Making accurate, quantitative predictions about biological systems improves biological understanding, increases the rate of experimental research, can inform experimental design, and provides a platform to rapidly test ideas about that particular system. However, as these biological systems are complex and fundamentally noisy, analysis becomes challenging. On the other hand, within this noise is information that can help build better, more quantitative models. Here, we focus on developing computational tools, with the help of the Finite State Projection Approach, to fit and predict single-cell RNA distributions obtained by single molecule in-situ hybridization (smFISH). To identify the system that best represents the biological reality, a variety of models can be tested for a given biological mechanism, and the best of these models can be used to make quantitative predictions. This type of analysis has been applied to the HOG – MAPK pathway in yeast. Yeast show different responses when exposed to different levels of osmotic shock. By capturing the gene expression dynamics of yeast to a particular input with smFISH, we can apply our approach and make predictions about gene expression dynamics in response to a different levels of osmotic shock.
