eduKateSG · Why Science?
Bridge anatomy across species without flattening the biological differences that make comparison meaningful
Choose the closest route, then return to the full index whenever you need the wider evidence chain.
Science learning becomes powerful when students can tell a shared identifier from an observation, a logical model from biological reality, and a useful comparison from a finished conclusion. Uberon is an integrated cross-species anatomy ontology covering animal anatomical structures. Its official documentation describes body parts, organs and tissues across species, with a focus on vertebrates, and the OBO Foundry lists its role in bridging species-specific anatomy ontologies. Uberon supports applications such as Bgee cross-species gene-expression comparison and sample annotation in ENCODE and FANTOM5. A shared class can connect comparable structures, but it does not mean every structure has identical development, function or evolutionary history. This guide supports Primary Science, PSLE Science, Secondary Science, O-Level Science, O-Level Biology and STEM education as a longform evidence-reading route; it is not medical advice, a laboratory protocol, an admission promise or a career guarantee.
Related eduKate reading: NCBI Taxonomy and stable organism identifiers; Gene Ontology and traceable gene-function annotations; Evidence and Conclusion Ontology; OBI and reproducible investigations; Science Learning Hub; Education Hub; STEM Education; Careers by Subject Capability.
Primary and official sources checked on 13 October 2026: OBO Foundry Uberon overview, checked 13 October 2026; Official Uberon documentation and release guide; Official Uberon source repository; Canonical Uberon OWL release; 2026 Singapore–Cambridge O-Level Biology syllabus; MOE G2/G3 Lower Secondary Science teaching and learning syllabus; 2026 MOE G2 Computing syllabus. Ontologies, definitions and release products can change, so preserve dated documentation, exact source records and file versions.
A deeper reading route
Anatomy is not one flat list. A structure can be part of an organ, develop from an earlier structure, overlap another region and occur only in particular taxa or life stages. Uberon represents these relations so data can be queried beyond local names.
The OBO Foundry describes Uberon as an integrated cross-species anatomy ontology covering animals. This makes it a bridge across organism-specific vocabularies, not a claim that all species share identical bodies.
Part-of relations create a partonomy. A query for an organ can retrieve data annotated to its components when the relation and query semantics justify expansion. Care is needed because spatial containment, structural part and developmental origin are different.
Developmental relations connect precursors with later structures. They help compare stages and organisms, but the assertion is a curated biological model rather than a direct observation in every dataset.
Taxon constraints protect against impossible annotations. A class may be broadly applicable, restricted to a clade or absent in a species. Those constraints depend on current biological knowledge and the taxonomy release.
Bridge files link Uberon terms with foreign, often species-specific, anatomy ontologies. The official documentation explains that each bridge contains axioms establishing relationships between Uberon and those terms.
A bridge can express equivalence under a taxon restriction rather than universal identity. That distinction matters when a local ontology describes a zebrafish, mouse or other organism-specific structure more precisely.
Uberon offers several release products, including canonical, base, basic, collected and composite variants. The exact product affects imports, redundancy and reasoning, so a publication must name the file actually used.
Cross-species expression databases can use Uberon to align tissue contexts. The comparison remains conditional on sampling, developmental stage, cell composition, assay and mapping quality.
The useful goal is translation with memory: connect related anatomical concepts while retaining the original term, species, stage, evidence and mapping rule for every comparative result.
Use the narrowest class supported by the specimen record. A generic label such as “brain” should not be silently upgraded to a named subregion.
Keep laterality, segment and developmental qualifiers explicit. Left and right, anterior and posterior, embryonic and adult structures may answer different questions.
Do not confuse homology with simple similarity. Comparable position or function does not by itself establish common evolutionary origin.
Preserve the source ontology term beside the Uberon mapping. Species specialists may need distinctions that a cross-species class intentionally groups.
Test relation-aware queries. Retrieving parts, precursors or overlapping regions uses different semantics and should be validated with known examples.
Treat anatomical boundaries as modelled. Dissection, imaging resolution and annotation practice can change which tissue a sample actually contains.
Version every ontology and bridge together. A canonical Uberon release paired with an undated bridge can produce unreproducible classifications.
Use uncertainty fields instead of forcing ambiguous specimens into one term. A broader parent plus a note is often the honest answer.
The optimistic lesson is that shared anatomy makes comparative biology easier while leaving room for species-specific expertise.
Anatomical queries should declare their closure rule. A search limited to exact Uberon IDs answers a narrower question than one expanded through every part-of descendant, and both can be valid when their scope is visible.
Specimen annotations benefit from two viewpoints: what the collector believed was sampled and what later quality control showed was present. Keeping both avoids rewriting field history after molecular results arrive.
Cross-species comparisons often require a common parent that is broader than either source term. The loss of detail should be quantified and shown, not hidden behind one harmonised label.
Developmental timing can reverse an apparent mapping. Structures that share an adult name may emerge through different intermediate stages, while transient precursors may have no direct adult counterpart.
Bridge quality is testable. Curators can sample mappings, inspect definitions and taxon restrictions, run reasoners, and ask organism specialists whether retrieved comparisons match biological expectations.
Release migration deserves its own table. For every changed identifier, record whether the class was renamed, replaced, split, merged or made obsolete, and whether analytical conclusions changed.
Spatial transcriptomics makes anatomical boundaries especially important because pixels or spots may contain mixtures that cross compartments. A precise ontology term cannot remove that measurement blur.
A responsible figure caption names the mapped structure, species, stage and release, then states whether the displayed grouping used exact terms, parts, precursors or bridge expansion.
Comparative anatomy becomes most powerful when shared structure and genuine difference are both searchable. Uberon provides the graph; specimen evidence and evolutionary reasoning provide the scientific judgment.
Before pooling anatomy-linked measurements, build a comparison matrix that lists the source term, mapped Uberon term, species, life stage, sex when relevant, sampled compartment, assay and mapping confidence. Review the matrix for structural mismatches and missing stages. This simple table prevents a broad shared parent from making biologically different specimens look interchangeable and gives readers a visible reason for every inclusion.
| Stage | Question | Evidence | Responsible output |
|---|---|---|---|
| Identify | Which anatomical structure is intended? | Uberon ID, definition and synonyms | A stable anatomy class |
| Constrain | In which species and life stage can it occur? | Taxon and developmental constraints | A biologically possible context |
| Relate | How does it connect to parts and precursors? | Partonomy and developmental relations | A queryable anatomical graph |
| Compare | What cross-species statement is justified? | Bridge axioms, evidence and original ontologies | A bounded comparison rather than false equivalence |
Inside this guide
1–12 · Foundations and description
- 1. Start with the comparison
- 2. Use the Uberon identifier
- 3. Read the definition
- 4. Inspect the is-a hierarchy
- 5. Traverse part-of relations
- 6. Trace develops-from relations
- 7. Record life stage
- 8. Record taxon
- 9. Check taxon constraints
- 10. Preserve laterality
- 11. Preserve regional position
- 12. Keep source terminology
13–24 · Relations, mappings and validation
- 13. Understand bridge axioms
- 14. Use taxon-restricted equivalence
- 15. Distinguish collected products
- 16. Distinguish composite products
- 17. Choose the canonical release carefully
- 18. Validate imports
- 19. Query cross-species expression
- 20. Annotate biosamples
- 21. Challenge dissection precision
- 22. Challenge imaging resolution
- 23. Challenge one-to-many mappings
- 24. Challenge many-to-one mappings
25–36 · Learning, interpretation and pathways
- 25. Do not infer homology automatically
- 26. Audit relation-aware expansion
- 27. Protect human metadata
- 28. Connect to Primary Science
- 29. Connect to Secondary Science
- 30. Connect to mathematics
- 31. Connect to computing
- 32. Connect to pathways
- 33. Did you know?
- 34. Write a reader-ready mapping note
- 35. Build a validation ladder
- 36. Finish with a reproducible bundle
Section 1 of 36
1. Start with the comparison
Define whether the study compares gene expression, morphology, development, phenotype or sample metadata. Each goal needs different anatomical granularity.
Working checkpoint — Uberon identifier: Write the species, stage and intended claim before choosing a term. Keep species, life stage, laterality, sampling method and original anatomy term beside the Uberon class.
For mathematics, treat the ontology as a graph whose edges carry different meanings.
Section 2 of 36
2. Use the Uberon identifier
A stable ID anchors the class even when preferred labels or synonyms change.
Working checkpoint — anatomical structure: Store namespace, identifier, label and release together. Inspect the relation used—part, precursor, overlap or bridge—because each supports a different query.
For computing, validate taxon constraints and unresolved identifiers before analysis starts.
Section 3 of 36
3. Read the definition
The definition sets the intended anatomical scope and prevents label-only assumptions.
Working checkpoint — part-of relation: Compare it with specimen notes and source-ontology meaning. Pin the Uberon product, bridge files, taxonomy and source ontology release together.
For writing, say “mapped to” or “comparable under this bridge” when equivalence is conditional.
Section 4 of 36
4. Inspect the is-a hierarchy
Broader and narrower classes support queries at several levels of resolution.
Working checkpoint — developmental relation: Choose the deepest term justified by the source evidence. Separate cross-species comparability from claims of structural, developmental or functional identity.
For students, ask whether the comparison is about position, parts, development, function or ancestry.
Section 5 of 36
5. Traverse part-of relations
Organs contain tissues and substructures, but part relationships are not the same as class inheritance.
Working checkpoint — taxon constraint: State whether a query expands through parts and why. Retain uncertain and one-to-many mappings so biological differences do not disappear during harmonisation.
For parents, a calm checkpoint is whether the anatomical label matches what was actually sampled.
Section 6 of 36
6. Trace develops-from relations
Embryonic precursors can connect to later anatomy through curated developmental relations.
Working checkpoint — bridge axiom: Separate developmental lineage from adult spatial similarity. Keep species, life stage, laterality, sampling method and original anatomy term beside the Uberon class.
For mathematics, treat the ontology as a graph whose edges carry different meanings.
Section 7 of 36
7. Record life stage
Anatomical identities and boundaries can change across embryonic, juvenile and adult stages.
Working checkpoint — Uberon identifier: Use an explicit stage ontology or documented local stage. Inspect the relation used—part, precursor, overlap or bridge—because each supports a different query.
For computing, validate taxon constraints and unresolved identifiers before analysis starts.
Section 8 of 36
8. Record taxon
A cross-species class still needs the organism identity for biological interpretation.
Working checkpoint — anatomical structure: Attach a stable taxonomy identifier to every specimen. Pin the Uberon product, bridge files, taxonomy and source ontology release together.
For writing, say “mapped to” or “comparable under this bridge” when equivalence is conditional.
Section 9 of 36
9. Check taxon constraints
Constraints can flag anatomy that should not occur in a taxon under the ontology model.
Working checkpoint — part-of relation: Review conflicts rather than deleting them automatically. Separate cross-species comparability from claims of structural, developmental or functional identity.
For students, ask whether the comparison is about position, parts, development, function or ancestry.
Section 10 of 36
10. Preserve laterality
Paired structures may require left, right or bilateral qualifiers.
Working checkpoint — developmental relation: Keep laterality as a structured field and in the specimen label. Retain uncertain and one-to-many mappings so biological differences do not disappear during harmonisation.
For parents, a calm checkpoint is whether the anatomical label matches what was actually sampled.
Section 11 of 36
11. Preserve regional position
Anterior, posterior, proximal, distal and segmental distinctions can matter experimentally.
Working checkpoint — taxon constraint: Avoid collapsing qualifiers that affect the scientific question. Keep species, life stage, laterality, sampling method and original anatomy term beside the Uberon class.
For mathematics, treat the ontology as a graph whose edges carry different meanings.
Section 12 of 36
12. Keep source terminology
The original anatomy ontology may carry species-specific detail and curation history.
Working checkpoint — bridge axiom: Store the source ID and mapping relation beside Uberon. Inspect the relation used—part, precursor, overlap or bridge—because each supports a different query.
For computing, validate taxon constraints and unresolved identifiers before analysis starts.
Section 13 of 36
13. Understand bridge axioms
Bridge files relate Uberon to foreign ontologies using explicit logical statements.
Working checkpoint — Uberon identifier: Inspect the axiom rather than assuming every mapping is exact. Pin the Uberon product, bridge files, taxonomy and source ontology release together.
For writing, say “mapped to” or “comparable under this bridge” when equivalence is conditional.
Section 14 of 36
14. Use taxon-restricted equivalence
A foreign term may equal an Uberon class only within a named taxon.
Working checkpoint — anatomical structure: Carry the taxon restriction into downstream queries. Separate cross-species comparability from claims of structural, developmental or functional identity.
For students, ask whether the comparison is about position, parts, development, function or ancestry.
Section 15 of 36
15. Distinguish collected products
Collected ontologies combine components while retaining source terms and may contain redundancy.
Working checkpoint — part-of relation: Name the exact product before reporting class counts. Retain uncertain and one-to-many mappings so biological differences do not disappear during harmonisation.
For parents, a calm checkpoint is whether the anatomical label matches what was actually sampled.
Section 16 of 36
16. Distinguish composite products
Composite products can remove redundant species-specific terms under their build rules.
Working checkpoint — developmental relation: Check whether provenance and mappings remain available. Keep species, life stage, laterality, sampling method and original anatomy term beside the Uberon class.
For mathematics, treat the ontology as a graph whose edges carry different meanings.
Section 17 of 36
17. Choose the canonical release carefully
The canonical PURL points to the latest main release, which supports discovery but not exact historical reruns.
Working checkpoint — taxon constraint: Archive the retrieved file or versioned URL and checksum. Inspect the relation used—part, precursor, overlap or bridge—because each supports a different query.
For computing, validate taxon constraints and unresolved identifiers before analysis starts.
Section 18 of 36
18. Validate imports
Uberon interoperates with cell, function, phenotype and taxonomy resources whose versions can alter reasoning.
Working checkpoint — bridge axiom: Record all imported ontology versions. Pin the Uberon product, bridge files, taxonomy and source ontology release together.
For writing, say “mapped to” or “comparable under this bridge” when equivalence is conditional.
Section 19 of 36
19. Query cross-species expression
Resources such as Bgee use anatomy mappings to compare where genes are expressed across animals.
Working checkpoint — Uberon identifier: Control for assay, stage, tissue composition and species. Separate cross-species comparability from claims of structural, developmental or functional identity.
For students, ask whether the comparison is about position, parts, development, function or ancestry.
Section 20 of 36
20. Annotate biosamples
Projects such as ENCODE and FANTOM5 use anatomy and cell terms to describe samples.
Working checkpoint — anatomical structure: Keep collection and dissection metadata with each annotation. Retain uncertain and one-to-many mappings so biological differences do not disappear during harmonisation.
For parents, a calm checkpoint is whether the anatomical label matches what was actually sampled.
Section 21 of 36
21. Challenge dissection precision
A sample labelled with an organ may include several tissues or adjacent structures.
Working checkpoint — part-of relation: Treat the recorded structure as a sampling claim, not microscopic purity. Keep species, life stage, laterality, sampling method and original anatomy term beside the Uberon class.
For mathematics, treat the ontology as a graph whose edges carry different meanings.
Section 22 of 36
22. Challenge imaging resolution
Microscopy may resolve boundaries differently from gross anatomy or bulk dissection.
Working checkpoint — developmental relation: Match ontology depth to measurement resolution. Inspect the relation used—part, precursor, overlap or bridge—because each supports a different query.
For computing, validate taxon constraints and unresolved identifiers before analysis starts.
Section 23 of 36
23. Challenge one-to-many mappings
One local class may correspond to several Uberon candidates depending on species or stage.
Working checkpoint — taxon constraint: Preserve alternatives and the rule used to choose among them. Pin the Uberon product, bridge files, taxonomy and source ontology release together.
For writing, say “mapped to” or “comparable under this bridge” when equivalence is conditional.
Section 24 of 36
24. Challenge many-to-one mappings
Several detailed source terms may map to one cross-species parent, hiding useful distinctions.
Working checkpoint — bridge axiom: Keep source IDs for any species-specific follow-up. Separate cross-species comparability from claims of structural, developmental or functional identity.
For students, ask whether the comparison is about position, parts, development, function or ancestry.
Section 25 of 36
25. Do not infer homology automatically
Similarity, location and shared names are not sufficient evidence of common ancestry.
Working checkpoint — Uberon identifier: Use explicit homology evidence and vocabulary when the claim requires it. Retain uncertain and one-to-many mappings so biological differences do not disappear during harmonisation.
For parents, a calm checkpoint is whether the anatomical label matches what was actually sampled.
Section 26 of 36
26. Audit relation-aware expansion
Queries that include parts or precursors can retrieve surprising records.
Working checkpoint — anatomical structure: Test a curated set of expected inclusions and exclusions. Keep species, life stage, laterality, sampling method and original anatomy term beside the Uberon class.
For mathematics, treat the ontology as a graph whose edges carry different meanings.
Section 27 of 36
27. Protect human metadata
Anatomical sample detail can increase re-identification risk when linked to clinical or genomic records.
Working checkpoint — part-of relation: Apply consent, access and minimum-necessary disclosure. Inspect the relation used—part, precursor, overlap or bridge—because each supports a different query.
For computing, validate taxon constraints and unresolved identifiers before analysis starts.
Section 28 of 36
28. Connect to Primary Science
Body parts, habitats and classification show how shared names organise observations.
Working checkpoint — developmental relation: Use animal diagrams and clearly label the activity as a model. Pin the Uberon product, bridge files, taxonomy and source ontology release together.
For writing, say “mapped to” or “comparable under this bridge” when equivalence is conditional.
Section 29 of 36
29. Connect to Secondary Science
Organisation from cells to tissues to organs illustrates nested biological systems and development.
Working checkpoint — taxon constraint: Ask which edge means class, part or precursor. Separate cross-species comparability from claims of structural, developmental or functional identity.
For students, ask whether the comparison is about position, parts, development, function or ancestry.
Section 30 of 36
30. Connect to mathematics
Graphs, sets and transitive relations explain why a query can expand through an anatomical network.
Working checkpoint — bridge axiom: Work with invented structures and draw labelled edges. Retain uncertain and one-to-many mappings so biological differences do not disappear during harmonisation.
For parents, a calm checkpoint is whether the anatomical label matches what was actually sampled.
Section 31 of 36
31. Connect to computing
Release files, identifiers, bridge axioms and validation rules are executable infrastructure.
Working checkpoint — Uberon identifier: Use tests for orphan IDs, taxon conflicts and cycles. Keep species, life stage, laterality, sampling method and original anatomy term beside the Uberon class.
For mathematics, treat the ontology as a graph whose edges carry different meanings.
Section 32 of 36
32. Connect to pathways
Comparative anatomy links zoology, developmental biology, genomics, informatics and data stewardship.
Working checkpoint — anatomical structure: Check official programme pages for current study routes. Inspect the relation used—part, precursor, overlap or bridge—because each supports a different query.
For computing, validate taxon constraints and unresolved identifiers before analysis starts.
Section 33 of 36
33. Did you know?
The official Uberon pages describe both collected and composite multispecies products, reflecting different ways to combine anatomy ontologies.
Working checkpoint — part-of relation: Ask which product a database used before comparing counts. Pin the Uberon product, bridge files, taxonomy and source ontology release together.
For writing, say “mapped to” or “comparable under this bridge” when equivalence is conditional.
Section 34 of 36
34. Write a reader-ready mapping note
Name the source term, Uberon term, species, stage, relation, evidence, release and unresolved alternatives.
Working checkpoint — developmental relation: Put conditional equivalence in plain language. Separate cross-species comparability from claims of structural, developmental or functional identity.
For students, ask whether the comparison is about position, parts, development, function or ancestry.
Section 35 of 36
35. Build a validation ladder
Move from identifier checks to definitions, taxon constraints, source-ontology review, expert curation and empirical compatibility.
Working checkpoint — taxon constraint: Stop the comparison at the strongest completed evidence. Retain uncertain and one-to-many mappings so biological differences do not disappear during harmonisation.
For parents, a calm checkpoint is whether the anatomical label matches what was actually sampled.
Section 36 of 36
36. Finish with a reproducible bundle
Deliver specimen terms, source and Uberon IDs, releases, bridge axioms, taxonomy, stage, mappings, tests and decision history.
Working checkpoint — bridge axiom: Make every comparative result traceable to its anatomy mapping. Keep species, life stage, laterality, sampling method and original anatomy term beside the Uberon class.
For mathematics, treat the ontology as a graph whose edges carry different meanings.
Contents · Previous section · Continue to the Science Learning Hub