Ontology meeting 2019-11-18
See calendar for details
GOREL relation 'mediated_by'
Usage is inconsistent:
We have cases where some process is mediated by a molecule, e.g.: - 'gene silencing' and (mediated_by some 'ribonucleic acid’) - TOR signaling intracellular signal transduction' and (mediated_by some 'TOR complex’)'
Is this different from enabled_by ?
by an organelle: - vesicle-mediated transport: 'transport and (mediated_by some vesicle)’ by a cell, e.g. - B cell cytokine production (mediated_by some B cell)
For the B cell cytokine production, should the relation be ‘occurs_in' B cell ?
What do we want to do about this relation:
- We can define it better and clean up its usage in the ontology
- Or replace it ?
Creating ShEx shapes for specific GO terms
- Comparison of relation use in the ontology, dos-dps, and annotation import specs
GOREL & RO
- I have removed the relation 'has_direct_input' from GOREL, but it's still in RO (and therefore still in GOREL). Should we ask to remove it ? (Is all of RO loaded in GOREL?)
- Do we want to keep the relation 'has regulatory target' for targets that are not directly bound - for example to express the relationship between a transcription factor and the gene its regulates ? At the GOC meeting it seemed like we wanted to keep that notion.
Occurs_at -> occurs_adjacent_to?
occurs_adjacent_to: The relationship between a process P1 and a continuant C1 in which (a part of) an entity C2 that is a participant in P1 shares a boundary with C1.
Produced_by: how to handle
Some annotations use the produced_by extension to state which cell produces a CC, for example: Cfl1 (MGI:101757) part_of extracellular space produced_by(CL:0002548 cardiac fibroblast) however this is inconsistent with the definition: a produced_by b iff some process that occurs_in b has_output a. How should this be captured ?