Scalable and flexible inference framework for stochastic dynamic single-cell models

Sebastian Persson, Niek Welkenhuysen, Sviatlana Shashkova, Samuel Wiqvist, Patrick Reith, Gregor W. Schmidt, Umberto Picchini, Marija Cvijovic

Abstract: Author summary Understanding the causes of heterogeneity and the means by which it can be controlled is crucial for manipulating cellular populations and treating diseases. To this end, single-cell time-lapse microscopy data is often combined with dynamic modelling. However, the construction of mechanistic models requires the ability to infer unknown model quantities from data, while simultaneously accounting for intrinsic and extrinsic noise. Here we propose a Bayesian inference framework which enabled us to elucidate sources of cell-to-cell variability in yeast signalling and provides deeper insights into the causes and consequences of heterogeneity. Our approach is versatile and can for example further be applied in pharmacokinetic and pharmacodynamic studies, epidemic studies, as well when modelling the behaviour of cancer cell populations.