Coupled biochemical reaction processes such as genetic regulatory networks display a wide range of nonlinear dynamics as well as complex adaptive responses. The components of such reactions can increasingly be engineered for use in vivo or in vitro for applications such as the control of living cells, sensing, and the design of biomaterial robots. While we have a range of methods for designing and validating the behavior of individual components and for constructing mechanistic models of these reaction networks, the resulting models often have critical gaps that emerge when designing networks, and current state of the art is the use of traditional hypothesis-based identification of these gaps and then hand-designing experiments to address them. Yet these systems are strong candidates for both automated experimentation and analysis as well as for the use of generative design. I will describe different classes of synthetic biochemical reaction networks, their applications, how these networks can be physically segregated into synthetic cells to create reaction-diffusion processes,and how networks can be coupled to other processes such as mechanical motion to build robots, self-assembly processes and regulate chemical synthesis. I will then describe our current methods for modeling and designing these systems as well as initial efforts toward different types of data-driven modeling and automated design and describe challenges and opportunities for thinking about the design of large-scale autonomous information processing systems and physical control systems using these chemical components.
Back to Workshop IV: Modeling Multi-Scale Collective Intelligences
Copyright. All Rights Reserved.