Analytical representation of protein distributions in stochastic models of gene expression

University Of Massachusetts Boston, Dorchester MA

Investigators

Abstract

One of the fundamental problems in biology is elucidating the molecular mechanisms that give rise to phenotypic variations among individuals in a population. It has been shown that phenotypic variations can arise without any underlying differences in the genotype or environmental factors. Such "nongenetic individuality" is observed in diverse cellular processes, ranging from bacterial persistence to HIV-1 viral infections, and is driven by randomness (noise) in the cellular levels of gene expression products such as mRNAs and proteins. To quantify the effects of noise in gene expression, recent single-cell experiments have obtained probability distributions characterizing protein levels across a population of cells. Correspondingly, there is a need to develop a general analytical framework for modeling and interpretation of the distributions obtained from such single-cell experiments that will lead to quantitative insights into how noise in gene expression is regulated. The goal of this project is to develop new approaches for obtaining analytical results for protein distributions in models of gene expression and its regulation. The project will contribute to a fundamental understanding of the role of noise in gene expression and its regulation in diverse cellular processes. The analysis requires tools and approaches from physics and applied mathematics and will be integrated with teaching efforts to effectively train students and future scientists in this fast-developing field of interdisciplinary research. Noise in gene expression is generally analyzed using coarse-grained stochastic models. However, obtaining exact analytical expressions for the corresponding protein distributions has been a challenging for all but the simplest models. The project research will involve integration of tools from queueing theory with approaches based on partitioning of Poisson processes to address such challenges. The approaches developed will be used to obtain both exact and approximate analytical results for models of gene expression and its regulation. The resulting analysis of models with regulatory mechanisms such as bursting, feedback and promoter-based regulation will lead to quantitative insights into noise characteristics of the basic building blocks of genetic circuits. The project research will also lead to analytical results that characterize the role of noisy inputs in regulating simple biochemical switches and suggest new approaches for estimating model parameters based on observations of noise. The analytical results derived in the project will have multiple applications ranging from synthetic biology to understanding phenotypic variation in clonal populations.