Molecular computation by cells helps them process signals in their environment to make important decisions. Such computation in cells is both analog, which uses signals with shades of grey and operations that are probabilistic or graded, and also digital, which uses "on" and "off" signals and operations that are logical. The hybrid analog-digital computation in the cell is an extremely important reason for why it is highly efficient in its use of energy and molecular parts (proteins, RNA, and DNA molecules), which are severely limited in number compared with artificial computational systems. For example, current digital microprocessors have nearly a billion electronic parts and a relatively plentiful supply of energy. The cell is several orders of magnitude more energy and part-count efficient than even the best digital computers of the far future are ever expected to be. For advancing the foundations of computer science beyond its boundaries, and for the future of biotechnology and medicine, it is important to understand and to artificially engineer hybrid analog-digital computation in living cells. The proposal focuses on how to engineer a hybrid analog-digital computational automaton in microbial cells that can enable it to sense, amplify, and process analog input molecular signals into pulsatile digital output molecular signals.
Intellectual Merit: This work combines innovations and knowledge from several disciplines to potentially create a paradigm-changing capability in the fields of synthetic biology, systems biology, molecular programming, biotechnology, computer science, and medicine: 1) Hybrid analog-digital automata can perform complex computations with very few parts and little energy, and the research here can help advance fundamental work in computer science in this area; 2) The highly efficient instantiation of a spiking-neuron-like automaton with only four genes in a microbe creates a fundamental new computational motif that can be ported to a wide range of molecular implementations including in-vitro molecular systems, in-vivo microbial, yeast, and mammalian cells; 3) The microbial fuel cell automaton enables sensitive and continuous molecular sensing to be built in a wide range of hosts without the need for bulky, expensive, non-continuous and toxic photo-bleaching optical systems; 4) By creating molecular cellular automata that serve to implement sophisticated molecular sensing, pattern recognition, and collective computation, the proposal could enable very large scale molecular computational systems to become practical.
Broader Impact: The hybrid analog-digital automata is a computational motif that could be widely applied to synthetic cell-based treatments in medicine. Microbial sensing and computational systems that are capable of sensing, amplifying, digitizing and classifying minute concentrations of input molecules could have wide application in the food, biotechnology, pharmaceutical, environmental, and security industries. PI participates in the Society of Women Engineers, MIT Summer Research Program, Saturday Engineering Enrichment Discovery Program, and computational and systems biology recruitment programs at MIT, which actively target women and underrepresented minorities.
