Hoodini Colab
Hoodini Colab is an interactive Jupyter widget that provides a visual interface for configuring and running Hoodini analyses. Designed specifically for Google Colab, it allows researchers to run complex genomic neighborhood analyses directly in their browser without any local installation.
No installation required! Just open the notebook in Google Colab and start analyzing.
Why Hoodini Colab?
Instead of remembering dozens of command-line parameters and flags, you get a visual interface where you can click, select, and configure everything through an intuitive web-based UI.
Quick Start
Open the Colab Notebook
Click the button below to open the interactive notebook in Google Colab:
Run the Setup Cell
Run the cell below only once. Running it multiple times may cause errors!
!pip install -q hoodini-colab
from hoodini_colab import create_launcher
launcher = create_launcher()
display(launcher)Configure and Run
Use the visual interface to configure your analysis, then click “Run Hoodini Analysis”.
First run takes 5-10 minutes: The first time you run Hoodini, it needs to install the tool and download reference databases. More annotation tools selected means more databases to download. Subsequent runs in the same session are much faster.
Key Features
The launcher interface is organized into collapsible sections covering all aspects of Hoodini’s functionality:
| Section | Description |
|---|---|
| Input/Output | Configure file paths and output directories |
| Remote BLAST | E-values and number of targets for database searches |
| Performance | Thread count and NCBI API keys |
| Neighborhood | Window sizes around target proteins |
| Clustering | Sequence grouping methods and thresholds |
| Tree Construction | Taxonomy, NJ, ML, or distance-based trees |
| Annotations | PADLOC, DefenseFinder, CCtyper, geNomad, and more |
| Links | Protein and nucleotide similarity connections |
Every parameter shows helpful descriptions explaining what it does. The generated command updates instantly as you make changes, and you can copy it to run manually later or click “Run” to execute immediately.
Local Installation
While optimized for Google Colab, you can also use hoodini-colab in local Jupyter notebooks:
pip install hoodini-colabThen in your notebook:
from hoodini_colab import create_launcher
launcher = create_launcher()
display(launcher)When running locally, you still need Hoodini installed. The widget will attempt to install it via pixi if not found.
Next Steps
- Quick Start Guide — Step-by-step tutorial
- Use Cases — Common analysis scenarios
- API Reference — Programmatic access to the widget