gene-symbol-normalizer

name	gene_symbol_normalizer
description	Normalize gene symbols, aliases, and species context before downstream lookup or analysis.
category	bio/literature
version	1
requires_tools	["ensembl_api","uniprot_api","python_repl"]
requires_network	true
user_invocable	true
tags	["gene-symbol","alias","identifier","normalization"]
aliases	["normalize_gene_symbol","gene_name_cleaner"]
species	any
modality	literature
stage	utilities
stability	stable
safety_level	low

Gene Symbol Normalizer

Purpose

Resolve gene aliases, outdated symbols, species ambiguity, and mixed capitalization before any downstream biological interpretation.

When to use

Use this skill when the user provides one or more gene names that may be aliases, old symbols, or mixed-species identifiers.

Required inputs

• genes: one or more gene symbols, aliases, or identifiers
• species (optional): human, mouse, or another organism if the user already knows it

Steps

1. Parse the input into a clean list of candidate gene names and restate any stated or inferred species assumptions.
2. Use ensembl_api as the primary source for canonical symbols, aliases, and organism mapping.
3. Use uniprot_api as a cross-check when alias families, protein naming, or species assignments remain unclear.
4. Use python_repl to assemble a compact normalization table with input term, canonical symbol, species, matched source, and ambiguity or confidence note.
5. Separate exact matches, likely alias matches, and unresolved collisions; do not silently choose between equally plausible species or symbols.
6. Return the normalized mapping together with the source basis, uncertainty, and the most useful downstream next step.

Output format

• Biological context or assumptions: stated or inferred species, identifier type, and any naming assumptions.
• Evidence or source basis: which ensembl_api and uniprot_api results supported each normalization.
• Normalized mapping: Input | Canonical symbol | Species | Match type | Note
• Caveats or ambiguity: unresolved aliases, multi-species collisions, or low-confidence mappings.
• Recommended next step: suggest the next skill or lookup if helpful.

Failure modes

• No confident match: report that the symbol could not be normalized and say which sources were checked.
• Multiple species match equally well: ask the user to specify species before choosing a canonical symbol.
• Alias collision: return all plausible mappings instead of choosing one blindly.

Examples

• "Normalize these genes: P53, Trp53, Cdkn1a."
• "Are PD1 and PDCD1 the same gene?"