Using gene expression noise to understand gene regulation

Back to The Sixth q-bio Conference.

Brian Munsky and Gregor Neuert

An important goal of systems biology is to understand and predict how cells sense and respond to their environments in the presence of biochemical noise. Although recent studies have revealed many components in these signal transduction and gene regulation pathways, it remains difficult to predict the phenotypic diversity of cellular dynamics. This tutorial will introduce a comprehensive approach to identify and validate gene regulation models using dynamic single-cell/single-molecule experiments and stochastic analyses. This integrated approach automatically generates hypotheses, proposes optimal experiments, and discards, validates or refines existing hypotheses. We show how to automatically generate thousands of hypotheses, and then systematically invalidate these hypotheses until we identify a single predictive model of gene regulation. We illustrate the approach on an example regulatory path in yeast, for which we uncover a single predictive model. We validate the final model with quantitative predictions at diverse environmental and genetic conditions. Our integrated experimental and computational approach is extremely general and applicable to any signal transduction and gene regulation pathway in bacteria, yeast, or human cells.