Skip to the content.

Discovering new molecules with desired functions or activities is crucial for human well-being by providing new medicines, securing the world’s food supply via agrochemicals, or enabling a sustainable energy conversion and storage to counter or mitigate climate change. However, the discovery of new molecules or molecular materials that are optimized for a particular purpose can often take up to a decade and is highly cost-intensive. Machine-learning (ML) methods can accelerate molecular discovery, which is of considerable importance generally, but especially in light of the COVID-19 crisis and future pandemics. To reach this goal of speeding up the discovery of new functional molecules, it is necessary to establish a dialogue between domain experts and ML researchers to ensure that ML positively impacts real world scenarios. The importance of this field has been acknowledged also by Stanford’s 2021 Artificial Intelligence index report which states that “Drugs, Cancer, Molecular, Drug Discovery” received the greatest amount of private AI investment in 2020, with more than USD 13.8 billion, 4.5 times higher than 2019. In this workshop, we are bringing together the expertise of excellent researchers in the field of ML and its applications to molecular discovery.

The workshop will be hosted by the ELLIS unit Cambridge and ELLIS unit Linz as a side-event to NeurIPS2021.

Schedule (CET time)

The workshop will take place on December 13th, 2021 via Zoom. We invite discussions and questions for the presenters on Discord.

Speaker Title
9:00 Opening remarks ML4Molecules Program Chairs
9:15-10:30 Workshop talks Session chair: Marwin Segler Ola Engkvist, AstraZeneca, Sweden AI for Drug Design, a personal perspective
Andrea Volkamer, Charite Berlin, Germany ML Models to Support Risk Assessment of Small Molecules
Johannes Margraf, Fritz Haber Institute Berlin, Germany Chemical ML beyond established benchmark datasets
Günter Klambauer, University of Linz, Austria Moving beyond narrow AIs in Drug Discovery -- a perspective PDF
Sereina Riniker, ETH Zurich, Switzerland Learning chemistry in the limit of finite training sets
Jose Miguel Hernandez Lobato, University of Cambridge, UK Bootstrap your Flow for Improved Sampling from Boltzmann Distributions
Francesca Grisoni, TU Eindhoven, Netherlands De novo design with chemical language models: between theory and experiments
10:30-11:30 Invited Talks
Session chair: Andrea Volkamer
Paula Marin Zapata, Bayer Pharmaceuticals, Germany Using cell morphology to guide de-novo hit design
Greg Landrum, ETH Zurich, Switzerland Some thoughts on the molecule part of Machine learning for molecule discovery
Camille Marini, Owkin, France MELLODDY: Accelerating Drug Discovery by Competitive Cooperation using privacy preserving federated learning
11:30-12:30 Poster session 1 Gathertown
12:30-14:30 Break
14:30-15:30 Workshop talks
Session Chair: Nadine Schneider
Michele Ceriotti, EPFL, Switzerland Machine learning for chemistry, beyond potentials
Cecilia Clementi, Freie Universität Berlin, Germany Designing protein models with machine learning and experimental data
Fernanda Duarte, University of Oxford, UK Efficient ML strategies to explore chemical reactivity
Gábor Csányi, University of Cambridge, UK Force fields using equivariant polynomials: a unique combination of accuracy and speed
Wengong Jin, MIT, USA Iterative refinement graph neural network for antibody sequence-structure co-design
Megan Stanley, Microsoft Research, UK Molecular Property Prediction as a Few-Shot Learning Problem
15:30-16:00 Contributed talks
Session Chair: Nadine Schneider
Andreu Vall BioassayCLR: Prediction of biological activity for novel bioassays based on rich textual descriptions
Yuanqi Du Interpreting Molecular Space with Deep Generative Models
Hannes Stärk 3D Infomax improves GNNs for Molecular Property Prediction
16:00-17:00 Invited talks
Session Chair: Jennifer Wei
Daniel Reker, Duke University, USA Active Molecular Machine Learning
Philippe Schwaller, EPFL, Switzerland Learning the Language of Chemical Reactions using Transformers
Heather Kulik, MIT, USA Putting DFT to the test in machine-learning-accelerated materials discovery
17:00-17:30 Panel discussion
Moderator: Marwin Segler
Ola Engkvist,
Bharath Ramsundar,
Jose Miguel Hernandez Lobato,
Francesca Grisoni
Current challenges and opportunities in Machine Learning for Molecule Discovery
18:00-19:00 Poster session 2 Gathertown
19:00-20:00 Break
20:00-21:30 Contributed talks
Session chair: Günter Klambauer
Miguel Garcia Ortegon DOCKSTRING, a better yardstick for ligand discovery
Loďc Dassi Kwate Identification of Enzymatic Active Sites with Unsupervised Language Modeling
Cristian Bodnar Weisfeiler and Lehman Go Cellular: CW Networks
Octavian-Eugen Ganea Independent SE(3)-Equivariant Models for End-to-End Rigid Protein Docking
Kexin Huang Benchmarking Molecular Machine Learning in Therapeutics Data Commons
Krzysztof Maziarz Learning to Extend Molecular Scaffolds with Structural Motifs
Aryan Deshwal Combining Latent Space and Structured Kernels for Bayesian Optimization over Combinatorial Spaces
Invited talk Tess Smidt, MIT, USA Unexpected properties of symmetry equivariant neural networks
21:30-21:45 Closing remarks


Please register for the event on Eventbrite: REGISTER

Important dates

Accepted contributions (oral):

Accepted contributions (poster):

Organizing Committee and Contact

Chairs: Jose Miguel Hernandez Lobato, Nadine Schneider, Günter Klambauer;
Michele Ceriotti, Cecilia Clementi, Connor Coley, Ola Engkvist, Francesca Grisoni, Sepp Hochreiter, Matt Kusner, Brooks Paige, Bharath Ramsundar, Sereina Riniker, Marwin Segler, Andrea Volkamer, Jennifer Wei.