Application of Neural Ordinary Differential Equations in Pharmacometrics

Machine learning (ML) is playing an increasingly important role in pharmacometric (PMX) modeling. Among the many approaches, Neural Ordinary Differential Equations (NODEs) have attracted special attention because they combine two powerful concepts: ordinary differential equations (ODEs), a standard tool in PMX, and neural networks (NNs).

In this workshop, participants will gain the theoretical background and hands-on experience with ML in PMX. In the first part, we will introduce the principles of NNs and NODEs in PMX and participants will gain hands-on experience applying them to pharmacokinetic data in R, exploring their strengths and limitations. In the second part, the focus will lay on low-dimensional NODEs, a special type of NODEs standing out for its comprehensibility and ease of use, even in common PMX software and in combination with mixed-effects approaches. Participants will be able to apply them in Monolix and in the framework of scientific ML, combining conventional compartmental modelling with NNs.

Instructors: Dominic Bräm and Gilbert Koch, University Children’s Hospital Basel, Switzerland

Dr. Dominic Bräm has a background in pharmaceutical sciences with a specialization in PMX and ML. He is a PostDoc at the University Children’s Hospital Basel in Switzerland. His research focus lies on the integration of ML into PMX and the development of ML-based automated model development algorithms. Dr. Bräm has developed two R-packages, the first one facilitating the integration of NODEs in PMX software, and the second one for optimal dose estimation.

Dr. Gilbert Koch is a mathematician with over 15 years of experience in PMX modeling. He is a Senior Research Fellow at the University Children’s Hospital Basel in Switzerland and an Adjunct Assistant Professor in the Department of Pharmaceutical Sciences at the State University of New York at Buffalo. His research focuses on developing clinical decision support systems and advancing PMX methodology, including the integration of PMX with artificial intelligence. Dr. Koch has made significant contributions to the PMX community, authoring more than 50 scientific publications.

The workshop is organized on behalf of the graduate programme in pharmaceutical sciences, Drug Research Academy and PopSim (Population Simulation Modelling section) under the Danish Pharmaceutical Society, by Associate Professor Trine Meldgaard Lund, Department of Drug Design and Pharmacology, Faculty of Health and Medical Sciences, University of Copenhagen.

Pre-registration is required: REGISTRATION

The workshop is free of charge and open to attendance by all interested parties. However, PhD students will be given priority. There is a limited capacity of 30 participants in the workshop. Please register before 13 October 2025, after which date confirmation on attendance will be sent out together with instructions on installation of R and Monolix.