Inferring cell division kinetics in Chlamydomonas reinhardtii from flow cytometry with Gaussian process regression

Green algae, such as Chlamydomonas reinhardtii, are promising candidates for various industrial applications. For microorganisms, cell size relates to metabolic activity and cellular composition, among others, and is thus a highly relevant property for bioprocess operations. Besides time-consuming microscopy measurements, cell sizes can be estimated indirectly for example with flow cytometry based on light scattering, resulting in a measured size distribution. However, calibration for absolute cell sizes is obstructed by the optical properties of the involved particles, resulting in less accurate size distributions, which in turn can hinder applications such as model development or process monitoring. In this study, a novel approach is proposed to improve the estimation accuracy of the cell size distribution by utilizing a physical model on light scattering around a sphere. Benchmarked with microscopy image analysis, the model shows substantial improvement. Using the corrected size distribution, the cell division rate is inferred with an extended Gaussian process regression used upon a population balance model. The resulting model is able to accurately describe the observed size distribution with the estimated division kinetics. The approach is tested using data obtained from the cultivation of C. reinhardtii as a model organism. The results provide mechanistic insight into C. reinhardtii cell division and cell size heterogeneity.