Abstract

Single-cell and single-molecule experimental techniques expose the randomness of cellular processes and invite a stochastic description. In this talk I will present our efforts to solve inverse problems related to stochastic cellular dynamics. First, we provide a inference framework that accounts for extrinsic and intrinsic noise contributions present in single-cell measurements. For that, we show that stochastic components of a cellular process can be marginalised exactly such that the inference remains tractable. Second, we present single-molecule experimental data to study transcriptional kinetics in live yeast cells. A stochastic models for the system is presented and biophysical parameters such elongation speed, termination rate etc are inferred from single transcription-site intensities. Moreover, optimal filtering or state estimation is performed to reconstruct the most likely position of single RNAP molecules on the gene.