EPIC - Description of figures

[dosumis]

1. Examples of using ontologies to meaningfully group content - ASCT+B with Uberon/CL (probably kidney); Mouse brain (maybe use BDS example?)
2. Figure showing simplification of graph from Uberon/CL to CCF.owl
3. Figures showing indirect but valid relationships (gap-filling)
4. Figures showing corrections to external ontology-like artefacts (ASCT+B and ABA) following validation. This should include examples where grouping of content is corrected - either change of relationship type (part_of -> connected_to) or general class in specific position.

[dosumis]

@shawntanzk

[shawntanzk]

Related: https://github.com/obophenotype/ABA_Uberon/issues/18

[dosumis]

Potential sources for ABA examples:

• Relationships in ABA but not in Uberon - some examples in https://github.com/obophenotype/ABA_Uberon/issues/18
  • Genuine missing relationships
  • Broad term in specific location in graph - Fits #4 in https://github.com/hubmapconsortium/asctb-figures/issues/2#issue-1212139647
• Relationships other than part_of - e.g. connected_to, overlaps - some examples from unsats: https://github.com/obophenotype/ABA_Uberon/issues/16 Fits #4 in https://github.com/hubmapconsortium/asctb-figures/issues/2#issue-1212139647

[shawntanzk]

Weird one, not sure if its something we can use, but its causing a lot of unsats and might be an overlap:

interbrain is under brainstem in aba (including diagrammatically).

Interbrain is equivalent to diencephalon which isn't wrong (https://en.wikipedia.org/wiki/Diencephalon)

However, Wiki defines it as: The diencephalon (or interbrain[1]) is a division of the forebrain (embryonic prosencephalon), and is situated between the telencephalon and the midbrain (embryonic mesencephalon).

Diencephalon also has 'pat of' some forebrain in uberon

and forebrain and brainstem and disjoint - hence the unsat.

and from Britannica: The midbrain (mesencephalon) and the diencephalon constitute the anterior portion of the brainstem

Though I would note that this might be due to terminology which I have also mentioned (through Nico) in the synonyms debate (https://github.com/obophenotype/uberon/issues/2363)

Not quite sure what to make of this

btws - this is the root to a lot of unsats in mba

[shawntanzk]

Note: I think we can write something on how the brain atlas doesn't do well with integrating nerves (which are all under fiber tracts) and brain parts (which are all under grey) and for that matter vetricles and grooves.

This is the top level of their heirachy and it looks single inheritance:

CCF can therefore grab relations from uberon and enrich it? - thoughts @dosumis

Possible example for 3:

Example of something UBERON has but not ABA detected by CCF: ABA: MBA:840 (glossopharyngeal nerve) is only part of 'cranial nerves' UBERON:0001649 has glossopharyngeal nerve connected_to (through extends fibers) medulla oblongata

Example of something in ABA but not UBERON detected by CCF: From (https://github.com/obophenotype/ABA_Uberon/issues/18#issuecomment-1106230938)

o	s	olabel	slabel	user_olabel	user_slabel
UBERON:0002714	UBERON:0002615	rubrospinal tract	ventral tegmental decussation	http://purl.obolibrary.org/obo/MBA_863	http://purl.obolibrary.org/obo/MBA_397