Frank Doyle: The Control of Time

A challenging problem in systems biology is to rapidly re-entrain circadian oscillations that have been forced out of synchrony with the environment due to external factors such as jet lag or shift work. Recent studies have identified small-molecule pharmaceuticals which effect dose-dependent changes in the mammalian circadian clock, providing an avenue for control of circadian amplitude and phase. In this work, we use a molecular model of the mammalian circadian gene regulatory network to inform a model predictive controller (MPC) for optimal re-entrainment of the circadian clock. Practical constraints on the re-entrainment process arise due to the wide variation in the parametric phase response curves (PRCs) describing drug response. We demonstrate that, in the presence of such practical constraints, careful consideration of the MPC prediction horizon, the cost function, and the controller sampling time is required for optimal phase resynchronization. We illustrate the effectiveness of circadian clock manipulation via MPC through in-silico experiments.

Seminar Details: https://cbe.nd.edu/frank-doyle

Originally published at https://cbe.nd.edu/