University of Minnesota
University Relations
myU OneStop

Go to unit's home.

Home | Seminars and Symposia | Past seminars/symposia: Tuesday, April 25, 2006

DTC Seminar Series

Modeling Biochemical Circuits: A Tutorial for Engineers


Anirvan Sengupta
Dept of Physics and BioMAPS Institute
Rutgers University

Tuesday, April 25, 2006
10:00 am

402 Walter Library

There has been enormous interest in the mathematical modeling of biochemical processes over the last decade. Originally meant to enhance our understanding of the dynamics of complex cellular regulatory circuits, quantitative analysis is becoming an essential part of "synthetic biology": the art of making artificial networks with known biological components. In this tutorial for the non-specialist, we will review several basic biochemical circuits. We will discuss signal transduction cascades, circuits that respond to external stimuli and let the cell adapt to changing conditions. We point out how concepts familiar to engineers, like gain, bandwidth or noise spectrum, help us understand the setting of biological parameters like protein abundances at various stages of an amplifying cascade. We discuss the role of feedback and gain control in these signaling systems. We also touch upon genetic switches, which feature strong positive feedback. These are multi-stable systems (like flip-flops), which hold on to different cellular fates despite the same genetic instruction (read same rule of dynamics). We will also mention how appropriately delayed strong negative feedback leads to biological oscillators — like the circadian clocks in our bodies that maintain our daily rhythms. We plan to also visit the general issue of robustness of these circuits to noise, as well as to changes of parameters. This issue brings us to some interesting questions. Can we say anything meaningful about generic features of circuits designed by evolution? To what extent are natural biological circuit parameters optimized for a specific purpose, and to what extent is it just a matter of chance? Why do biological circuits often come with so many bells and whistles when a much more simple and sparse circuit can do the same job? Time permitting, we will give an example of a moderately complex circuit: the segment polarity network in fly development. There has been much discussion of the evolutionary robustness of this circuit. We formulate a hybrid analog-digital version of this circuit and obtain insights not easily accessible via simulations alone. We claim that understanding the geometry of the space of admissible parameters, the region where the circuit can function appropriately, is crucial for figuring out which circuits are picked by evolution.


Anirvan Sengupta was trained as a theoretical physicist. After obtaining a PhD from the Tata Institute of Fundamental Research (on abstruse mathematical subjects like two dimensional strings and black holes), he decided to dive into the complexities of the real world. He started to work on strongly interacting systems at the Ecole Normale Superieure in Paris, where he was from 1992 to 1994. He joined the Bell Laboratories Theoretical Physics Group in 1994. At Bell labs, with its heady mixture of pure and applied science (and engineering), he found himself drawn towards problems that are not considered part of conventional physics research, but are tantalizingly close. These projects ranged from the application of random matrix theory to communication channels, all the way to the evolution of genetic networks. As his research focussed more and more on biological questions, Anirvan decided to venture into experimental molecular biology to complement his theoretical and computational efforts. On leaving Bell labs in 2002 to join the newly formed Biological Physics Group in the Physics Department of Rutgers University, he started his own wet lab. He is currently Associate Professor; his group focusses on the biology of gene regulation and approaches the problem from various angles: biophysics, evolutionary theory, systems biology and bioinformatics. Since its inception, Anirvan has been actively involved in the Rutgers BioMaPS Institute, which brings physical and mathematical scientists together with biologists. He is also a member of the Aspen Center for Physics. He has helped coordinate scientific activities in the New York area as a part of the Systems Biology Discussion Group of the New York Academy of Sciences. His publications of outside dedicated biology journals have appeared in Science, PNAS, Physical Review Letters and various IEEE Transactions, among others. His research is currently funded by the NIH.