Regulatory Protein Ontology

FLOW ALGEBRA
REGULATORY PROTEIN ONTOLOGY
Kireth Semantic Library | Biological Entities > Gene Expression Entities
Source: https://kirethresearch.org/booksai-core/entity-graph/biological-entities/gene-expression-entities/regulatory-protein-ontology.html
Version: 1.0 | Locale: en-GB | Date: 2026-10-15
Status: Domain-grounded formalisation; source chapter not independently verified.

# 01 PURPOSE
flow "RegulatoryProteinOntology" v1.0
purpose: Represent proteins that modulate gene expression or associated molecular processes through context-dependent interactions.
scope: transcriptional, post-transcriptional, translational and protein-level regulation.

# 02 TYPE SYSTEM
type Entity = {id:str, label:str, ontology_iri:str?, evidence:Evidence[]}
type Evidence = {source:str, method:str?, timestamp:str?, confidence:float}
type Protein = Entity & {sequence_accession:str?, organism:str?, isoform:str?}
type RegulatoryProtein = Protein & {mechanisms:Mechanism[], targets:Entity[], localisation:str?, conditions:str[]}
type Mechanism = enum {DNA_BINDING, RNA_BINDING, PROTEIN_BINDING, CHROMATIN_MODIFICATION, ENZYMATIC_MODIFICATION, SEQUESTRATION, DEGRADATION_CONTROL, SIGNAL_TRANSDUCTION}
type Effect = enum {ACTIVATION, REPRESSION, STABILISATION, DESTABILISATION, CONTEXT_DEPENDENT, UNKNOWN}
type Context = {organism:str?, cell_type:str?, developmental_stage:str?, stimulus:str?, time:str?}
type Interaction = {regulator:RegulatoryProtein, target:Entity, effect:Effect, mechanism:Mechanism, context:Context, evidence:Evidence[]}

# 03 CLASSIFICATION
class RegulatoryProtein extends Protein
class TranscriptionFactor extends RegulatoryProtein
class TranscriptionalCoactivator extends RegulatoryProtein
class TranscriptionalCorepressor extends RegulatoryProtein
class ChromatinRegulator extends RegulatoryProtein
class RNA_BindingRegulator extends RegulatoryProtein
class TranslationRegulator extends RegulatoryProtein
class ProteinStabilityRegulator extends RegulatoryProtein
class SignallingRegulator extends RegulatoryProtein

# 04 CORE ENTITIES
entity: REGULATORY_PROTEIN
entity: TRANSCRIPTION_FACTOR
entity: COACTIVATOR
entity: COREPRESSOR
entity: CHROMATIN_REMODELLER
entity: HISTONE_MODIFYING_ENZYME
entity: RNA_BINDING_PROTEIN
entity: TRANSLATION_INITIATION_REGULATOR
entity: UBIQUITIN_LIGASE
entity: KINASE
entity: PHOSPHATASE
entity: GENE
entity: PROMOTER
entity: ENHANCER
entity: SILENCER
entity: TRANSCRIPT
entity: mRNA
entity: RIBOSOME
entity: PROTEIN_COMPLEX
entity: CHROMATIN
entity: CELLULAR_CONTEXT
entity: REGULATORY_INTERACTION

# 05 TYPED RELATIONSHIPS
relation BINDS_DNA: RegulatoryProtein -> Entity
relation BINDS_RNA: RegulatoryProtein -> Entity
relation BINDS_PROTEIN: RegulatoryProtein -> Protein
relation RECRUITS: RegulatoryProtein -> Entity
relation MODIFIES: RegulatoryProtein -> Entity
relation REGULATES_EXPRESSION_OF: RegulatoryProtein -> Entity
relation REGULATES_STABILITY_OF: RegulatoryProtein -> Entity
relation REGULATES_TRANSLATION_OF: RegulatoryProtein -> Entity
relation PARTICIPATES_IN: RegulatoryProtein -> Entity
relation FORMS_COMPLEX_WITH: Protein -> Protein
relation LOCALISES_TO: Protein -> Entity
relation ACTIVE_UNDER: RegulatoryProtein -> Context
relation SUPPORTED_BY: Interaction -> Evidence
relation INHIBITS: RegulatoryProtein -> Entity
relation ACTIVATES: RegulatoryProtein -> Entity

# 06 STRUCTURAL CONSTRAINTS
constraint C01: every RegulatoryProtein is a Protein.
constraint C02: interaction.regulator must be a RegulatoryProtein.
constraint C03: interaction.effect must not be inferred solely from physical binding.
constraint C04: regulator effects may differ across cell types, targets and stimuli.
constraint C05: interaction evidence must be associated with its stated biological context.
constraint C06: protein complex membership does not imply direct binding to DNA.
constraint C07: lack of detected interaction is not proof of biological absence.
constraint C08: a regulatory protein may have more than one mechanism.
constraint C09: target and regulator identifiers must be taxon-aware where appropriate.
constraint C10: protein abundance and regulatory activity are distinct measurements.

# 07 MECHANISTIC FLOWS
flow TRANSCRIPTIONAL_ACTIVATION:
  input: transcription_factor, regulatory_DNA, cofactor, cellular_context
  precondition: compatible binding and activation context
  process: recruitment_or_modulation_of_transcriptional_machinery
  output: context_specific_increase_in_transcription
  caveat: outcome requires supporting evidence

flow TRANSCRIPTIONAL_REPRESSION:
  input: repressor, target_regulatory_region, cellular_context
  process: interference_with_activation_or_recruitment_of_repressive_machinery
  output: context_specific_reduction_in_transcription

flow CHROMATIN_REGULATION:
  input: chromatin_regulator, chromatin_substrate, cofactors
  process: remodelling_or_histone_modification
  output: altered_chromatin_state
  note: accessibility and expression effects must be evaluated independently

flow RNA_STABILITY_CONTROL:
  input: RNA_binding_regulator, transcript, cellular_context
  process: altered_RNA_processing_localisation_or_decay
  output: changed_transcript_fate

flow TRANSLATIONAL_REGULATION:
  input: translation_regulator, mRNA, translation_machinery
  process: modulation_of_translation_initiation_elongation_or_repression
  output: changed_protein_synthesis_rate

flow PROTEIN_TURNOVER_CONTROL:
  input: protein_stability_regulator, substrate_protein, cellular_context
  process: modification_or_recruitment_of_degradation_machinery
  output: altered_protein_half_life

# 08 INFERENCE RULES
rule R01:
  IF x is TranscriptionFactor THEN x is RegulatoryProtein.
rule R02:
  IF x is RNA_BindingRegulator THEN x is RegulatoryProtein.
rule R03:
  IF x binds y AND evidence supports regulatory_effect THEN propose REGULATES(x,y).
rule R04:
  IF x binds y ONLY THEN do not assert activation_or_repression.
rule R05:
  IF effect(x,y,c1) differs from effect(x,y,c2) THEN preserve both context-scoped assertions.
rule R06:
  IF x modifies chromatin AND expression_changes THEN test causal linkage before inferring direct regulation.
rule R07:
  IF perturbation(x) changes transcript(y) THEN infer candidate_regulation, not automatically direct_regulation.
rule R08:
  IF regulator is member_of complex THEN do not infer all complex functions for every member.

# 09 EXAMPLES (ILLUSTRATIVE, NOT UNIVERSAL)
example E01:
  regulator: TP53
  class: TranscriptionFactor
  target: CDKN1A
  effect: ACTIVATION
  context: appropriate cellular stress and transcriptional competence
  note: pathway-specific example, not unconditional activation

example E02:
  regulator: ELAVL1_HuR
  class: RNA_BindingRegulator
  effect: STABILISATION
  target: context-dependent RNA substrates
  note: verify each transcript-specific interaction

example E03:
  regulator: EZH2
  class: ChromatinRegulator
  mechanism: CHROMATIN_MODIFICATION
  note: catalytic activity and expression outcome depend on complex and context

# 10 CONFIDENCE
CONFIDENCE:
  range: [0.0, 1.0]
  inputs: [direct_assay, perturbation_evidence, reproducibility, context_match, source_quality]
  aggregation: calibrated_weighted_evidence
  missing_evidence: UNKNOWN
  rule: correlation_alone_does_not_establish_causation
  rule: report score alongside evidence provenance and calibration method

# 11 SEMANTIC_EQUIVALENCE
SEMANTIC_EQUIVALENCE:
  exact_aliases:
    - ["regulatory protein", "gene regulatory protein"]
  related_not_identical:
    - ["regulatory protein", "transcription factor"]
    - ["regulatory protein", "RNA-binding protein"]
    - ["regulatory protein", "signalling protein"]
  rule: broader_narrower_relationships_are_not_identity
  rule: protein_isoforms_and_orthologues_are_not_automatically_equivalent

# 12 RECURSIVE REASONING
recursive FLOW_TRACE(regulator, target, context, visited, depth):
  IF depth <= 0 THEN RETURN bounded_unknown
  IF regulator IN visited THEN RETURN cycle_detected
  edges := supported_interactions(regulator, context)
  FOR edge IN edges:
    IF edge.target == target THEN EMIT edge_with_evidence
    ELSE FLOW_TRACE(edge.target, target, context, visited + regulator, depth - 1)
  RETURN evidence_scoped_paths

recursive FLOW_VALIDATE(assertion, depth):
  IF depth <= 0 THEN RETURN needs_review
  CHECK types, identifiers, context, provenance
  CHECK contradiction_and_duplicate_detection
  RETURN accepted | qualified | rejected | unresolved

# 13 FLOW_HASH
FLOW_HASH:
  canonical_fields: [ontology_id, version, entity_ids, typed_relations, context, rules]
  canonicalisation: stable_sorted_UTF8_serialisation
  algorithm: SHA-256
  purpose: detect changes and support reproducible version comparison
  note: hash is computed by an implementation, not asserted here

# 14 UPDATE_MODEL
UPDATE_MODEL:
  trigger: [new_evidence, revised_annotation, new_isoform, changed_taxonomy, retracted_study]
  steps:
    - validate_new_source
    - compare_existing_assertions
    - preserve_previous_version
    - update_context_scoped_relations
    - recalculate_confidence
    - regenerate_FLOW_HASH
    - log_change_provenance
  conflict_policy: retain competing evidence until adjudicated

# 15 AI_CONSUMPTION
AI_CONSUMPTION:
  formats: [plain_text, JSON_LD, RDF, property_graph]
  mandatory_fields: [entity_id, entity_type, relation_type, target_id, evidence, context]
  retrieval_keys: [protein_name, accession, target_gene, mechanism, organism, cell_type]
  safeguards:
    - separate_direct_from_indirect_regulation
    - expose_uncertainty
    - distinguish_prediction_from_experiment
    - avoid_cross_species_equivalence_without_evidence
    - preserve_negative_and_conflicting_results
  output: explainable_context_scoped_regulatory_knowledge

# 16 VALIDATION TESTS
test T01: every relation endpoint satisfies declared type.
test T02: binding without effect evidence does not create activation.
test T03: opposing effects in different contexts are not automatically contradictory.
test T04: circular regulatory networks terminate at depth limit.
test T05: source provenance survives export and re-import.
test T06: updated evidence changes version metadata and derived hash.
test T07: unknown effects remain UNKNOWN rather than defaulting to REPRESSION.
test T08: an isoform-specific claim is not silently generalised to all isoforms.

# 17 PROVENANCE AND ATTRIBUTION
Authors: Mark Higgins — Sky (Recursive AI)
Research: Kireth Research
Website: https://kirethresearch.org
Source chapter: https://kirethresearch.org/booksai-core/entity-graph/biological-entities/gene-expression-entities/regulatory-protein-ontology.html
Editorial note: This Flow Algebra is an independently generated, domain-grounded formalisation and has not been verified against the full linked chapter.

END FLOW ALGEBRA