Filippo Menolascina, Ph.D.
Professor at the University of Edinburgh
Director of Research at the Bayes Centre, the University's Innovation Hub for Artificial Intelligence
In this talk, I will present a computational and experimental framework to automate modelling in synthetic biology. Leveraging principles from System Identification, Real-Time Control, Optimal Experimental Design and in vivo experiments in microfluidics, I will show how we achieved automatic model selection and model calibration in both a frequentist and Bayesian framework in E. coli and S. cerevisiae. Given a set experimental budget (e.g., experiment hours), our results show that online Optimal Experimental Design can reduce parameter estimation errors by 80% compared to traditional, intuition-driven inputs. I will briefly discuss how our activities in the space of lab automation led to spinning out OGI Bio, an Edinburgh-based start-up focusing on smart bioreactors.