Machine Learning for Molecule Discovery
Mission
To accelerate molecular discovery using machine learning and AI, with the goal of tackling global challenges like disease and climate change and promoting collaboration between AI and molecular science experts.
Highlights
Over the past four years, the ML4Molecules ELLIS program has emerged as a leading initiative at the intersection of machine learning, chemistry, and molecular sciences, fostering a vibrant community of researchers dedicated to accelerating molecular discovery through AI. The workshops have consistently brought together experts from academia and industry to tackle fundamental challenges such as molecular representation learning, generative modeling, reaction prediction, and property estimation. From the foundational discussions in 2021 to the integration of large-scale models and multi-modal learning in 2024, the program has evolved to reflect the rapid advancements in both machine learning and molecular science.
A central ambition of the ML4Molecules initiative has been to bridge disciplinary gaps, promote open science, and critically assess the applicability of ML methods in real-world chemical and pharmaceutical contexts. Each workshop has emphasized rigorous benchmarking, interpretability, and data efficiency, while also exploring cutting-edge innovations like foundation models, diffusion-based generation, and LLMs for chemistry. By fostering collaboration across domains and encouraging reproducible research, the program continues to shape the future of AI-driven molecular discovery, with a strong commitment to both scientific excellence and societal impact.
Collaborations
The Machine Learning for Molecule Discovery program maintains active collaborations with other ELLIS programs such as ELLIS Health and Geometric Deep Learning to share methodologies and co-organize events. It is closely connected to institutions including the ELLIS units in Cambridge, Berlin, Linz, and Zürich, and includes experts from industry, including organizations like Novartis, AstraZeneca, Microsoft Research, and the Fritz-Haber-Institut.
Notes (Other Media)
ML4Molecules Workshop 2021: Website
ML4Molecules Workshop 2022: Website
ML4Molecules Workshop 2023: Website
ML4Molecules Workshop 2024: Website
