Probability seminar: Investigating cell-to-cell variability with Bayesian model selection and approximate likelihood
SM3, School of Mathematics
Sarah Filippi, University of Oxford
At the molecular level every cell is unique and differences between cells of the same type can have profound biomedical implications. The mechanisms driving this variability can be divided into two classes: within-cell variability arising from stochasticity in the biochemical reactions and cell-to-cell fluctuations in biochemical reaction rates. Given the growing abundance of ’omics data resolved at the single-cell level, it is becoming increasingly important to include these sources of noise in mathematical models. Here we present a general modelling and Bayesian inference methodology for single-cell data that explicitly takes into account both intrinsic and extrinsic noise. Using quantitative image cytometry (QIC) single-cell proteomics data, we apply our methodology in order to study MEK/ERK phosphorylation dynamics.