Identification of stochastic models to predict single-cell gene regulation dynamics

Abstract: Stochastic fluctuations can cause identical cells or individual molecules to exhibit wildly different behaviors. Often labeled “noise,” these fluctuations are frequently considered a nuisance that compromises cellular responses, complicates modeling, makes predictive understanding and control all but impossible. However, if we computationally examine fluctuations more closely and carefully match them to discrete stochastic analyses, we discover virtually untapped, yet powerful sources of information and new opportunities. In this talk, I will present our collaborative endeavors to integrate single-cell and single-molecule experiments with precise stochastic analyses to gain new insight and quantitatively predictive understanding for signal-activated gene regulation. I will explain how we experimentally quantify transcription dynamics at high temporal and spatial resolutions; how we use precise computational analyses to model this data and efficiently infer biological mechanisms and parameters; how we predict and evaluate the extent to which model constraints (i.e., data) and uncertainty (i.e., model complexity) contribute to our understanding. We will examine how different data statistics (e.g., expectation values versus probability densities) contribute to model bias and uncertainty, and we will show how these affect predictive power. Finally, we will introduce a new approach to compute the Fisher Information Matrix, and we will illustrate its application for the improved design of single-cell experiments.

Bio: Dr. Munsky received B.S. and M.S. degrees in Aerospace Engineering from the Pennsylvania State University in 2000 and 2002, respectively, and his Ph.D. in Mechanical Engineering from the University of California at Santa Barbara in 2008. Following his graduate studies, Dr. Munsky worked at the Los Alamos National Laboratory — as a Director’s Postdoctoral Fellow (2008-2010), as a Richard P. Feynman Distinguished Postdoctoral Fellow in Theory and Computing (2010-2013), and as a Staff Scientist (2013). In 2014, he joined the Colorado State University Department of Chemical and Biological Engineering and the School of Biomedical Engineering, in which he is now an Assistant Professor. Dr. Munsky is best known for his discovery of Finite State Projection algorithm, which has enabled the efficient study of probability distribution dynamics for stochastic gene regulatory networks. Dr. Munsky’s research interests are in the integration of discrete stochastic models with single-cell experiments to identify predictive models of gene regulatory systems. Dr. Munsky was the recipient of the 2008 UCSB Department of Mechanical Engineering best Ph.D. Dissertation award, the 2010 Leon Heller Postdoctoral Publication Prize, and the 2012 LANL Postdoc Distinguished Performance Award for his work in this topic. Dr. Munsky became a Keck Scholar in 2016. Dr. Munsky is the contact organizer of the internationally recognized, NIH-funded q-bio Summer School (q-bio.org), where he runs a course on single-cell stochastic gene regulation.

Hosted by: Paul Bogdan