Summer School
From (foundational) interatomic potentials to generative AI
Funded by COST Action DAEMON
ML4Chem is a three-day school on machine learning methods for chemistry and materials science, held 3–5 August 2026 at Leipzig University. It is aimed at PhD students, postdocs, and early-career researchers in computational chemistry, materials science, and related fields.
The scientific core is four sessions, one per half-day. Each session opens with a 15-minute contributed talk from a participant, continues with a lecture by the invited speaker, and closes with a two-hour tutorial run by the same speaker.
Equivariant architectures, foundational models, and fine-tuning strategies for accurate and scalable molecular simulations.
Machine learning approaches for excited states, photodynamics, and nonadiabatic processes.
Hybrid quantum mechanics/molecular mechanics methods enhanced by machine learning for large-scale, condensed-phase, and biological systems.
Generative AI for the inverse design of molecules and materials with targeted properties.
The school closes on Wednesday morning with a career round table: Prof. Dr. Leticia González (University of Vienna), Prof. Dr. Reinhard J. Maurer (University of Vienna), Jun.-Prof. Dr. Leonie Weissweiler (Leipzig University / ScaDS.AI), and Daniel Zügner (Microsoft Research) take questions on research careers inside and outside academia, followed by closing remarks.
Each session is taught by the people who lead it: the invited speaker gives the lecture, and the speaker or a member of their group runs the hands-on tutorial. Wednesday’s round-table guests come from two different directions: three hold professorships, one does research in industry.
Csányi Group, Department of Engineering, University of Cambridge
Professor of AI in Theoretical Chemistry, Leipzig University
Professor of Molecular Modeling and Simulation, BOKU University, Vienna
Institute of Molecular Modeling and Simulation, BOKU University, Vienna
Machine Learning Group, TU Berlin and BIFOLD
Machine Learning Group, TU Berlin and BIFOLD
Professor of Theoretical Chemistry, University of Vienna
Professor of Computational Materials Discovery, University of Vienna
Junior Professor of Natural Language Processing, Leipzig University / ScaDS.AI
Senior Researcher, Microsoft Research, Berlin
On Monday and Tuesday the program repeats one rhythm twice a day: contributed talk, lecture, coffee, tutorial. Wednesday belongs to the career round table. All lectures and tutorials take place in the Experimental Lecture Hall (Experimentalhörsaal) of the Faculty of Chemistry; the poster session is held in the Freiarbeitsbereich.
ML-accelerated adsorption dynamics on semiconductor surfaces
Hendrik Bergner-Weiske, Leipzig University
Vibrational Fingerprints of Ultrafast Spin Crossover in an Fe(III) Complex
Dilara Farkhutdinova, University of Vienna
Incorporating long-range interactions via the multipole expansion into ground and excited-state molecular simulations
Rhyan Barrett, Leipzig University
Generative Inverse Design of Tetrathiafulvalene-Based Molecular Materials for Organic Electronics
Patrícia Ramos, Technical University of Porto
Nonadiabatic Excited-State Pathways from SOCT-ISC to Hydration-Driven Proton Transfer: A Trajectory-Based Perspective
Moumita Banerjee, Linköping University
ML vs. expert intuition for AD-HoC Nickel Cross-Coupling: Genetic algorithms coupled with regression algorithms to learn from heterogeneous experimental data
Daniel Bitterlich, Leipzig University
Gas-Phase Vibrational Spectroscopy of PF6−(H2O)0–18: From Isolated Clusters to the Condensed Phase
Hannah Buttkus, Leipzig University
Toward Closed-Loop Design of Enantioselective N-Heterocyclic Carbene Catalysts for the Benzoin Reaction
Shuai Chen, University of Groningen
A High-Throughput Generative Workflow for Data-Driven Reaction Environment Optimization
Robin Curth, Leipzig University
Combining genetic algorithms with omits in kinematical refinements of cRED data: pathways towards higher accuracy
Alexander Feige, Leipzig University
Machine learning for smell: ordinal odor strength prediction of molecular perfumery components
Peter Fichtelmann, Leipzig University
First principles construction of Newns-Anderson Hamiltonians for hydrogen chemistry at surfaces
Nils Hertl, University of Warwick
The Solvation Line: Mapping Close-Range Interactions in Reaction Mechanisms
Federica Lauria, University of Turin
Dataset Design for Buffer Region Neural Networks: from Sampling to Simulation
Bettina Lier, BOKU University, Vienna
Mixed Metal Oxides in the Gas Phase: The Study of the FeAlO3+ and Al2O3+ Clusters
Lucas W. de Lima, Humboldt-Universität zu Berlin
E(3)-VERDE: A 3D-Equivariant Neural Network for the Complete Ground- and Excited-State Redox Landscape of Organic Photocatalysts
Andres Lopez, Northeastern University
High-Precision Modeling of Nuclear Quantum Effects at Low Temperatures in CO/NaCl(100) Adsorption Using MACE Potentials and Path-Integral Methods
Mateo Pescador Arboleda, University of Potsdam
Quantum nuclear and band-dispersion effects recover near-UV absorption in short-hydrogen-bonded organic crystals
Muhammad Nawaz Qaisrani, Technische Universität Ilmenau
Basin-Hopping Monte Carlo with Transferable Neural Network Potentials: Mapping Germanium and Aluminium Distributions in Zeolites
Indranil Saha, RWTH Aachen University
Investigating ZnPd nanoparticles supported on ZnO using atomistic simulations with machine learning potentials
Roy Schumacher, Forschungszentrum Jülich
Restricted or unrestricted? The boundaries of universal MLIPs for organic photochemistry
Henrik Seng, ETH Zürich
Lighting up CO2 reduction: are organic photosensitizers paving the way to metal-free artificial photosynthesis?
Lucrezia Villa, TU Wien
Exploring Excited-State Energy Landscapes Across Chemical Compound Space with Machine Learning: The Example of Green Fluorescent Protein Chromophores
Sophia Wesely, Leipzig University
From Analytical Microchips towards Self-Driving Chemistry – Experimental Challenges in the Age of Digitalisation
Hannes Westphal, Leipzig University
Enhancing Product Specificity of Chalcone Synthase Using Natural Deep Eutectic Solvents
Mohammad Yousefi, Leibniz Institute of Plant Biochemistry, Halle
Another diamine cation challenges quantum chemistry methodology
Neda Zolghadri Jahromi, Nicolaus Copernicus University in Toruń
Assessing the performance of quantum-mechanical descriptors in physicochemical and biological property prediction
Alejandra Irene Hinostroza Caldas, TU Dresden
Johannisallee 29, 04103 Leipzig, Germany
Accommodation recommendations will be shared soon.
Johannisallee 29, 04103 Leipzig
Get directions →Philipp-Rosenthal-Straße 31, 04103 Leipzig
Get directions →The Destille is a short walk from the Faculty of Chemistry, just around the corner on Philipp-Rosenthal-Straße.
Registration is closed.
Grants for travel and accommodation, funded through COST Action DAEMON, are allocated by the organizing committee. Allocation is not guaranteed; applicants are notified directly.
The Tuesday BBQ at the Destille is an optional add-on: €30, paid at on-site registration.
Personal data is processed in line with the GDPR. Details are shown at the top of the registration form.