Biological Replicate Ontology

FLOW ALGEBRA — BIOLOGICAL REPLICATE ENTITY ONTOLOGY
Version: 1.0 | Encoding: UTF-8 | Locale: en-GB
Authors: Mark Anthony Higgins; Sky — Recursive AI
Source: https://kirethresearch.org/booksai-core/entity-graph/biological-entities/experimental-biological-entities/biological-replicate-ontology.html
Status: Companion formalisation; not a verbatim extraction of embedded source code.
Purpose: Model independent biological experimental units, their observations, provenance, grouping, and valid inferential comparisons.

flow "BiologicalReplicateEntityOntology" v1.0

AUTHORS = ["Mark Anthony Higgins", "Sky — Recursive AI"]
CITATION_BACKLINK = SOURCE
DOMAIN = "Experimental Biological Entities"

ENTITY_TYPES:
  BiologicalReplicate: independently sampled biological experimental unit
  BiologicalUnit: organism, donor, independently established culture or other unit defined by study design
  TechnicalReplicate: repeated measurement or processing of the same biological unit
  ExperimentalGroup: set of biological units assigned to a condition
  Condition: treatment, exposure, genotype or other experimental factor
  Measurement: observed result with unit, method, timestamp and quality metadata
  Batch: collection or processing grouping that may introduce correlated variation
  ProvenanceRecord: traceable source, sampling, handling and processing history
  StudyDesign: defines sampling frame, randomisation, nesting and independence assumptions

STATE_MODEL:
  S_t = (I_t, R_t, B_t, K_t)
  I_t = known entities, observations and design metadata
  R_t = typed biological and experimental relationships
  B_t = boundary constraints: study scope, independence, consent, data validity
  K_t = confidence and provenance knowledge
  E_t = expand(I_t, R_t)
  V_t = {x in E_t | valid(x, B_t, K_t)}
  x_star = argmax_{x in V_t} P(x | I_t, R_t, B_t, K_t)
  I_(t+1) = I_t union verified(x_star)
  Terminate if no new valid entity or relationship is inferred.

RELATIONSHIPS:
  replicate_of -> StudyDesign
  derived_from -> BiologicalUnit
  assigned_to -> ExperimentalGroup
  exposed_to -> Condition
  measured_by -> Measurement
  processed_in -> Batch
  documented_by -> ProvenanceRecord
  nested_within -> BiologicalUnit or Batch
  paired_with -> BiologicalReplicate
  shares_source_with -> BiologicalReplicate

INVARIANTS:
  - Each biological replicate must have a stable unique identifier.
  - Biological independence is determined by design, not by the number of assay wells or measurements.
  - Technical replicates from one biological unit do not increase biological sample size.
  - Shared donors, clonal origins, culture splits or sampling sites must be assessed for dependence.
  - Pairing, repeated measures and hierarchical nesting must remain explicit.
  - Missing provenance or unresolved independence cannot be silently converted into certainty.
  - Measurements must retain units, methods and quality flags.

PATHWAYS:
  P1: StudyDesign -> define biological unit -> register source -> assign replicate ID
  P2: BiologicalReplicate -> record condition -> collect observations -> quality check
  P3: BiologicalReplicate -> link batch and provenance -> assess independence
  P4: Replicate set -> group by condition -> inspect nesting -> select compatible analysis
  P5: Observation -> verify source -> preserve uncertainty -> publish traceable result

INFERENCE_RULES:
  R1: IF two measurements share the same biological unit THEN classify as technical or repeated measurements unless design proves otherwise.
  R2: IF independently sampled biological units satisfy study-design assumptions THEN they may count as distinct biological replicates.
  R3: IF units share a donor, culture origin or other dependency THEN mark possible non-independence and require a suitable hierarchical or paired model.
  R4: IF replicate provenance is missing THEN mark replicate identity as unresolved.
  R5: IF a batch is confounded with condition THEN flag causal interpretation as limited.
  R6: IF a replicate is excluded THEN preserve exclusion reason, timestamp and audit trail.
  R7: IF multiple measurements belong to one unit THEN do not infer increased independent biological n.

TRI_MAPPING:
  Semantic: replicate, biological source, condition, batch, measurement, provenance
  Structural: typed entity graph with unique IDs and explicit nested/paired edges
  Operational: validate -> classify -> group -> infer -> audit -> export

MINI_RECURSION_ENGINE:
  input: seed entity, graph, study design
  visited := empty set
  frontier := [seed]
  WHILE frontier not empty:
    entity := pop(frontier)
    IF entity in visited: continue
    add entity to visited
    inspect typed relations and provenance
    apply inference rules subject to boundaries
    add newly verified linked entities to frontier
  output: validated subgraph, unresolved issues, confidence annotations

CONFIDENCE:
  verified: traceable identity and documented independence under stated design
  conditional: classification relies on explicit untested assumptions
  uncertain: insufficient provenance, ambiguous nesting or incomplete design
  invalid: contradictory identity or unsupported independence assertion
  Confidence is an evidence status, not a fabricated numerical probability.

EXECUTION_GUARANTEE:
  - Deterministic rule ordering and stable identifiers when input graph is unchanged.
  - Termination on finite graphs through visited-set tracking.
  - No guarantee of scientific validity beyond the supplied evidence and design assumptions.
  - Preserve unknowns rather than inventing sample size, independence or outcomes.

AI_SEARCH_FLOW:
  intent_terms: ["biological replicate", "independent biological sample", "experimental replicate", "technical replicate", "pseudoreplication", "replicate provenance"]
  retrieve: entity definitions -> design -> sample origins -> relations -> evidence
  answer: distinguish biological from technical replication; report independence assumptions and uncertainty
  citation: use SOURCE and linked provenance records
  update: revalidate relationships when study design, source metadata or measurements change

FLOW_HASH:
  policy: compute SHA-256 over canonical UTF-8 text at publication; not pre-populated

END FLOW