To translate gene symbols, gene names, NCBI Gene IDs and UniProt accession numbers accurately, a translator must separate biological meaning from database identity. Genomics papers, laboratory reports, bioinformatics pipelines, product documentation and educational resources often place approved gene symbols, full gene names, stable database identifiers and protein accessions side by side. The human-language descriptions may need translation; the identifiers normally must remain exact. A small change in capitalization, species context or accession can point the target document toward a different gene or protein.
This guide explains how to translate genetics and protein information without changing HGNC-approved symbols, HGNC IDs, NCBI GeneIDs or UniProtKB accession numbers. It solves one distinct search intent: what does each identifier represent, which elements can be localized, why gene symbols are not ordinary abbreviations, how gene and protein identity differ, and how multilingual scientific documents can stay traceable to the same biological record after translation.
The working rule is translate the descriptive name, preserve the approved symbol or database identifier, record the species and database context, and verify every relationship. HGNC assigns approved names, symbols and stable HGNC IDs to human genes; NCBI Gene assigns a stable unique GeneID across taxa; UniProtKB uses stable accession numbers to identify protein entries. These systems are related through cross-references but they do not identify the same biological layer. This specialist branch connects to Master Art of Translation, the protected Vocabulary Learning Hub, and How English Works without replacing the broader biotechnology or life-sciences translation owners.
One locus can have many labels
A human gene can have an approved symbol, approved full name, aliases, previous symbols, database IDs and linked protein records. Treating all labels as interchangeable can fragment identity or merge different biological objects. Build a gene-and-protein identity map before translating. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
HGNC is the authority for human gene nomenclature
HGNC assigns approved symbols and names to human genes and loci. A translator who invents a target-language symbol can create a symbol that the scientific community does not recognize. Preserve the approved symbol and translate explanatory prose separately. The verification question is not only “Did the characters survive?” but “Does this target still point to the same biological record?” Use the authoritative database, keep the namespace visible, and record any historical or current-name difference instead of silently rewriting it.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Approved gene symbols are not ordinary abbreviations
HGNC symbols are standardized identifiers used in scientific communication and electronic retrieval. Expanding or translating a symbol because its letters resemble English words destroys nomenclature consistency. Keep the symbol exactly as approved. For data pipelines, store the identifier as text and exclude it from locale formatting, spell-checking and machine-generated expansion. Let the surrounding scientific explanation carry the target language while the database key remains stable for search and reproducibility.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
HGNC approved names are human-readable terminology
The full approved gene name explains the gene in natural language more than the compact symbol does. A literal target translation may not be an official nomenclature form. Preserve the approved source name and add a target explanation when needed. A useful reviewer habit is to check identity in both directions: start from the target term and confirm the code, then start from the code and confirm the species, record type and current approved name. Bidirectional checking catches plausible-looking gene and protein mix-ups.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
HGNC IDs are stable gene identifiers
Each HGNC-approved gene has a unique HGNC ID such as HGNC:number. Symbols and names can change over time while the HGNC ID remains stable. Use HGNC ID as a durable cross-reference and preserve it exactly. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Previous symbols matter in historical literature
Older papers may use symbols that were once approved but later changed. Replacing every historical symbol with the current symbol can erase what the source actually said. Preserve historical wording and add current cross-reference separately. The verification question is not only “Did the characters survive?” but “Does this target still point to the same biological record?” Use the authoritative database, keep the namespace visible, and record any historical or current-name difference instead of silently rewriting it.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Aliases are not always approved symbols
Literature and databases can contain informal aliases and legacy names. A familiar alias can be ambiguous or reused for different genes. Distinguish approved symbol from alias in the target. For data pipelines, store the identifier as text and exclude it from locale formatting, spell-checking and machine-generated expansion. Let the surrounding scientific explanation carry the target language while the database key remains stable for search and reproducibility.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Gene symbols are species-sensitive
Similar symbols can exist across species and ortholog conventions differ. Dropping species context can make a correct symbol point readers toward the wrong organism. Keep species name or taxonomic context visible. A useful reviewer habit is to check identity in both directions: start from the target term and confirm the code, then start from the code and confirm the species, record type and current approved name. Bidirectional checking catches plausible-looking gene and protein mix-ups.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Human gene symbols use capitalization conventions
HGNC-approved human symbols use defined uppercase conventions. Automatic sentence case or title case can corrupt the scientific form. Exclude gene symbols from ordinary style transformations. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Italics are a formatting convention, not identity
Scientific style often italicizes gene symbols while protein symbols are handled differently. Losing italics can create ambiguity in prose even when the letters remain correct. Follow the publication’s gene-versus-protein style consistently. The verification question is not only “Did the characters survive?” but “Does this target still point to the same biological record?” Use the authoritative database, keep the namespace visible, and record any historical or current-name difference instead of silently rewriting it.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Gene and protein are related but not identical
A gene can encode one or more protein products, and the protein record belongs to a different biological layer. Using a protein accession where a gene identifier is expected can misstate the object. Label gene and protein fields separately. For data pipelines, store the identifier as text and exclude it from locale formatting, spell-checking and machine-generated expansion. Let the surrounding scientific explanation carry the target language while the database key remains stable for search and reproducibility.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
NCBI GeneID is a stable database identifier
NCBI Gene assigns a stable unique GeneID and notes that GeneID is unique across taxa. A preferred symbol can change or be non-unique in some contexts while the GeneID remains the database key. Preserve GeneID and its NCBI context. A useful reviewer habit is to check identity in both directions: start from the target term and confirm the code, then start from the code and confirm the species, record type and current approved name. Bidirectional checking catches plausible-looking gene and protein mix-ups.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
GeneID is not the same as HGNC ID
HGNC ID and NCBI GeneID come from different resources even when they cross-reference the same human gene. Copying one into the other’s field creates a valid-looking but false identifier. Keep database namespace with the number. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Database namespace is part of identity
A bare integer is ambiguous when many biological databases use numeric IDs. Removing prefixes such as HGNC or labels such as GeneID weakens traceability. Store identifier and database name together. The verification question is not only “Did the characters survive?” but “Does this target still point to the same biological record?” Use the authoritative database, keep the namespace visible, and record any historical or current-name difference instead of silently rewriting it.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
NCBI preferred symbol can follow nomenclature authorities
NCBI Gene integrates nomenclature information from recognized authorities where available. Translators should not overwrite an official symbol with a home-made localized abbreviation. Use approved nomenclature as the anchor. For data pipelines, store the identifier as text and exclude it from locale formatting, spell-checking and machine-generated expansion. Let the surrounding scientific explanation carry the target language while the database key remains stable for search and reproducibility.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
LOC symbols can be provisional-style database labels
NCBI may use LOC plus a GeneID when an official published symbol is unavailable. Translating LOC as if it were an English word changes the database label. Preserve the exact LOC symbol until the record is updated. A useful reviewer habit is to check identity in both directions: start from the target term and confirm the code, then start from the code and confirm the species, record type and current approved name. Bidirectional checking catches plausible-looking gene and protein mix-ups.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Gene history matters
NCBI tracks discontinued or replaced GeneIDs through history resources. Silently replacing a historical GeneID can make archival data impossible to reproduce. Preserve source-period ID and document any replacement separately. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
UniProtKB accessions identify protein entries
UniProtKB assigns stable accession numbers to entries and recommends the primary accession for citation. A protein accession is not a gene symbol or gene database ID. Label the protein layer explicitly. The verification question is not only “Did the characters survive?” but “Does this target still point to the same biological record?” Use the authoritative database, keep the namespace visible, and record any historical or current-name difference instead of silently rewriting it.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Primary UniProt accession is the citable key
UniProt distinguishes primary and secondary accessions after merges or other record history. Using a secondary accession as if it were the current primary key can confuse citation and linking. Prefer the current primary accession when updating, but preserve historical source data when translating archives. For data pipelines, store the identifier as text and exclude it from locale formatting, spell-checking and machine-generated expansion. Let the surrounding scientific explanation carry the target language while the database key remains stable for search and reproducibility.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
UniProt entry name is not the accession
UniProt entry names are mnemonic identifiers while accessions provide stable entry identity. Treating the entry name as the stable accession can break long-term linking. Keep accession and entry-name fields distinct. A useful reviewer habit is to check identity in both directions: start from the target term and confirm the code, then start from the code and confirm the species, record type and current approved name. Bidirectional checking catches plausible-looking gene and protein mix-ups.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Protein names can be translated cautiously
Protein names are human-readable biological descriptions and may need explanation in the target language. A casual translation can erase enzyme class, receptor family or subunit distinctions. Retain the official protein name and add target explanation when appropriate. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Isoforms complicate protein identity
A gene can generate multiple transcripts and protein isoforms. Calling one gene symbol a single unique protein can oversimplify the record. Preserve isoform or sequence context when the source supplies it. The verification question is not only “Did the characters survive?” but “Does this target still point to the same biological record?” Use the authoritative database, keep the namespace visible, and record any historical or current-name difference instead of silently rewriting it.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Transcript IDs are another layer
RefSeq and Ensembl transcript accessions can identify particular RNA records. A transcript identifier is not interchangeable with GeneID or UniProt accession. Keep transcript, gene and protein namespaces separate. For data pipelines, store the identifier as text and exclude it from locale formatting, spell-checking and machine-generated expansion. Let the surrounding scientific explanation carry the target language while the database key remains stable for search and reproducibility.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Reference sequence versions matter
Accession.version strings can distinguish sequence revisions. Dropping the version can make downstream variant or sequence references ambiguous. Preserve the full accession with version when the source relies on it. A useful reviewer habit is to check identity in both directions: start from the target term and confirm the code, then start from the code and confirm the species, record type and current approved name. Bidirectional checking catches plausible-looking gene and protein mix-ups.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Ensembl IDs are separate database identifiers
Ensembl provides gene, transcript and protein identifiers with its own namespace. Replacing an Ensembl ID with an NCBI GeneID without an explicit mapping changes the data model. Preserve source database identity. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Cross-references are mappings, not synonyms
HGNC, NCBI, Ensembl and UniProt can cross-reference related records. A mapping between databases does not mean every ID identifies the same biological object. Record the relation and direction of mapping. The verification question is not only “Did the characters survive?” but “Does this target still point to the same biological record?” Use the authoritative database, keep the namespace visible, and record any historical or current-name difference instead of silently rewriting it.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Orthologs are not the same gene record
Genes in different species can be orthologous and have related nomenclature. Replacing a human gene with a mouse ortholog because symbols resemble each other changes species identity. Verify organism before mapping or translating. For data pipelines, store the identifier as text and exclude it from locale formatting, spell-checking and machine-generated expansion. Let the surrounding scientific explanation carry the target language while the database key remains stable for search and reproducibility.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Paralogs need symbol precision
Gene families can contain closely related paralogs whose symbols differ by only one digit or letter. A minor typo can point to another valid gene. Use exact symbol matching and family context. A useful reviewer habit is to check identity in both directions: start from the target term and confirm the code, then start from the code and confirm the species, record type and current approved name. Bidirectional checking catches plausible-looking gene and protein mix-ups.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Gene families share roots
HGNC often uses common symbol roots for related genes. A translator may mistake the shared root for an interchangeable generic term. Preserve the full symbol including family-member distinctions. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Pseudogenes have their own identity
Pseudogenes can have approved symbols and IDs distinct from functional genes. Dropping a suffix or qualifier can collapse pseudogene and gene records. Keep locus type and symbol exact. The verification question is not only “Did the characters survive?” but “Does this target still point to the same biological record?” Use the authoritative database, keep the namespace visible, and record any historical or current-name difference instead of silently rewriting it.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Non-coding RNA genes are not protein-coding genes
HGNC covers many locus types including non-coding RNA genes. Assuming every gene symbol maps to a UniProt protein can create false links. Check locus type before connecting gene and protein records. For data pipelines, store the identifier as text and exclude it from locale formatting, spell-checking and machine-generated expansion. Let the surrounding scientific explanation carry the target language while the database key remains stable for search and reproducibility.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Mitochondrial genes use specific nomenclature
Mitochondrial gene symbols follow defined naming conventions. Ordinary style correction can alter their recognized form. Use authoritative nomenclature and preserve case. A useful reviewer habit is to check identity in both directions: start from the target term and confirm the code, then start from the code and confirm the species, record type and current approved name. Bidirectional checking catches plausible-looking gene and protein mix-ups.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Fusion gene notation is structured
Gene fusions can be represented with standardized separators and gene symbols. Replacing separators or translating symbol components can corrupt the fusion identity. Protect the whole fusion notation as structured scientific text. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Readthrough and combined loci need exact notation
Some gene records use standardized punctuation to represent special locus relationships. Typography tools can convert or remove the punctuation. Exclude nomenclature strings from smart punctuation. The verification question is not only “Did the characters survive?” but “Does this target still point to the same biological record?” Use the authoritative database, keep the namespace visible, and record any historical or current-name difference instead of silently rewriting it.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Scientific capitalization can carry meaning
Protein and gene naming conventions can depend on capitalization and italics. Sentence-case automation can make a symbol visually wrong even when letters are otherwise unchanged. Apply scientific styles after translation. For data pipelines, store the identifier as text and exclude it from locale formatting, spell-checking and machine-generated expansion. Let the surrounding scientific explanation carry the target language while the database key remains stable for search and reproducibility.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Hyphens are not free punctuation
Some approved symbols or protein descriptions can include governed hyphens or other separators. Replacing a hyphen with an en dash creates a different visible identifier. Preserve nomenclature punctuation exactly. A useful reviewer habit is to check identity in both directions: start from the target term and confirm the code, then start from the code and confirm the species, record type and current approved name. Bidirectional checking catches plausible-looking gene and protein mix-ups.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
OCR can misread gene symbols
Scanned papers can confuse O/0, I/1, S/5 and similar characters. A misread string may still be another real gene symbol. Verify against HGNC or the relevant database. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Spreadsheets can damage long IDs
Numeric GeneIDs and accession-like strings can be reformatted by spreadsheet software. Scientific notation or dropped leading characters corrupts database identity. Import biological identifiers as text. The verification question is not only “Did the characters survive?” but “Does this target still point to the same biological record?” Use the authoritative database, keep the namespace visible, and record any historical or current-name difference instead of silently rewriting it.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
CSV schema matters
CSV does not preserve data types on its own. The receiving application may reinterpret identifiers as numbers or dates. Define identifier columns as strings in the workflow. For data pipelines, store the identifier as text and exclude it from locale formatting, spell-checking and machine-generated expansion. Let the surrounding scientific explanation carry the target language while the database key remains stable for search and reproducibility.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Machine translation can alter symbols
Language models and MT engines may expand abbreviations or change capitalization. A helpful-looking expansion can create a non-existent symbol. Lock symbols and accessions before automated translation. A useful reviewer habit is to check identity in both directions: start from the target term and confirm the code, then start from the code and confirm the species, record type and current approved name. Bidirectional checking catches plausible-looking gene and protein mix-ups.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Translation memory can spread outdated symbols
Older projects may contain previous approved symbols or obsolete protein names. Blind reuse can reintroduce outdated nomenclature. Check current nomenclature while preserving historical source context. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Historical source versus current nomenclature
An archival paper should remain faithful to the terms used at publication. Replacing every old symbol with the current one can distort quotations and search history. Add current nomenclature as a note, not a silent rewrite. The verification question is not only “Did the characters survive?” but “Does this target still point to the same biological record?” Use the authoritative database, keep the namespace visible, and record any historical or current-name difference instead of silently rewriting it.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Database release dates matter
Biological databases update continuously. A record description can change while the stable identifier remains. Record retrieval date or release when reproducibility matters. For data pipelines, store the identifier as text and exclude it from locale formatting, spell-checking and machine-generated expansion. Let the surrounding scientific explanation carry the target language while the database key remains stable for search and reproducibility.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Gene symbol alone may be ambiguous in older literature
Old symbols can collide across species or disciplines. Search based only on the symbol can retrieve the wrong gene. Use HGNC ID, GeneID, species and context together. A useful reviewer habit is to check identity in both directions: start from the target term and confirm the code, then start from the code and confirm the species, record type and current approved name. Bidirectional checking catches plausible-looking gene and protein mix-ups.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Protein accession alone needs database context
An accession should be paired with UniProtKB or another source name. Bare alphanumeric strings can resemble accessions from other databases. Keep the database namespace visible. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Sequence records can merge or split
UniProt entries can be merged or demerged and retain history through primary and secondary accessions. A current entry can differ from the exact historical protein record cited by the source. Preserve temporal context during archival translation. The verification question is not only “Did the characters survive?” but “Does this target still point to the same biological record?” Use the authoritative database, keep the namespace visible, and record any historical or current-name difference instead of silently rewriting it.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Functional descriptions can change
Biological understanding evolves and database names can be updated. Updating explanatory text can make a historical source sound more certain than it was. Separate translation from scientific revision. For data pipelines, store the identifier as text and exclude it from locale formatting, spell-checking and machine-generated expansion. Let the surrounding scientific explanation carry the target language while the database key remains stable for search and reproducibility.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Laboratory assay labels may contain gene symbols
Assay names can embed gene symbols beside ordinary words such as mutation panel or expression test. Translating the symbol portion can break kit or assay identification. Protect symbols inside mixed labels. A useful reviewer habit is to check identity in both directions: start from the target term and confirm the code, then start from the code and confirm the species, record type and current approved name. Bidirectional checking catches plausible-looking gene and protein mix-ups.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Commercial product names can resemble gene symbols
Biotech products may use gene names in branding or panel names. Treating every capitalized token as a gene symbol can freeze ordinary brand text incorrectly. Classify tokens before applying protection rules. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Clinical reports can contain gene and protein together
A report may mention a gene variant and an associated protein consequence. Using one identifier namespace for both can confuse molecular level. Label gene, transcript and protein layers explicitly. The verification question is not only “Did the characters survive?” but “Does this target still point to the same biological record?” Use the authoritative database, keep the namespace visible, and record any historical or current-name difference instead of silently rewriting it.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Worked example: gene record
A fictional human gene record lists approved symbol, HGNC ID and NCBI GeneID. The target should translate explanatory prose while preserving all three identity fields. Use the stable IDs to verify that the translated name still refers to the same gene. For data pipelines, store the identifier as text and exclude it from locale formatting, spell-checking and machine-generated expansion. Let the surrounding scientific explanation carry the target language while the database key remains stable for search and reproducibility.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Worked example: protein entry
A fictional report links the human gene to a UniProtKB protein accession. The protein accession is related to but not interchangeable with the gene ID. Keep the biological relationship visible in the target. A useful reviewer habit is to check identity in both directions: start from the target term and confirm the code, then start from the code and confirm the species, record type and current approved name. Bidirectional checking catches plausible-looking gene and protein mix-ups.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Error clinic
Common failures include symbol translation, case changes, species loss, GeneID/HGNC swaps and protein/gene conflation. These errors can survive fluent prose review. Run a dedicated biological-identifier QA pass. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Release checklist
Approved symbol, HGNC ID, GeneID, UniProt accession, species and record type all need verification. Checking only spelling is insufficient. Verify tokens, namespace, species and biological layer before release. The verification question is not only “Did the characters survive?” but “Does this target still point to the same biological record?” Use the authoritative database, keep the namespace visible, and record any historical or current-name difference instead of silently rewriting it.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Frequently asked questions
Gene symbols and database IDs should not be translated, while descriptive names can be explained or localized. Users often confuse gene names with protein names. Teach the difference between nomenclature, database identity and biological relationship. For data pipelines, store the identifier as text and exclude it from locale formatting, spell-checking and machine-generated expansion. Let the surrounding scientific explanation carry the target language while the database key remains stable for search and reproducibility.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Connection to eduKateSG architecture
This specialist owner protects gene and protein identity rather than broad biotechnology prose. A general genomics hub would overlap existing life-sciences owners. Use this article for identifiers and existing owners for domain-wide translation strategy. A useful reviewer habit is to check identity in both directions: start from the target term and confirm the code, then start from the code and confirm the species, record type and current approved name. Bidirectional checking catches plausible-looking gene and protein mix-ups.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Authoritative sources
HGNC, NCBI Gene and UniProt publish the governing nomenclature and identifier guidance. Third-party databases can lag or normalize names differently. Verify identity against the authoritative source for each namespace. In translation, mark the human-readable description as translatable, the symbol or database ID as protected, and the species or biological layer as required context. That prevents a correct string from being attached to the wrong organism, transcript or protein.
Before release, confirm what is being named: gene, transcript, protein, isoform or database entry. The target should let a reader recover the same authoritative record without guessing from translated terminology alone.
Reference verification route
For human gene nomenclature, use the HUGO Gene Nomenclature Committee, its gene nomenclature guidelines, and symbol reports. For NCBI identifiers, use the NCBI Gene resource and Gene documentation. For protein-entry identity, use UniProt’s accession-number guidance and UniProtKB records.