Published on July 21, 2026
Artificial intelligence in bioproduction: from self-adaptive sampling to gene network analysis. Application to GMP insulin production
Pharmaceutical bioproduction suffers from high rejection rates, revealing the limitations of traditional empirical approaches. Artificial intelligence opens up transformative perspectives along two complementary axes. - The first concerns intelligent monitoring. GenSensor's GenSampler performs aseptic robotic sampling whose frequency adapts in near real-time to the culture's state. Analysis of cellular flow during sampling instantly provides density and morphology, and allows access to biological information hidden by the stress of conventional sampling. - The second axis leverages deep learning to decipher gene regulatory networks. Transcriptomic analysis identifies metabolic bottlenecks and gene modules associated with performance, transforming empirical understanding into actionable knowledge. Applied to commercial-scale insulin production, this convergence embodies Quality by Design: robustness through anticipation of deviations, rational optimization of critical parameters, and documentation meeting regulatory requirements. The real-time adaptability of monitoring, coupled with systemic intelligence, defines a new paradigm for bioprocess control.
Speakers :
Charles Hébert, CEO, GenSensor
Barbara Jacques, Responsable Réseau Polepharma
