Researcher in Chemical University of Basel
ETH Zurich
Switzerland

The integration of deep learning with chemical reaction prediction is a fast-evolving field with the potential to revolutionize small molecule drug discovery and design. This project aims to develop advanced machine learning methods for chemical reaction prediction, focusing on condition, yield and binary prediction, substrate screening, forward reaction prediction for compound library design, reaction classification, and multi-step reaction prediction. The successful candidate will enhance existing workflows in active learning, graph neural networks, and language models to address these challenges.

The Postdoc will work closely at the interface of various teams in Small Molecule Research at pRED, including their Computer-Aided Drug Design group, the High-Throughput Experimentation laboratory in Medicinal Chemistry, and Compound Library Enhancement and Logistics (CLEL). The developed models will be integrated with existing CADD workflows, such as property and bioactivity prediction, and applied to compound optimization case studies. This collaborative setting between the ETH and pRED provides an ideal environment for contributing to both theoretical advancements and practical applications in drug discovery.

The Computer-Assisted Drug Design (CADD) group of the Department of Biosystems Science and Engineering (D-BSSE) at ETH Zürich, is seeking a postdoctoral researcher to work in tight collaboration with Pharma Research and Early Development (pRED) at Roche Innovation Center Basel. The CADD group focuses on developing innovative concepts, algorithms, and software for rapid identification of bioactive tool compounds and pharmaceutical lead structures. Our research leverages AI for drug discovery and design, advancing machine learning methods to tackle diverse challenges in medicinal and bioorganic chemistry.

As a Postdoc, you will work simultaneously at the D-BSSE located in Basel and at the Roche Innovation Center Basel. Your focus will be on developing and refining deep learning models for reaction prediction within the context of small molecule drug discovery. Your responsibilities include data curation, designing and developing computational methods such as active learning, graph neural networks, language models, and quantum chemistry for condition prediction, substrate prediction, reaction design and classification, yield and binary prediction, substrate screening, and forward and multi-step reaction prediction for library design. This work will be closely coordinated with members of the CADD, Medicinal Chemistry, and CLEL groups to integrate the models into existing workflows.

We are seeking a highly motivated and technically skilled candidate with the following qualifications:

• A PhD in chemistry, chemical engineering, computational science, physics, or related fields, or the expectation of obtaining such a degree before Q1 2025.
• Fluency in Python and proficiency with cheminformatics tools such as RDKit and OpenBabel.
• Experience in designing and developing deep learning models in PyTorch, demonstrated through at least one public GitHub repository and related first author publication.
• A strong understanding of chemical reactions and synthesis.
• Familiarity with chemical databases and molecular modeling.
• The ability to work independently and effectively in an international team at the intersection of academia and industry.
• Excellent communication skills in English (both written and spoken).

Desirable criteria:

• Familiarity with graph neural networks (e.g., with PyTorch Geometric) and/or chemical language models (RNNs, Transformers).
• Experience with quantum-chemical simulations (e.g., Psi4, GFN2-xTB).
• Know-how of graphical user interface development.

This position offers a unique opportunity to work at the intersection of machine learning and drug discovery, contributing to cutting-edge research in a collaborative and interdisciplinary environment.

We look forward to receiving your online application with the following documents:

• a cover letter stating your motivation for this position
• a comprehensive CV including publications and presentations
• diplomas and a list of courses attended (transcript)
• at least one letter of reference

Please note that we exclusively accept applications submitted through our online application portal. Applications via email or postal services will not be considered.

For further information please contact Dr. Jan Hiss, jan.hiss@pharma.ethz.ch (no applications!), and visit our website.