D3net: Driving drug design with deep multimodal and multi-task learning

Term: 07/2019 – 06/2022 (36 Monate)

Partner: UCB Biopharma sprl

In this project, UCB Biopharma and the Institute for Machine Learning at Johannes Kepler University Linz team up to investigate, assess and enable the use of Deep Learning-based AI methods in various stages of drug design. Deep Learning methods have recently led to a new state-of-the-art performance for bioactivity and toxicity models, enabled generative models for molecules, and even paved the way for AI-based synthetic route planning. This project aims at advancing the technologies further to increase the efficiency & effectiveness of drug discovery, by decreasing toxic effects such as drug-induced liver injury & cardiotox, while assisting designers in further improving the design of compounds.