the last One World ABC Seminar of the season is behind the corner!

On Thursday, June 29, at 9.30am UK time, Jonas Arruda and Yannik Shälte (University of Bonn) will talk about "Amortized simulation-based inference for non-linear mixed-effects models" with an reported below.

Abstract: Non-linear mixed-effects models are a powerful tool for studying heterogeneous populations in biology, medicine, engineering, and other fields. However, fitting these models is computationally challenging when the description of individuals is complex and the population is large. To address this issue, we propose the use of neural density estimation to approximate entity-specific posterior distributions in an amortized fashion. This approximation is then used to efficiently infer population-level parameters. Applying this approach to problems from single-cell biology, we demonstrate its scalability to large datasets in an unprecedented manner. Moreover, it enables accurate uncertainty quantification, simple testing of hypotheses, and extends to stochastic models, going beyond what established methods can give. In summary, our method constitutes a highly flexible novel approach has the potential to substantially improve the analysis of heterogeneous populations.