Single-cell ‘omics technologies have the potential to revolutionize our understanding of stochasticity and heterogeneity in biology, yet such measurements are inevitably affected by high levels of noise and technical artifacts. To distinguish genuine biological variability from confounding factors, it is therefore essential to adopt analysis methodologies that model such noisy effects. In this review, we discuss model-based approaches that tackle this problem within the framework of Bayesian statistics. We start by revisiting the fundamental concepts and illustrate how they are used in a number of single-cell RNA sequencing analyses, highlighting the merits and still unmet challenges within this expanding area of research.
