Hackathon on predictive modelling in systems pharmacology: design, implementation, and outcomes

Abstract

The growing availability of multi-omics data and the rise of network-based and machine-learning methods have created both an opportunity and a challenge for the systems pharmacology community: the opportunity to model drug action at unprecedented mechanistic resolution at different scales, and the challenge of equipping a new generation of researchers with the computational literacy to do so. To address this need, we organised a 4-day intensive hackathon on Predictive Modelling in Systems Pharmacology under the Horizon Europe CANVAS and French National DIGPHAT consortia. The event brought together 16 early-career researchers spanning biology, bioinformatics, and computational science, with the dual aim of fostering a multidisciplinary consortium community and providing hands-on training that bridges classical pharmacometrics with omics data science. Three scientific tracks were pursued: (i) pharmacokinetic/ pharmacodynamic modelling, (ii) drug perturbation modelling through network diffusion in gene regulatory networks, and (iii) machine learning for pharmacogenomics. A theory-to-practice pedagogical framework structured each day, with morning workshops grounding participants in methodological foundations and afternoon coding sprints applied to real-world datasets. Here we describe the scientific rationale, datasets, methodological frameworks, and outcomes for each track, along with the pedagogical design choices, technical infrastructure, and qualitative participant feedback. We propose that this format focuses on interdisciplinary community-building and constitutes a replicable model for computational systems pharmacology training at the European level.