Mathematics Colloquia and Seminars

Return to Colloquia & Seminar listing

Biological systems are notoriously high dimensional and highly nonlinear. Mechanistic models are important for understanding underlying dynamics driving system response. However, generating predictive models through a mechanistic approach remains a challenge due to limited observable states in a highly complex system. Data-driven methods will help to bridge the gap between theory and biomedical applications. In order to direct cellular response, we consider data-driven methods to controlling biology. In other words, the achievement of an intended and predicted response in a biological system. We present NN-based predictors and feedback controllers in order to direct cellular response with no model a priori and no offline training. The algorithms learn in real-time as information is received. The control algorithm adapts according to the difference between the desired response and actual response. Hence, we must also consider careful design of the desired response in real time to achieve the intended end response. To this end, we also introduce efforts towards data-driven state-transition models of complex biological processes in order to identify and drive systems towards desired reachable states via our NN-based controller.