Molecular Probe Ontology

FLOW ALGEBRA — MOLECULAR PROBE ENTITY ONTOLOGY
Version: 1.0 | Date: 2026-10-28 | Encoding: UTF-8
Authors: Mark Anthony Higgins; Sky — Recursive AI
Source: https://kirethresearch.org/booksai-core/entity-graph/biological-entities/experimental-biological-entities/molecular-probe-ontology.html
Status: Companion formalisation based on the page title and established molecular-probe concepts; not a transcription of inaccessible embedded code.
Citation: Higgins, M. A. & Sky — Recursive AI, Molecular Probe Ontology, Kireth Research.
Purpose: Represent molecular probes, targets, recognition, signals, evidence, and uncertainty as typed, provenance-aware graph transitions.

1. ENTITY TYPES
MolecularProbe(p): an agent used to detect, label, localise, or quantify a molecular target.
Target(t): molecular species, sequence, epitope, or structural feature under investigation.
Sample(s): material in which target presence or abundance is assessed.
RecognitionElement(r): probe component responsible for target association.
Reporter(q): component or property producing an observable response.
Assay(a): observation procedure linking probe, sample and readout.
Signal(z): measured observation, including background and instrument effects.
Evidence(e): record of observation, calibration, controls and provenance.

Subtypes: NucleicAcidProbe, AntibodyProbe, AptamerProbe, SmallMoleculeProbe,
FluorescentProbe, RadioLabelledProbe, EnzymeLinkedProbe, AffinityProbe.
Subtype axes are non-exclusive: recognition chemistry and reporter modality differ.

2. STATE MODEL
S_t = (I_t, R_t, B_t, K_t)
I_t = known entities, identities, measurements and metadata.
R_t = typed biological and assay relationships.
B_t = admissibility bounds: sample identity, controls, calibration, assay scope.
K_t = evidence, confidence, limitations and provenance.
E_t = expand(I_t, R_t)
V_t = {x in E_t | satisfies(x, B_t)}
x*_t = argmax_{x in V_t} P(x | I_t, R_t, B_t, K_t)
I_(t+1) = update(I_t, x*_t, observed_evidence)
No update may turn a predicted association into a verified observation without evidence.

3. RELATIONSHIPS
probe HAS_RECOGNITION_ELEMENT recognition_element
probe HAS_REPORTER reporter
probe INTENDED_TO_DETECT target
probe USED_IN assay
assay ANALYSES sample
sample MAY_CONTAIN target
probe MAY_BIND target
binding_event MAY_GENERATE signal
signal RECORDED_AS evidence
evidence SUPPORTS_OR_CONTRADICTS hypothesis
probe HAS_SELECTIVITY_PROFILE target_set
assay HAS_CONTROL control
assay HAS_CALIBRATION calibration
sample HAS_PROVENANCE provenance_record

4. CONCEPTUAL TRANSITIONS
Uncharacterised -> Characterised: identity and modality documented.
Characterised -> AssayEligible: compatibility and controls documented.
AssayEligible -> Observed: valid signal and metadata recorded.
Observed -> Interpreted: background, controls and uncertainty assessed.
Interpreted -> Corroborated: independent evidence agrees within stated limits.
Any -> Indeterminate: missing controls, contamination, saturation or ambiguity.
These are logical evidence states, not laboratory instructions.

5. OBSERVATION MODEL
z_obs = g(target, probe, matrix, instrument, conditions) + background + noise
z_corrected = z_obs - estimated_background (when justified)
Detection != identification; signal != target abundance without calibration.
A positive readout can reflect cross-reactivity, interference or artefacts.
A negative readout cannot establish absence below a method's detection capability.

6. INFERENCE RULES
R1: IF probe has documented recognition element AND target compatibility,
    THEN propose possible recognition, status=PREDICTED.
R2: IF validated signal exceeds an assay-specific decision criterion AND
    controls pass, THEN infer target-associated evidence, status=SUPPORTED.
R3: IF controls fail OR sample provenance is unknown,
    THEN mark target inference INDETERMINATE.
R4: IF off-target association is plausible,
    THEN reduce specificity confidence; retain alternatives.
R5: IF independent orthogonal evidence agrees,
    THEN increase corroboration confidence without declaring certainty.
R6: IF reporter signal saturates,
    THEN prohibit unqualified quantitative interpretation.
R7: IF multiple probes share a target,
    THEN preserve separate measurement and error provenance.

7. RECURSIVE PATHWAY
Input query -> identify probe and target -> retrieve assay context ->
expand candidate relations -> apply validity bounds -> inspect controls ->
rank evidence-supported interpretations -> update graph ->
repeat only if new independent evidence exists.
Stop if no admissible new evidence, contradictory controls, or confidence plateau.

8. TRI-MAPPING
Entity graph: probe -> recognition element -> target -> sample.
Process graph: selection -> observation -> validation -> interpretation.
Evidence graph: signal -> controls -> calibration -> provenance -> confidence.
A valid conclusion requires consistency across all three mappings.

9. CONFIDENCE MODEL
Confidence is conditional on target, matrix, assay and controls.
Record analytical sensitivity, specificity, reproducibility and uncertainty
when measured; never invent numeric scores.
States: UNKNOWN, PREDICTED, SUPPORTED, CORROBORATED, CONTRADICTED.
Confidence cannot exceed the strength of its supporting evidence.

10. EXECUTION GUARANTEES (CONDITIONAL)
G1: Every asserted measurement retains sample and assay provenance.
G2: A probe's intended target is not automatically its verified target.
G3: Biological and technical replicates are not interchangeable.
G4: Unsupported transitions are rejected or labelled as hypotheses.
G5: Contradictory evidence remains visible; no silent overwrite.
G6: Outputs distinguish presence, identity, localisation and quantity.

11. AI SEARCH FLOW
flow: MolecularProbeOntology
version: 1.0
intent: resolve molecular probe entities and evidence-backed associations
input: {probe?, target?, sample?, assay?, observation?, context?}
retrieve: [entity_graph, relationship_graph, assay_evidence, provenance]
expand: [probe_types, targets, reporters, controls, observations]
filter: [identity_consistency, evidence_quality, scope, safety]
rank: evidence_strength_with_explicit_uncertainty
output: {entities, relations, pathways, claims, limitations, citations}
update_policy: append_provenance_preserving_evidence; never overwrite history
semantic_equivalence: contextual_only
hash_policy: hash canonical UTF-8 content at publication; no invented hash

12. HOW TO READ
The ontology separates what a molecular probe is designed to recognise
from what an assay actually demonstrates. Its state transitions are logical
constraints for AI reasoning, not protocols for conducting experiments.

END FLOW