For Dre: 9 V_gamma fields need ontology descriptors (placeholder: nodes to resolve)

[stevevanhooser]

Context

Branch claude/select-ontology-documents-hT22V annotated several V_gamma document types with _ontology field-concept descriptors. Nine field annotations could not be mapped to a confident, verified term in the currently-registered ontology set and were left with placeholder: _node values so consumer tooling will warn (the placeholder prefix is _approximate: true in the registry).

This issue tracks resolving those nine. Scope reminder: _ontology describes what the field itself means (the concept the field denotes), not what its value is — see V_gamma_SPEC.md "Ontology Annotation Object".

Currently registered prefixes that may help

After this branch, the CURIE registry includes: iao, uberon, pato, emapa, ncit, schema, allen_ccf_v3, stato, obi, empty (Waltham Empty Ontology — Experimental Measurements, Purposes, and Treatments), plus the placeholder: prefix for the unresolved entries below.

The 9 fields

neuron_extracellular.json

1. quality — current: placeholder:spike_sorting_quality_label ("good" / "mua" / "noise")

   • Concept: the categorical quality assessment a sorter assigns to a putative unit.
   • Search: NIFSTD (spike-sorting / unit-quality branch), EDAM (data-quality concepts), empty (would be a natural home), SIO.

2. num_spikes — current: placeholder:action_potential_count

   • Concept: count of action potentials assigned to one unit.
   • Search: GO (GO:0001508 "action potential" — biological process), NIFSTD, empty, PATO (count/amount qualities), NBO (Neuro Behavior Ontology). May need composition (count-of × action-potential).

stimulus_presentation.json

3. presentation_order — current: placeholder:stimulus_presentation_order

   • Concept: the temporal order in which a trial sequence was presented.
   • Search: OBI (study-design / trial concepts), NCIT (Sequence / Order), STATO, empty, EDAM.

4. num_trials — current: placeholder:trial_count

   • Concept: count of stimulus trials in the presentation.
   • Search: OBI (has investigation-trial concepts; OBI may have a "number of trials" or "replicate count"), STATO (sample-size / replicate metrics), NCIT, empty.

orientation_direction_tuning.json

5. preferred_direction — current: placeholder:preferred_direction_of_motion (degrees, 0–360)

   • Concept: the motion direction eliciting the maximum response in a tuning curve.
   • Search: NIFSTD (visual physiology), CogPO, empty (strong candidate — exactly the experimental-measurement domain), NBO, PATO (direction qualities).

6. preferred_orientation — current: placeholder:preferred_orientation (degrees, 0–180)

   • Concept: the visual orientation eliciting the maximum response.
   • Search: same set as preferred_direction. Note PATO PATO:0000133 "orientation" is a generic morphological orientation, not preferred-orientation-of-tuning — probably wrong fit but worth confirming.

7. direction_selectivity_index — current: placeholder:direction_selectivity_index (0–1)

   • Concept: derived metric quantifying directional selectivity from a tuning curve (DSI).
   • Search: NIFSTD, STATO (derived-statistic concepts), empty. May need to be minted in empty if no ontology has it.

8. orientation_selectivity_index — current: placeholder:orientation_selectivity_index (0–1)

   • Concept: derived metric quantifying orientation selectivity from a tuning curve (OSI).
   • Search: same set as DSI.

9. tuning_width — current: placeholder:tuning_curve_half_width_at_half_height (degrees)

   • Concept: half-width-at-half-height of a tuning curve.
   • Search: STATO (distribution-width / FWHM concepts), PATO PATO:0000921 "width" (probably too generic), NIFSTD, empty.

Suggested resolution flow

1. For each field, search the candidate ontologies (the EBI OLS browser is the easiest cross-ontology search).
2. If a verified term is found, replace the _node and _name.
3. If no real-world term exists, mint one in empty (this is the natural home for fields 5–9 in particular — they're neuroscience-experimental measurements that aren't covered well elsewhere) and replace the placeholder: reference with the new empty: term.
4. If the concept is genuinely too compositional to ontologize as a single class (e.g., "count of action potentials"), document the decision in the field's _documentation and pick the closest parent concept.

Files to edit

• schemas/V_gamma/neuron_extracellular.json
• schemas/V_gamma/stimulus_presentation.json
• schemas/V_gamma/orientation_direction_tuning.json

Cross-references

• Branch: claude/select-ontology-documents-hT22V
• V_gamma SPEC: schemas/V_gamma_SPEC.md — see "Ontology Annotation Object" and "CURIE Registry" sections.
• CURIE registry: schemas/V_gamma/CURIE_lookups_meta.json