Bioprocess Optimization: Active and Reinforcement Learning for bioprocess design and optimization

Date: September 2025
Duration: 1 month
Home laboratory: Systems & Synthetic Biology (SSB), Bioprocess Engineering (BPE) chair groups - Wageningen University & Research (WUR), The Netherlands
Host laboratory: Laboratoire BRS (Bio-RetroSynth), Institu MICALIS, INRAE, France

This project is aligned with the main goal of the B-BEST call on the efficient development of a bio-based industry and economy. We are developing a computational framework for bioprocess design to simulate the performance of a microbial cell factory under different reactor operating modes and to find the ideal values for process variables. This tool relies on Bayesian parameter estimation and Design of Experiments approaches. However, other datadriven techniques to both explore and exploit larger parameter spaces are required. The purpose of my mobility is to incorporate Active and Reinforcement Learning methods developed in Dr. Jean-Loup Faulon’s lab into the framework described. The use of these approaches would facilitate addressing a large solution space to significantly accelerate the finding and exploitation of values for variables in process design.