unibind-database

researchgoogle-deepmind1.9K installs

Queries the UniBind database for experimentally validated transcription factor (TF) binding sites. Use when retrieving direct TF-DNA interaction datasets, downloading binding site coordinates (BED/FASTA) for local analysis, or listing available datasets by species, cell line, or TF name. Don't use to query specific intervals, locations, genes, motif models or expression data.

Install

npx skills add https://github.com/google-deepmind/science-skills --skill unibind-database

SKILL.md

---name: unibind-databasedescription: >----

UniBind Database Skill

UniBind is a database of direct TF–DNA interactions across 9 species, integrating ChIP-seq peaks with JASPAR TF binding profiles via the DAMO framework.

Prerequisites

uv: Read the uv skill and follow its Setup instructions to ensure uv is installed and on PATH.
User Notification: If .licenses/unibind_database_LICENSE.txt does not already exist in the workspace root directory then (1) prominently notify the user to check the terms at https://unibind.uio.no/ and https://unibind.uio.no/api/overview, then (2) create the file recording the notification text and timestamp.

Quick Start

Query commands print JSON to stdout by default. Most outputs are small enough to read directly. For large outputs (list_cell_lines, list_tfs), pipe through jq to extract only the fields you need.

uv run <SKILL DIR>/scripts/unibind_api.py list_species

The download_tfbs command writes BED/FASTA files to --output-dir instead. You may optionally use --output <path> on any query command to save results to a file if needed.

Core Rules

Use the Wrapper: ALWAYS execute the provided helper scripts to query the database rather than accessing the database directly. The scripts automatically enforce the required rate limit gracefully.
Output: Query commands print JSON to stdout. Most responses are compact and can be read directly.
Large Results: list_cell_lines and list_tfs produce large output. Pipe these through jq to extract specific fields rather than reading the full output into context.
Saving to File: Use --output <path> when you need to reference the data later or when processing very large results with jq.
Pagination: Use --page and --page-size (max 1000) to chunk large result sets.
Ordering: Use --order field_name (prefix with - for descending) on any list command.
Notification: If this skill is used, ensure this is mentioned in the output.

Utility Scripts

Replace <SKILL DIR> with the absolute path to this skill's directory.

1. List Species

uv run <SKILL DIR>/scripts/unibind_api.py list_species

2. List Collections

uv run <SKILL DIR>/scripts/unibind_api.py list_collections

3. List Cell Lines & TFs (large output — use `jp`)

These commands return large datasets. Use uvx --from jmespath jp to extract only the fields you need.

uv run <SKILL DIR>/scripts/unibind_api.py list_cell_lines | uvx --from jmespath jp "results[].name"
uv run <SKILL DIR>/scripts/unibind_api.py list_tfs | uvx --from jmespath jp "results[].tf_name"

4. List and Filter Datasets (and Profile-Specific Datasets)

Filter datasets using the following arguments:

--species (e.g., "Homo sapiens")
--tf-name (e.g., "CTCF")
--cell-line (e.g., "mESC")
--collection (e.g., Permissive, Robust)
--search (a search term)
--biological-condition (biological condition or source)
--data-source (source of data, e.g., "ENCODE")
--has-pvalue ("true" or "false")
--identifier (e.g., "GSE60130")
--jaspar-id (JASPAR database profile matrix ID)
--model (prediction model)
--summary (summary filter)
--threshold-pvalue (p-value threshold)

Use list_datasets for standard datasets, or list_specific_datasets for profile-specific queries.

uv run <SKILL DIR>/scripts/unibind_api.py list_datasets --species "Homo sapiens" --tf-name "CTCF" --data-source "ENCODE"
uv run <SKILL DIR>/scripts/unibind_api.py list_specific_datasets --species "Mus musculus" --cell-line "mESC"

5. Get Dataset Details

uv run <SKILL DIR>/scripts/unibind_api.py get_dataset "EXP047889.HMLE-Twist-ER_breast_cancer.SMAD3"

6. Download TFBS Files (BED / FASTA)

Downloads all TFBS files for a dataset to a local directory. Use --format bed (default) or --format fasta.

uv run <SKILL DIR>/scripts/unibind_api.py download_tfbs "EXP047889.HMLE-Twist-ER_breast_cancer.SMAD3" --output-dir /tmp/tfbs --format bed

Anti-Patterns

DON'T attempt to use the UniBind API to query specific genomic intervals, locations, or genes.
DON'T guess or hallucinate genome coordinates. Always use ensembl-database as an external check if you're pulling local BED tracks for offline bedtools intersection.
DON'T use for motif models (PFMs). Use the jaspar-database skill instead.
DON'T use for gene expression data. UniBind only stores binding events.
DON'T assume tissue-specific expression from dataset lists alone.
DON'T use cat to read large JSON output files into context. The output is too large. Use jq or write your own code to parse the output files.