Authors Andreas Mardt Tim Hempel Cecilia Clementi Frank Noe tackle the challenge of modeling dynamics of large protein by decomposing it into smaller independent subsystems, simultaneously modeling each subsystem’s kinetics ensuring mutual independence. go.nature.com/3FOgF4C

As neural networks scale up, the most valuable people in deep learning field today are the people who can make systems 10-100x more efficient.

You can now watch the recorded material from #NeurIPS2022 online without registration at: slideslive.com/neurips-2022

Final proceedings from the inaugural edition of the Learning on Graphs Conference are now available on PMLR! Our sincere gratitude to the program committee, authors, supporters, and research community for making LoG possible 🤗 📖: proceedings.mlr.press/v198/

We're pleased to share this new infographic on work from Marin Zapata et al., showing that phenotypic features of cell images can guide a GAN to drug candidates. Read the #openaccess article: doi.org/10.1039/D2DD00… #bioimageanalysis #deeplearning #ML #cellpainting #drugdiscovery

I will be opening two fully funded PhD positions (5 year) in my group in the near future. If you are interested in working with generative models in molecular applications (design and simulation) please feel free to reach out.

Call for reviewers and area chairs AI for Science #NeurIPS2023! Please sign up if you are interested!

LoG Conference update: a little bit over one month to go till the abstract submission deadline! OpenReview portal opens soon, and we're always looking out for reviewers who care.

The AiChemist project has officially started! This boundary-pushing #phd programme brings together experts in #AI, #chemoinformatics and #chemistry from #academia and #pharma to train future #XAI specialists in the life sciences. aichemist.eu/openpositions --> apply until 10.09!

Our work "Navigating the Design Space of Equivariant Diffusion-Based Generative Models for De Novo 3D Molecule Design" is now on arxiv: arxiv.org/abs/2309.17296 Shared co-first authorship with Tuan Le Many thanks to @ktschuett and @OkkoClev! https://bsky.app/profile/aidd.bsky.social 👇

📢Help us spread the word: We are organizing a workshop on **Optimal Transport in Berlin 2024**, connecting the theoretical and applied research, with themes in Dynamics, OPT, ML. We’ve got a list of excellent speakers! ✨Super excited! Website: sites.google.com/view/ot-berlin…

"Hallucinations", "lies", "creativity" -- too many anthropomorphisms to describe a probability distribution sampling data outside of the training data support. When we sample from a Normal density and get a 5sigma sample, we don't say it's a "hallucination", it's just a sample.

141 submissions for our workshop "Machine Learning for Life and Material Sciences" at ICML Conference on July 26! 🎉 We would love your help with reviewing 🙏. In exchange, you get a chance to see something new and exciting! 🙂 Fill in shorturl.at/b6ILv or DM me if interested

Cool new paper with Julian Cremer, Tuan Le, Okko Clevert and @ktschuett, introducing a new way to do conditional generation through diffusion models via importance sampling of the denoising pathway. Applied to ligand generation in proteins. arxiv.org/abs/2405.14925

Each week, we want to highlight Learning on Graphs Conference 2024 papers! ✨Check out "Representation Learning on Biomolecular Structures Using Equivariant Graph Attention" by Tuan Le , Frank Noe , and Djork-Arné Clevert! proceedings.mlr.press/v198/le22a/le2…

BREAKING NEWS The Royal Swedish Academy of Sciences has decided to award the 2024 #NobelPrize in Physics to John J. Hopfield and Geoffrey E. Hinton “for foundational discoveries and inventions that enable machine learning with artificial neural networks.”

After a 4-year journey, we are super happy to see this paper out in Nature Communications - Katarina Elez et al: High-throughput molecular dynamics + active machinelearning enable efficient identification of an experimentally-validated broad coronavirus inhibitor. nature.com/articles/s4146…

Attending #NeurIPS2025 in San Diego (Dec 1–7)! Happy to chat about generative modeling for structure-based drug design and proteomics—DM me to connect. Also presenting our paper “Diffusion Generative Modeling on Lie Group Representations” on Fri Dec 5, Poster 3611 (Hall C/D/E).