown on existing triage tickets. I don't think 3 checkboxes will work to cover that diversity of issues.

In Uberon we just have part_of - not functional vs physical part_of. We can and do define terms for functional systems and these have parts (e.g. limbic system). I don't think useful to make this a category on tickets.

[dosumis]

Also - work on brain should probably be part of wider collaboration & bring in Allen Brain Atlas ontologies (which extend Uberon).