Our Nature Machine Intelligence paper is out! 🧬💊

Our latest work, Towards a more inductive world for drug repurposing approaches, has been published in Nature Machine Intelligence (Nature Portfolio)! 🧬💊 In this work, we compare the current state-of-the-art models for DTI prediction, focusing on their advantages and disadvantages, and propose guidelines for their development

This work arose from a strong collaboration between Cima Universidad de Navarra, Tecnun - Universidad de Navarra, Medical Research Institute La Fe (IIS La Fe), and Stanford University, bringing together expertise in machine learning, bioinformatics, and drug discovery.

💡Key contributions:

✅ Comprehensive benchmarking of current models, showing that inductive graph learning enables better generalization for drug repurposing.

🔬 A biologically driven negative-edge subsampling method, uncovering previously unknown interactions, validated in vitro.

🛠️ A standardized benchmarking toolbox, ensuring reproducibility and model fairness, contributing to future model design.

This work would not have been possible without the dedication and expertise of our outstanding team! 🙌

A special thanks to Jesús de la Fuente and Guillermo Serrano, as well as the PIs of this work, Olivier Gevaert and Mikel Hernaez, for their guidance and support.