Installation

The alphagenome package is a lightweight Python client for the hosted AlphaGenome models and the AlphaGenome Atlas. It runs on Linux, macOS, and Windows, and no local GPU is required.

Model predictions and Atlas queries both need a free AlphaGenome API key.

Install the package

Install the latest release from PyPI inside a virtual environment:

pip

python3 -m venv .venv
source .venv/bin/activate
pip install alphagenome

uv

uv venv
source .venv/bin/activate
uv pip install alphagenome

Or add directly to an existing uv project:

uv add alphagenome

conda

conda create -n alphagenome python -y
conda activate alphagenome
pip install alphagenome

From source

Clone the GitHub repository for an editable install:

git clone https://github.com/google-deepmind/alphagenome.git
cd alphagenome
pip install -e .

To pull upstream changes later:

git pull
pip install -e .

Google Colab

Every tutorial notebook starts with an install cell that installs alphagenome (and any extra packages that notebook needs) in your Colab runtime:

! pip install alphagenome

Add API key to secrets

To authenticate from Google Colab notebooks without pasting your API key into code cells:

1. Open your Google Colab notebook and click the Secrets icon (key symbol) in the left sidebar.

2. Click + Add new secret and name it ALPHA_GENOME_API_KEY.

3. Paste your AlphaGenome API key into the Value field.

4. Toggle Notebook access on for the current notebook.

Local API key configuration

When running Python scripts or Jupyter notebooks locally, export ALPHA_GENOME_API_KEY in your shell environment (or add it to ~/.zshrc / ~/.bashrc):

export ALPHA_GENOME_API_KEY="YOUR_API_KEY"

colab_utils.get_api_key() reads this environment variable first and falls back to Colab Secrets, so the same code runs locally and in Colab.

Verify your installation

Confirm that the package is installed and your API key works:

from alphagenome import colab_utils
from alphagenome.models.v2 import dna_client

dna_model = dna_client.create(colab_utils.get_api_key())
print(len(dna_model.supported_conditions()), 'supported human conditions')

Next steps

• Continue to Getting started to choose a tool and walk through an AlphaGenome 2 workflow.

• Follow the interactive Tutorials covering reference conditions, expression conditioning (cell_conditioning=), and imputed tracks.