Quickgo Database
Pull Gene Ontology annotations from QuickGO to link genes (UniProtKB) to biological processes, functions, and components for bioinformatics agents and science apps.
Overview
QuickGO Database is an agent skill most often used in Build (also Idea research) that searches QuickGO Gene Ontology annotations linking genes to processes, functions, and cellular components.
Install
npx skills add https://github.com/google-deepmind/science-skills --skill quickgo-databaseWhat is this skill?
- `annotation search` subcommand maps gene products to GO Biological Process, Molecular Function, and Cellular Component
- Filter by `geneProductId`, `goId`, `aspect`, `taxonId` (e.g. human 9606), and ECO evidence codes
- Example workflows for UniProtKB IDs (e.g. P04637) and process terms like apoptosis (GO:0006915)
- JSON export via `--output` for downstream ML or review pipelines
- Runs with `uv run scripts/quickgo_tool.py` per SKILL reference
- Documents `annotation search` as the primary QuickGO subcommand for gene-to-GO mapping
- Example queries use `--limit 50` with JSON `--output` files
Adoption & trust: 538 installs on skills.sh; 1.7k GitHub stars; 3/3 security scanners passed (skills.sh audits).
What problem does it solve?
You need authoritative GO annotation rows for a gene or term without manually clicking through QuickGO or hand-writing API clients in every session.
Who is it for?
Solo builders shipping bioinformatics helpers, lab automation, or AI research agents that must cite GO mappings with filters.
Skip if: General biology tutoring without database lookups, clinical diagnosis workflows, or users unfamiliar with GO/ECO identifiers.
When should I use this skill?
User needs GO annotations, QuickGO lookups, gene product to biological process/function/component links, or UniProtKB-driven functional mapping.
What do I get? / Deliverables
The agent runs filtered QuickGO annotation searches and saves structured JSON you can feed into analysis, reports, or feature engineering.
- JSON annotation result files (e.g. `annotations.json`, `p53_bp_annotations.json`)
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Primary shelf is Build because usage is scripted API-style queries via `quickgo_tool.py`, typical when implementing research features or data pipelines. Integrations reflects external QuickGO/GO database access rather than notebook-only exploration.
Where it fits
Wire annotation search into a prototype that enriches UniProt IDs with GO terms for a dashboard API.
Pull apoptosis-related annotations to validate whether a niche functional-genomics tool idea has enough public GO coverage.
Generate JSON annotation caches the backend serves to an agent explaining protein roles to end users.
How it compares
Database integration skill for GO annotations—not a full variant caller or literature review skill.
Common Questions / FAQ
Who is quickgo-database for?
Developers and computational biologists using agentic coding tools who need QuickGO GO annotation queries scripted from SKILL.md examples.
When should I use quickgo-database?
During Build integrations when implementing annotation lookup features, and during Idea research when exploring gene function or apoptosis-related gene sets before scoping a science product.
Is quickgo-database safe to install?
It runs local Python tooling that calls external QuickGO services; review the Security Audits panel on this Prism page and validate network usage and outputs before production pipelines.
SKILL.md
READMESKILL.md - Quickgo Database
# Gene Ontology Annotations Reference ## QuickGO `annotation` Subcommand Use the `annotation` subcommand to search for GO annotations linked to gene products. This is the primary functional mapper linking a gene directly to Biological Processes, Molecular Functions, and Cellular Components. ### Searching Annotations ```bash # Find experimentally-validated (EXP=ECO:0000269) annotations for a specific UniProtKB ID uv run scripts/quickgo_tool.py annotation search --geneProductId "UniProtKB:P04070" --taxonId 9606 --evidenceCode "ECO:0000269" --limit 50 --output annotations.json # Find all annotations for a specific GO ID (e.g. apoptosis) uv run scripts/quickgo_tool.py annotation search --goId "GO:0006915" --goUsage desc --taxonId 9606 --limit 50 --output apoptosis_annotations.json # Find Biological Process annotations for a specific UniProtKB ID uv run scripts/quickgo_tool.py annotation search --geneProductId "UniProtKB:P04637" --aspect "biological_process" --limit 50 --output p53_bp_annotations.json ``` **Parameters:** - `--geneProductId`: The database identifier for the gene product (e.g., `UniProtKB:P04637`). - `--goId`: The Gene Ontology ID (e.g., `GO:0006915`). - `--aspect`: Filter by GO aspect (`biological_process`, `molecular_function`, `cellular_component`). - `--taxonId`: NCBI Taxonomy ID (e.g., `9606` for Human). - `--evidenceCode`: The ECO ID corresponding to the evidence type (e.g., `ECO:0000269` for EXP, experimental evidence). Note that many electronic annotations are assigned `ECO:0000501` (IEA). - `--goUsage`: How to use the `goId` parameter. Can be `exact` (only annotations exactly matching the ID), `desc` (annotations matching the ID or any of its descendants), or `slim` (treat the IDs as a GO slim). - `--qualifier`: Qualifier such as `enables`, `part_of`, `involved_in`, `acts_upstream_of`, etc. - `--limit`: Maximum number of results to return per page (max: 100, default: 25). - `--page`: Page number for pagination (default: 1). - `--output`: The JSON file to save the results. # Evidence & Conclusion Ontology (ECO) Terms Reference ## QuickGO `eco` Subcommand Use the `eco` subcommand to search and retrieve details about Evidence & Conclusion Ontology terms. These terms are used as evidence codes in GO annotations (e.g. ECO:0000269 for "experimental evidence used in manual assertion"). ### 1. Searching for ECO Terms Search the Evidence & Conclusion Ontology for a specific query string. ```bash uv run scripts/quickgo_tool.py eco search --query "experimental" --limit 5 --output eco_search_results.json ``` **Parameters:** - `--query`: The text you are looking for. - `--limit`: Maximum number of results to return per page (max: 100, default: 25). - `--page`: Page number for pagination (default: 1). - `--output`: The JSON file to save the results. ### 2. Getting ECO Term Details Fetch detailed information about a specific ECO term or a set of ECO terms by their IDs. ```bash # Get core attributes of an ECO term uv run scripts/quickgo_tool.py eco terms --ids "ECO:0000269" --output eco_term_details.json # Get ancestors of an ECO term uv run scripts/quickgo_tool.py eco terms --ids "ECO:0000269" --relation ancestors --output eco_term_ancestors.json ``` **Parameters:** - `--ids`: Comma-separated list of ECO IDs (e.g., "ECO:0000269"). - `--relation`: Optional. Can be `ancestors`, `descendants`, `children`, `complete`, or `paths`. - `--output`: The JSON file to save the results. # Gene Product Reference ## QuickGO `geneproduct` Subcommand Use the `geneproduct` subcommand to search for gene products across databases like UniProtKB, RNAcentral, and ComplexPortal. This is useful when you have a common gene symbol (e.g., "PROC") but you need its formal database identifier (e.g., "UniProtKB:P04070") to perform a strict annotation search. ### Searching Gene Products ```bash # Find gene product by symbol uv run scripts/quickgo_tool.p