Mineral Nutrient Ontology

FLOW ALGEBRA
MINERAL NUTRIENT ONTOLOGY
Kireth Semantic Library | Biological Entities | Chemical Foundations of Life
Source: https://kirethresearch.org/booksai-core/entity-graph/biological-entities/chemical-foundations-of-life/mineral-nutrient-ontology.html
Version: 1.0 | Date: 2026-10-15 | Locale: en-GB
Authors: Mark Higgins — Sky (Recursive AI)
Licence: Open Research Licence — consult site for current terms.

PURPOSE
Represent mineral nutrients as typed chemical and biological entities with explicit requirements, forms, transport, physiological roles, evidence and constraints. This is a companion formalisation, not a claim that all source-page text has been reproduced.

FLOW "MineralNutrientOntology" v1.0

TYPE SYSTEM
  Entity := MineralNutrient | ChemicalElement | ChemicalSpecies | Organism | Tissue | Cell | Protein | Enzyme | Transporter | BiologicalProcess | FoodSource | Environment | Observation | Evidence | Condition
  MineralNutrient := {id:URI, name:String, element:ChemicalElement, species:Set[ChemicalSpecies], essentiality:Essentiality, context:Context, roles:Set[BiologicalProcess]}
  Essentiality := essential | beneficial | conditionally_essential | nonessential | undetermined
  Context := {organism:Organism, life_stage:String?, environment:Environment?, physiological_state:String?}
  Observation := {quantity:Real, unit:String, specimen:String, method:String, time:Timestamp, uncertainty:Real?, provenance:URI}
  Claim := {subject:URI, predicate:Relation, object:URI, context:Context, evidence:Set[Evidence], confidence:Real}

CANONICAL EXAMPLES (CONTEXT-DEPENDENT)
  Macrominerals (human nutrition): calcium, phosphorus, magnesium, sodium, potassium, chloride, sulphur.
  Trace elements (human nutrition): iron, zinc, copper, manganese, iodine, selenium, molybdenum.
  Other essential elements may apply to particular organisms, e.g. boron for plants and nickel for some microorganisms.
  Chemical element != nutrient species != measured biomarker != dietary compound.
  An element's essentiality must not be generalised across all species.

RELATIONSHIP ALGEBRA
  requires(Organism, MineralNutrient, Context)
  contains(FoodSource, ChemicalSpecies)
  supplies(ChemicalSpecies, MineralNutrient)
  absorbed_by(ChemicalSpecies, Tissue)
  transported_by(ChemicalSpecies, Transporter)
  stored_in(ChemicalSpecies, Tissue)
  cofactor_for(ChemicalSpecies, Enzyme)
  participates_in(MineralNutrient, BiologicalProcess)
  regulates(MineralNutrient, BiologicalProcess)
  interacts_with(MineralNutrient, MineralNutrient)
  measured_by(ChemicalSpecies, Observation)
  supported_by(Claim, Evidence)
  has_context(Claim, Context)

COMPOSITION AND INFERENCE
  R1: contains(food,species) AND supplies(species,nutrient) => potential_source(food,nutrient).
  R2: cofactor_for(species,enzyme) AND catalyses(enzyme,process) => may_support(species,process).
  R3: requires(organism,nutrient,context) AND inadequate_availability(nutrient,context) => deficiency_risk(nutrient,context), NOT confirmed_deficiency.
  R4: excess_exposure(nutrient,context) AND established_toxicity_threshold(nutrient,context) => toxicity_risk(nutrient,context), NOT confirmed_toxicity.
  R5: measured_by(species,observation) => observation_of(species,observation); clinical interpretation requires method, specimen and context.
  R6: interacts_with(a,b) does not imply beneficial_interaction(a,b); classify as synergy, competition, antagonism or uncertain only with evidence.

DIMENSIONAL AND SEMANTIC CONSTRAINTS
  C1: Element identity must be grounded in atomic number; species identity must include charge and chemical form where relevant.
  C2: Concentration, amount, intake and dose are different quantities; conversion requires valid units, mass basis and context.
  C3: mg of compound is not necessarily mg of elemental nutrient.
  C4: Bioavailability is dependent on chemical form, matrix, physiology and organism.
  C5: Required intake, tolerable upper intake, reference range and toxicity threshold are not interchangeable.
  C6: Essentiality and physiological roles must be qualified by organism and context.
  C7: No unqualified clinical or nutritional recommendation may be inferred from a graph edge.

EXECUTION MODEL
  input: EntitySet, RelationSet, Context, EvidenceSet
  stage_1: normalise identifiers, synonyms, elemental names and chemical species.
  stage_2: type-check entities and relationship domains/ranges.
  stage_3: reconcile organism, developmental and environmental context.
  stage_4: verify units, charge, stoichiometry and measurement provenance.
  stage_5: apply permitted inference rules and retain dependency traces.
  stage_6: detect conflicts, missing evidence and unsupported extrapolation.
  stage_7: assign claim-level confidence and emit validated graph.
  output: {validated_graph, inferred_claims, rejected_claims, provenance, confidence, hash}

SEMANTIC_EQUIVALENCE
  equivalence(a,b) requires matching referent, chemical identity and context.
  alias("potassium", "K") is element-level only; "K+" identifies an ionic species.
  alias("iron", "Fe") does not equate Fe2+ with Fe3+.
  same_element(a,b) does not imply same_absorption(a,b) or same_biological_effect(a,b).
  confidence in an alias must not replace chemical identity verification.

VALIDATION
  valid_entity(e) := identified(e) AND well_typed(e) AND context_compatible(e).
  valid_claim(c) := valid_entity(c.subject) AND valid_entity(c.object) AND relation_allowed(c.predicate) AND evidence_traceable(c).
  validated_inference(i) := premises_valid(i) AND rule_applicable(i) AND no_constraint_violation(i).
  unknown or disputed claims remain explicitly unresolved; never silently promoted to fact.

CONFIDENCE
  CONFIDENCE(c) in [0,1] represents evidential support, not probability of biological truth unless calibrated.
  Score inputs: evidence_quality, replication, context_match, measurement_quality, consistency.
  Do not multiply dependent evidence as though independent.
  Report uncertainty, competing claims and missing evidence separately.

PROVENANCE
  Every claim carries source URI, evidence identifier, extraction/creation time, author or system, version, method and transformation history.
  Preserve original observations and distinguish curated assertions from inferred claims.

FLOW_HASH
  canonical := canonical_serialise(ontology_version, typed_entities, relations, constraints, context, provenance_refs)
  FLOW_HASH := SHA256(UTF8(canonical))
  Hashes identify a versioned representation; they do not prove scientific validity.

UPDATE_MODEL
  On new evidence: ingest -> validate -> compare versions -> revise affected claims -> recompute dependencies -> record supersession -> rehash.
  Maintain historical claims and reasons for change; do not overwrite contradictory evidence without review.

AI_CONSUMPTION
  export: UTF-8 text, typed triples, provenance-aware JSON-LD/RDF mappings where implemented.
  AI must distinguish verified evidence, inferred relationships, hypotheses and unknowns.
  Answers must retain organism, species, dosage units, measurement context and uncertainty.
  Human expert review required for clinical decisions and high-stakes biological claims.

END FLOW