PICO

Pipeline for Interpretable Classification on Omics

Description

The PICO is a tool to apply machine learning algorithms to omics datasets. The goal is to extract the most important features because they are potential novel biomarkers. The interface is made to be easy to use and intuitive even for those with small to nonexistant experience in programming and AI.

More generally, it is a tool to apply interpretable machine learning algorithms to tabular data. It focuses on analyzing the features selected by the models.

The documentation

You can find the documentation here. It explains how to use PICO but also how it works.

Authors and contributors

• Élina Francovic-Fontaine
• Vincent Primpied
• Vincent Vilain
• Gabriel Leclerc
• Thierry Moszkowicz
• Mathieu Bazinet
• Thibaud Godon
• Pier-Luc Plante
• Baptiste Bauvin
• Louis-Philippe Vignault

Disclaimer

PICO is still in development. If you encounter any issue or have any suggestion, feel free to contact us at elina.francovic-fontaine.1@ulaval.ca. Or you can leave an issue here with the tag "bug".

For developers

Setup

Clone the project with :

git clone https://github.com/ElinaFF/PICO.git

It is recommanded to setup a virtual environment. When it's done, use your isolated python and install pico package locally and in editable mode with :

python -m pip install -e ".[dev]"

Collaborative development

The entire development process has been done following Python development guidelines. Similarly, we adhered to scikit-learn's developer recommendations and provided code understandable to any user familiar with the domain. Therefore, PICO could become an asset for the research community. Indeed, collaborative development ensures that the improvement directions are the ones needed by the user-base.

License

Pipeline for Interpretable Classification on Omics (PICO) © 2026 by Elina Francovic-Fontaine is licensed under CC BY-NC-SA 4.0