alterlab-gget

gget

Overview

gget is a command-line bioinformatics tool and Python package providing unified access to 20+ genomic databases and analysis methods. Query gene information, sequence analysis, protein structures, expression data, and disease associations through a consistent interface. All gget modules work both as command-line tools and as Python functions.

Project home: development moved to the scverse organisation (github.com/scverse/gget); the manual stays at pachterlab.github.io/gget.

Important: The databases queried by gget are continuously updated, which sometimes changes their structure. gget modules are tested automatically on a biweekly basis and updated to match new database structures when necessary.

Installation

Install gget in a clean virtual environment to avoid conflicts:

# Install (or upgrade) into a clean environment
uv pip install --upgrade gget

In Python/Jupyter

import gget


## Quick Start

Basic usage pattern for all modules:

```bash
# Command-line
gget <module> [arguments] [options]

# Python
gget.module(arguments, options)

Most modules return:

Command-line: JSON (default) or CSV with -csv flag
Python: DataFrame or dictionary

Common flags across modules:

-o/--out: Save results to file
-q/--quiet: Suppress progress information
-csv: Return CSV format (command-line only)

Module Catalog

Pick a module, then see references/module_examples.md for worked CLI + Python examples and references/module_reference.md for the full parameter table.

Module	Purpose	Queried source
`ref`	Reference genome download links/metadata	Ensembl
`search`	Find genes by name/description	Ensembl
`info`	Gene/transcript metadata (~1000 IDs max)	Ensembl, UniProt, NCBI
`seq`	Nucleotide/amino-acid sequences (FASTA)	Ensembl
`blast`	BLAST against standard databases	NCBI BLAST
`blat`	Genomic position of a sequence	UCSC BLAT
`muscle`	Multiple sequence alignment	Muscle5 (local)
`diamond`	Fast local protein/translated alignment	DIAMOND (local)
`pdb`	Experimental protein structures + metadata	RCSB PDB
`alphafold`	Predict 3D protein structure (setup req.)	AlphaFold2 (local)
`elm`	Eukaryotic linear motifs (setup req.)	ELM
`archs4`	Correlated genes / tissue expression	ARCHS4
`cellxgene`	Single-cell RNA-seq (setup req.)	CZ CELLxGENE Census
`enrichr`	Ontology/pathway enrichment	Enrichr
`bgee`	Orthologs and expression	Bgee
`opentargets`	Disease/drug associations	OpenTargets
`cbio`	Cancer genomics heatmaps	cBioPortal
`cosmic`	Somatic cancer mutations (license/account)	COSMIC
`mutate`	Generate mutated sequences	local
`virus`	Download filtered virus genome datasets	NCBI Virus
`g2p`	Residue-level structural/functional annotations	Genomics 2 Proteins portal
`gene_expression`	Mean/variance of normalized expression per partition	8cubeDB
`psi_block`	ψ_block block-level specificity scores	8cubeDB
`specificity`	Gene-level ψ / ζ specificity statistics	8cubeDB
`gpt`	Natural-language text generation (setup req.)	OpenAI API
`setup`	Install third-party deps for a module	local

cbio is exposed in Python as gget.cbio_search() and gget.cbio_plot().

Setup-required modules (gget setup <module> before first use): alphafold (~4GB params, needs uv pip install openmm first), cellxgene, elm, gpt, and cbio.

When to Use This Skill

Quick interactive lookup (gene info, BLAST, one structure, one enrichment) → use gget directly; see references/module_examples.md.
Batch processing / advanced BLAST → use the biopython skill.
Multi-database Python workflows → use the bioservices skill.

Does NOT Trigger

Scenario	Use Instead
Local BLAST+ database builds and large CLI searches	`alterlab-blast`
Scripted Entrez/SeqIO pipelines and file parsing	`alterlab-biopython`
One workflow spanning many web services in Python	`alterlab-bioservices`
Serious CELLxGENE Census querying beyond a one-liner	`alterlab-cellxgene`
Running AlphaFold properly (complexes, confidence analysis)	`alterlab-alphafold`

Chaining several gget modules into a pipeline → see references/workflows.md and the ready-made scripts/ (gene_analysis, batch_sequence_analysis, enrichment_pipeline).

Best Practices (essentials)

Use --limit to bound large queries; save with -o/--out for reproducibility.
Gene symbols are case-sensitive in cellxgene ('PAX7' vs 'Pax7').
Run gget setup before first use of alphafold, cellxgene, elm, gpt.
Process max ~1000 Ensembl IDs at once with gget info.
Database structures change; keep gget updated: uv pip install --upgrade gget.
Use virtual environments to avoid dependency conflicts.

Output Formats

Command-line: JSON default; -csv for CSV; FASTA (seq, mutate); PDB (pdb, alphafold); PNG (cbio plot).
Python: DataFrame/dict default; json=True for JSON; save=True or out="filename" to write; AnnData for cellxgene.

References

references/module_examples.md — worked CLI + Python examples for every module
references/module_reference.md — full parameter tables for all modules
references/database_info.md — queried databases and their update frequencies
references/workflows.md — extended multi-module workflow examples

For additional help:

Official documentation: https://pachterlab.github.io/gget/
GitHub issues: https://github.com/pachterlab/gget/issues
Citation: Luebbert, L. & Pachter, L. (2023). Efficient querying of genomic reference databases with gget. Bioinformatics. https://doi.org/10.1093/bioinformatics/btac836

Part of the AlterLab Academic Skills suite.