Gaining insight into the operation of biochemical reaction networks in the cell is an important problem. Understanding the mechanisms by which biochemical networks function will help us develop the design principles to systematically build synthetic biochemical networks. Unfortunately, the dynamics of processes inside cells cannot be observed directly, limiting our ability to design and analyze biochemical networks. We will investigate problems related to estimation and observation in stochastic biochemical networks using dynamic data generated by single cells. Our approach to the problems of state estimation and parameter estimation is based on the theory of stochastic chemical reaction networks, applicable to systems observed through experimental methods like time-lapse microscopy in which stochastic phenomena dominate.
Intellectual Merit
The PI proposes to test his approach experimentally by applying the theory to synthetic gene regulatory networks that we construct in E. coli and observe with time-lapse microscopy. Success will lead to the development of new tools for evaluating time-lapse microscopy data and motivate new techniques for experiment design. The results will impact research in systems and synthetic biology by introducing new methods for verifying the performance of engineered networks.
Broader Impact
The PIs will develop an educational module on the quantitative techniques used in this research and integrate it into the already existing three-course sequence on systems and synthetic biology offered at the University of Washington. In addition, the PIs will work with several University of Washington programs that promote diversity by offering research opportunities to under-represented students.
Developing an understanding of biological phenomena through modeling requires a notion of a state that captures the essential components of the system and a model that describes its essential functions. When a collection of cells is considered in aggregate, measurement noise is usually primarily responsible for complicating the problem of identifying state and model parameters in genetic networks. At the single-cell level, the presence of cellular variability in experimental data introduces systemic noise that further complicates this problem. The stochastic phenomenon of systemic noise in individual cells can be detected by observing the variation that occurs during the growth the of isogenic colonies observed using time-lapse microscopy. The movies produced by these methods do not provide a full measurement of the system’s state but instead provide measurements of only a few species, such as fluorescing proteins, and these data are corrupted by measurement noise. In our work we developed (1) new synthetic strains of bacteria and yeast so that certain proteins in genetic regulatory networks we constructed are observable under the microscope using fluorecence imaging; (2) methods for preparing the cells to be repeatedly imaged as they grow under the micscrosope; (3) methods for parsing the image data that result; (4) software that automatically estimates the levels invisible proteins involved in the genetic circuits in the cells. In addition to our contributions to research, we taught several senior and graduate level courses in which these approaches were used in part to teach students how to estimate the levels of proteins in cells from low-dimensional output data.
