AF:Small:Collaborative Research:Kinetics and Thermodynamics of Chemical Computation

University Of California-Davis, Davis CA

Investigators

Abstract

Biology is replete with smart molecular systems that perform nanoscale assembly, sense environmental stimuli, create chemical signals, and produce physical motion - each of these tasks coordinated by information processing circuits implemented with chemical reactions. Learning how to build artificial chemicals that compute autonomously in complex environments would bring about groundbreaking advances in manufacturing, chemical sensing, and medicine. The theory of computation has proven invaluable in enabling information processing in electronic man-made systems, and much-studied algorithms underlie the behavior of everything from communication networks to video games. However, a thorough understanding of the principles of chemical computation is still lacking. The goal of this proposal is to use rigorous mathematical models to investigate the capabilities and limitations of chemical information processing. The proposed research will bring the fields of physics, chemistry, biology, and computer science closer intellectually and mutually enrich them. For example, conceptual frameworks and mathematical tools capturing the manipulation of information at the molecular level may yield critical insights into the design principles of evolved biological regulatory networks. Further, understanding how information processing is possible in the disordered world of chemistry could result in error-robust electronic computing. The project will also contribute to the development of undergraduate and graduate courses, which will train students to apply the principles of computer science and electrical engineering in traditionally incompatible domains. This will encourage the next generation of scientists to break through traditional disciplinary barriers and create the scientific and engineering fields of tomorrow. This proposal will answer foundational questions about the computational power of chemical kinetics (chemical reaction networks). How can chemicals be programmed to have desired behaviors? How much molecular energy does such computation consume? How much "more computation" does every additional chemical reaction enable? Recent advances in DNA nanotechnology (strand displacement cascades) demonstrate that molecular systems of complex functionality can be designed and constructed from the ground up. This proposal will help precisely delineate the capabilities and limitations of this technology, resulting in smaller, simpler DNA-based circuits. This proposal also introduces a new paradigm, based on the laws of thermodynamics, for programming DNA-DNA interactions. As chemical and biological systems are comprised of molecules that are inherently information-rich and programmable, principles of computer science will help design smart molecules.