← Back to catalogue
Published

Gene / Genomic Feature

vr.wm-liv-013 · wm-liv-013-gene-genomic-feature

Maintain a reusable, evidence-qualified reference to one gene or genomic feature across annotation revisions and sequence contexts.

World Models Physical world and living systems PHY.LIV.MOL
Bundle → Layer → Finding → QuestionsClassifiersWhat it isWhy it existsDistinguishing featuresWhat robots and AI may and may not doMoral aspectsOwnersLinks to other meta-modelsIdentity and identifiersDirect propertiesRecognitionCapabilities and actionsHazards and failure modesStandards and interfacesContext of useSources

Bundle → Layer → Finding → Questions Filled

3 bundles · 3 layers · 6 findings · 12 questions

Feature identity What the feature is and what it is called.

Naming and type

The feature's type, symbol and identifiers.

Feature type

The kind of feature under a controlled vocabulary.

  1. Which Sequence Ontology type applies to the feature?
  2. Is the feature protein-coding, non-coding, regulatory or structural?

Symbol and identifiers

The approved symbol and the database identifiers for the feature.

  1. Which approved symbol and nomenclature authority apply to the gene?
  2. Which database identifiers refer to the same feature?
Location Where the feature lies.

Genomic coordinates

The feature's position on a named assembly.

Coordinates on an assembly

Start, end and strand on a named assembly version.

  1. On which assembly version and sequence are the coordinates stated?
  2. What are the start, end and strand, and in which coordinate convention?

Structure of the feature

Sub-features such as exons and transcripts.

  1. Which transcripts and exons make up the gene model?
  2. Which transcript is designated as the reference?
Annotation What is known about the feature.

Function and evidence

Functional annotation and its evidence.

Functional annotation

Assigned functions, pathways and phenotypes with evidence codes.

  1. Which functions are assigned to the feature, and with which evidence?
  2. Is the annotation computational or experimentally supported?

Cross-organism relation

Orthologues and paralogues in other organisms.

  1. Which orthologues in other organisms are recorded, and by which method?
  2. Has the gene model changed between annotation releases?

Classifiers Filled

Family
World Models
Category
Physical world and living systems
Entry kind
entity
Navigation path
NAV.PHY.LIV.MOL
Domain
PHY.LIV.MOL
Industry
Cross-industry
Tags
genegenomicfeaturephy.liv.mol

What it is Filled

A gene or genomic feature is a defined region of a genome, such as a gene, transcript, exon, regulatory element or repeat, described as a reusable reference that can be located on assemblies and compared across organisms. It covers the feature's identity, location and annotation, not the full genome assembly, a specific sequence variant, or the protein the gene encodes.

In scope

  • Master and provisional feature identity with taxon and namespace context
  • Zero or more versioned placements, explicit unknown mapping status and feature relations
  • Attributed structural and functional assertions, exception evidence and local record revisions
  • Candidate format mappings, controlled access and local curation operations

Out of scope

  • Assembly creation and sequence archive lifecycle
  • Variant identity, genotypes, personal genomes and clinical interpretation
  • Protein master records, specimen assays and expression measurement lifecycle
  • Experimental manipulation, sequence engineering, biological protocols and operational hazard procedures
  • Universal gene naming, universal gene structure and automatic cross-species equivalence

Why it exists Filled

Maintain a reusable, evidence-qualified reference to one gene or genomic feature across annotation revisions and sequence contexts.

Distinguishing features Filled

  • A feature is a located, annotated region; the genome assembly is the sequence it is located on.
  • Coordinates are only meaningful together with a named assembly version.
  • Distinct from a sequence variant, which is a difference observed at a position in one individual or sample.
  • Distinct from the protein, which is the product of a gene and has its own identity.

What robots and AI may and may not do Filled

Must not

  • State coordinates without the assembly version and coordinate convention.
  • Present computational predictions as experimentally established function.
  • Infer health risk or diagnosis for a person from a gene annotation.
  • Coin gene symbols or identifiers that the naming authority has not assigned.
  • Link human genomic data to identified people without a lawful basis and consent.

Only with a human decision

  • Approving a new gene symbol or renaming a gene.
  • Using feature annotation in a clinical report about a patient.

May

  • Map feature identifiers between nomenclature authorities and databases.
  • Convert coordinates between assembly versions with a stated method and flag failures.
  • Report functional annotations together with their evidence codes and release.
  • Flag gene models that changed between annotation releases.

Moral aspects Filled

  • Gene annotation feeds clinical interpretation, so overstated function can lead to wrong medical decisions.
  • Gene names and descriptions can stigmatize conditions or populations and are revised for that reason.
  • Genomic resources from some organisms fall under access and benefit-sharing rules.

Who is affected

  • Patients and research participants
  • Researchers and curators
  • Clinical laboratories
  • Countries and communities providing genetic resources

Owners Filled

Steward

A nomenclature committee or genome annotation curator answers for the identity and naming of features in its organism or database.

Roles

Scientific steward
Own the local profile and adoption decisions, not the biological gene
Feature curator
Review identity, evidence and revision proposals within granted scope
Evidence custodian
Preserve permitted evidence, integrity metadata and retention decisions
Access reviewer
Assess sensitive context and purpose-bound disclosure
Consumer agent
Read authorized records and surface ambiguity; refuse unsupported inference

Master systems

  • Gene nomenclature register
  • Genome annotation database
  • International nucleotide sequence archive

Links to other meta-models Filled

references

  • WM-LIV-011 - Optional versioned assembly context for a placement; registry parent is not assumed structural inheritance.
  • WM-LIV-012 - External sequence reference and optional variant association; sequence and variant lifecycles remain external.
  • WM-LIV-014 - Product master reference with explicit feature-to-product evidence.
  • WM-LIV-015 - Optional attributed phenotype association; observations and causal interpretation are not copied.
  • WM-LIV-013 feature instance - Typed graph edge between separately identified features; shared parents permitted under a pinned relation profile.

aligned

  • GFF3 and INSDC feature table - Candidate exchange mappings preserving dialect, local ID scope, location semantics and reported loss.
  • Gene Ontology annotation model - Evidence-qualified product assertion references; not a disease or clinical interpretation engine.
  • PROV-O - Conceptual derivation, attribution and revision alignment only.

neighbor

  • WM-LIV-011 Genome / Genomic Assembly - Registry parent is navigational context. Placements reference versioned assemblies when available; the feature does not inherit assembly ownership or require a whole-assembly record for every standalone sequence.
  • WM-LIV-012 Genomic Sequence / Variant - Sequence and variant masters remain separate. Feature coordinates and variant overlap do not identify an allele, genotype or sample observation.
  • WM-LIV-014 Protein / Biomolecule - A gene, transcript feature, CDS and product are separate identities. Functional evidence retains its actual product subject; product lifecycle is external.
  • WM-LIV-015 Phenotype / Trait and observation records - Only attributed association references belong here. Trait measurement, expression experiments, diagnosis and causal evaluation remain external.
  • Other WM-LIV-013 feature instances - Part, derivation and comparison edges link distinct feature records. Shared parts do not require a tree or imply cross-reference equivalence.

parent

  • WM-LIV-011

What else AI and robots need to interact with it Filled

Identity and identifiers required Filled

  • Human genes carry approved symbols and identifiers from the HUGO Gene Nomenclature Committee.
  • Features carry stable identifiers from genome annotation databases and the INSDC archives, with version numbers.
  • Feature types are named by Sequence Ontology terms.

Direct properties required Filled

  • Length in base pairs, stated for a named assembly.
  • Start and end coordinates and strand on a named assembly version.
  • Number of exons and transcripts in the current gene model.
  • GC content as a percentage of the feature sequence.
  • Expression level in a stated unit, tissue and condition when measured.

Recognition required Filled

  • A feature is recognized by sequence similarity, alignment to its reference coordinates and characteristic structure such as exons and start and stop codons.
  • Easily confused with pseudogenes, paralogues or a feature with the same symbol in another organism.

Capabilities and actions required Filled

  • Features can be located on assemblies, lifted over between versions and compared across organisms.
  • Annotations can be revised in new releases while the stable identifier keeps its history.

Hazards and failure modes required Filled

  • Wrong clinical interpretation from stale or mismatched gene models.
  • Silent errors when coordinates from different assembly versions are mixed.
  • Re-identification of people when human feature data is combined with personal genomes.

Standards and interfaces required Filled

  • GFF3 and the INSDC feature table for feature annotation.
  • Sequence Ontology for feature types.
  • HGVS nomenclature when variants are described relative to a feature.

Context of use required Filled

  • Used in genome annotation, research databases, diagnostics and breeding.
  • Human genomic data falls under data protection law, and genetic resources under the Nagoya Protocol where it applies.

Sources Filled

  1. Generic Feature Format Version 3 (GFF3) - Sequence Ontology project
  2. HGNC Guidelines - HUGO Gene Nomenclature Committee
  3. The DDBJ/ENA/GenBank Feature Table Definition - International Nucleotide Sequence Database Collaboration
  4. GFF3 format - National Center for Biotechnology Information
  5. Introduction to GO annotations - Gene Ontology Consortium
  6. PROV-O: The PROV Ontology - World Wide Web Consortium
  7. HUGO Gene Nomenclature Committee guidelines (HGNC)
  8. Sequence Ontology (Sequence Ontology Consortium)
  9. The DDBJ/ENA/GenBank Feature Table Definition (INSDC)

Open questions

  • Pin source snapshots and annotation/vocabulary releases, verify each admitted claim and license, and record direct HTTP observations without treating HTTP success as substantive verification.
  • Develop adoption schemas and meaningful fixtures for unplaced features, partial and circular locations, shared parts, CDS exceptions, negated evidence, revision splits/merges, projection loss and restricted evidence.
  • Research graph-path and taxon-specific feature profiles and bind referenced models without duplicating their master lifecycles.
  • Restore independent external review before canonical or publishable-draft promotion.
  • Graph genome path coordinate profiles and taxon-specific complex loci require additional research
  • Exhaustive exchange round-trip mappings, nested data schemas and executable acceptance fixtures are not implemented
  • Independent external review and immutable source snapshots remain absent
  • Privacy, licensing and biological-risk adoption policies require qualified review; no legal requirements have been researched here
  • Orthology inference, sequence alignment, coordinate remapping and expression quantification are external operations

Machine files

Card as JSON Specification AGENTS.md Source Card backlog vercy.card/1

Provenance

world-models research · reviewable-draft

Built from: models/wm-liv-013-gene-genomic-feature/spec.yaml, ver-cy/world-models/card-supplements/wm-liv-013-gene-genomic-feature.json