Physics of Information Processing in Living Systems - Details of our work

Yuhai Tu's Group

Physics of Biological Information Processing

About the Group: We are in the Physical Sciences Department at IBM T. J. Watson Research Center. The PI is Dr. Yuhai Tu.

Research Interests: Our general research interests in biophysics are two-fold: we are interested in understanding important biological phenomena (e.g., signal transduction) by using tools from physics (computational modeling, statistical physics, dynamical systems analysis); we also wish to uncover general design principles and novel physics (e.g., non-equilibrium thermodynamics) from studying biological systems. Currently, we are working on:

  1. To understand dynamics of biochemical information processing in molecular signaling networks.
  2. To understand the mechanics and the control of nanoscale biological machines (such as the rotary flagellar motor of bacteria).

Our approach is to combine systems level modeling with quantitative experiments (done in various experimental collaborators' labs). One of the specific model systems we focus on is bacterial chemotaxis. Some general questions we are trying to address include:

  • How do biological systems obtain (sensing) and retain information (memory) about their environment?
  • How do biological systems process (compute) the external and internal information?
  • How does a cell make decisions in complex environment with multiple cues?
  • What are the physical limits and costs of these cellular biological computations and decision making processes?
  • How do biological systems carry out these decisions by controlling their molecular machineries, such as the flagellar motor?

Besides studying cellular level signaling based on molecular biochemical networks, we have recently started studying the spatio-temporal dynamics (e.g., various brain oscillations) of realistic neural networks (with Dr. Roger Traub).

Current Projects in Physical Biology:


Other Biology-Related Projects:

  • Large scale gene-expression analysis
  • Physics of micro-arrays
  • Flocking dynamics: Collective motion of self-propelled particles

I. Structure and function of the receptor clusters

Many sensory receptors form large scale complexes or clusters on cell membrane. One such example is the MCP chemo-receptors in bacteria, which form large polar clusters. Our goal is to understand the functions of these receptor clusters. Our strategy is to build models of the receptor cluster based on the biochemical properties of the individual receptors as well as the structure of the receptor clusters. We will use the models to determine the responses of the system to different stimuli and compare these predictions to experimental measurements. These comparisons are used to test hypothesis made in our models and to improve these models. We are also interested in understanding the structural basis (conformational changes) for chemo-receptor function by large scale MD simulation of the MCP receptor (with membrane and explicit water) using IBM Blue-Gene computer, and in understanding the formation and function of the receptor cluster that form the basis for signal amplification and integration.

Receptor cluster HI-resolution image

  • “Discovery of Novel Chemoeffcetors and Rational Design of E. coli Chemoreceptor Specificity”, S. Bi, D. Yu, G. Si, C. Luo, T. Li, Q. Ouyang, V. Jakovljevic, V. Sourjik, Y. Tu, and L. Lai, Proc. Natl. Acad. Sci. USA (PNAS),110(42), 16814-16819, 2013.
  • "Lateral density of receptor arrays in the membrane plane influences sensitivity of the E. coli chemotaxis response", C. M. Khursigara, G. lan, S. Neumann, X. Wu, S. Ravindran, M. J. Borgnia, V. Sourjik, J. Milne, Y. Tu, and S. Subramaniam, The EMBO J. , 25 March 2011; doi:10.1038/emboj.2011.77.
  • "Adapt locally and act globally: strategy to maintain high chemoreceptor sensitivity in complex environments", G. Lan, S. Schulmeister, V. Sourjik, Yuhai Tu, Molecular Systems Biology 7:475, 2011.
  • "An allosteric model for heterogeneous receptor complexes: Understanding bacterial chemotaxis responses to multiple stimuli", B. Mello and Yuhai Tu, PNAS, 102(48), 17354-17359 (2005).
  • "Effects of receptor interaction in bacterial chemotaxis", B. Mello, L. Shaw, and Yuhai, Tu, Biophysical Journal, 87(3), 1578-1595(2004).
  • "Quantitative Modeling of Sensitivity in Bacterial Chemotaxis: The Role of Coupling Between Different Chemoreceptor Species", B. Mello and Yuhai Tu, PNAS, 100(14), 8223-8228 (2003).



II. Dynamics of the signal processing network

We study the molecular mechanism for signal processing in biological networks, e.g., signal amplification; signal integration and signal differentiation in bacterial chemo-sensory system.

We have developed a simple model to describe the signaling dynamics, which incorporated the two most important ingredients, signal amplification and accurate adaptation, in the simplest way that are consistent with the known knowledge of the pathway. This "standard model" of bacterial chemotaxis is used to explain a wide range of existing experiments within a simple unified model framework. It is also used to predict responses to time-dependent signals (ramps and sine-waves). These predictions are verified by experiments.

Standard model of E. coli chemotaxis HI-resolution image

  • "A modular gradient-sensing network for chemotaxis in E. coli revealed by responses to time-varying stimuli", T. S. Shimizu, Yuhai Tu, and Howard C. Berg, Molecular Systems Biology 6: 382, 2010.
  • "Modeling the chemotactic response of E. coli to time-varying stimuli", Y. Tu, T. S. Shimizu and H. Berg, PNAS, 105(39), 14855-14860 (2008)
  • "Effects of adaptation in maintaining high sensitivity over a wide range of backgrounds for E. coli chemotaxis", B. Mello and Y. Tu, Biophysical Journals, 92(4), 2329-2337 (2007).
  • "Perfect and near perfect adaptation in a model of bacterial chemotaxis", B. Mello and Y. Tu, Biophysical Journal, 84(5), 2843-2856 (2003).

A recent review article: our study on these two aspects of the E. coli chemotaxis signaling patwhay -- signal amplification and accurate adaptation -- has been summarized in a recent review paper.

  • “Quantitative Modeling of Bacterial Chemotaxis: Signal Amplification and Accurate Adaptation”, Yuhai Tu, Annu. Rev. Biophys. 42: 337-59, 2013.



III. Bacterial flagellar motor: Mechanics and control

Bacterial flagellar motor has one rotor and multiple stators in a circular ring-like structure of roughly 45nm in diameter. Torque is generated by interaction of stator units, anchored to the peptidoglycan cell wall, with the rotor, driven by the proton motive force (pmf) across the cell membrane. The motor's rotation direction can be regulated by intracellular response regulator (CheY-P) as well as external conditions, such as load and temperature.

Our goal is to understand both the rotational motion (mechanics) of the flagellar motor as well as its switching (between CCW and CW directions) behavior in a unified model framework. 1) Mechanics: we have proposed a mathematical model of the motor dynamics. This model provides the microscopic mechanism for the macroscopic characteristics of the motor, such as the torque-speed relationship and the dependence of the maximum velocity on the number of torque generating units. 2) Switching: we have proposed a non-equilibrium mechanism for the BFM switching, in which the BFM is switched by certain dissipative mechanism, akin to the Maxwell daemon. Though the origin of the Maxwell daemon is unknown, they are shown to be necessary in order to explain certain distinctive features, such as peaks in CCW and CW duration time distribution function, measured in experiments. We are now developing the model in order to link the mechanical part, associated with the rotational motion, with the signaling part, associated with the switching process.

It was recently discovered that the flagellar motor is not just a fixed machine, instead it can adjust itself in response to external (chemical and mechanical) stimuli. We are interested in studying the underlying mechanisms for these motor-level adaptations. Bacterial Flagellar Motor HI-resolution image

  • “Tandem adaptation with a common design in Escherichia coli Chemotaxis”, Yuhai Tu, Howard C. Berg, J. Mol. Bio. 423, 782-788, 2012.
  • "Dynamics of the Bacterial Flagellar Motor: The Effects of Stator Complaince, Back Steps, Temperature, and Rotational Asymmetry", G. Meacci, G. Lan, and Yuhai Tu, Biophysical Journal, 100 (8), 1986-1995, 2011.
  • "Dynamics of the bacterial flagellar motor with multiple stators", G. Meacci and Yuhai Tu, PNAS, 106(10), 3746-3751 (2009).
  • "Nonequilibrium mechanism for a biological switch: Sensing by Maxwell's demons", Y. Tu, PNAS, 105(33), 11737-11741 (2008).
  • "How white noise generates power-law switching in bacterial motors", Y. Tu and G. Grinstein, PRL, 94, 208101(2005)



IV. Pathway-based modeling of cell motility: From pathway dynamics to behaviors

A computational model, based on a coarse-grained description of the cell's underlying chemotaxis signaling pathway dynamics, has been developed to study Escherichia coli chemotactic motion in any given environments that can change in both space and time. We have used this Signaling-Pathway-based E. coli Chemotaxis Simulator (SPECS) to study well known chemotaxis assay, such as the capillary assay, and measurements in recent microfluidics devices. By combining SPECS and the microfluics measurements (by our collaborators lab), we have studied the logarithmic sensing ability of the E.coli cells and their frequency-dependent behaviors.

Signaling-Pathway-based E. coli Chemotaxis Simulator (SPECS) HI-resolution image

Recently, we have developed a continuum model for E. coli chemotaxis population dyanmics based on the underlying molecular level signaling pathway dynamics. We have also tested some of the predictions for cellular behaviors under spatio-temporal varying conditions by using microfluidic experiments.

  • "Pathway-based mean-field model for Escherichia coli chemotaxis", G Si, T Wu, Q Ouyang, Y Tu - Physical review letters, 109(4), 048101-048105 (2012).
  • "Frequency-dependent Escherichia coli chemotaxis behavior", X. Zhu, G. Si, N. Deng, Q. Ouyang, T. Wu, Z. He, L. Jinag, C. Lou, and Yuhai Tu, Phys. Rev. Lett., 108, 128101-128105 (2012).
  • "Quantitative Modeling of E. coli Chemotaxis Motion in Environments Varying in Space and Time", L. Jiang, Q. Ouyang, Y. Tu, Plos Comp. Bio., 6(4), e100735, 2010.
  • "Logarithmic sensing in Escherichia coli bacteria chemotaxis", Yevgeniy V Kalinin, Lili Jiang, Yuhai Tu, and Mingming Wu, Biophysical Journal, 96(6), 2439-2448 (2009).



V. Noise in biology

We are interested in understanding the noise in biological systems, in particular in signaling and regulatory systems. We want to understand their origins and their effects; how they propagate through the signaling pathway and what are the cell's strategies in filtering the noise. In the case of the bacterial flagellar motor (BFM), we are intetested in understanding the origin and characteristics of the noise in stepping and switching of the BFM. We are also interested in understanding the energetics cost of such noise control and other regulatory functions in biology.

Noise and filtering in adaptive signaling pathways HI-resolution image

  • "Noise Filtering Strategies in Adaptive Biochemical Signaling Networks", P. Satori and Yuhai Tu, J. Stat. Phys., 142 (6), 1206-1217, 2011.
  • "Dynamics of the bacterial flagellar motor with multiple stators", G. Meacci and Yuhai Tu, PNAS, 106(10), 3746-3751 (2009).
  • "How white noise generates power-law switching in bacterial motors", Y. Tu and G. Grinstein, PRL, 94, 208101(2005)



VI. The energy cost of biological sensory adaptation

We are interested in studying the costs (especially the energetic cost) of performaning biological functions (especially the regulatory functions). Our goal is to understand how such costs relate to (or limit) the performance of these biological functions, and if or how a biological system optimizes its efficiency. We have studied the dynamics and energetics of the basic biochemical networks underlying biological sensory adaptations. We found that there is a universal trade-off relationship between the energy cost of the network and the performance of the system charcaterized by the accuracy and speed of adaptation.

  • “The energy-speed-accuracy trade-off in sensory adaptation”, G. Lan, P. Sartori, S. Neumann, V. Sourjik, and Yuhai Tu, Nature Physics 8, 422–428, 2012.
  • “The cost of sensitive response and accurate adaptation in networks with incoherent type-1 feed-forward loop”, Ganhui Lan and Yuhai Tu, J. R. Soc. Interface, 10(87), 2013.



VII. Other taxis system: precision sensing and its underlying mechanisms

We are also interested in other taxis system, in particular thermotaxis, where the same chemotaxis gradient sensing machinery is re-tooled to execute what we called "precision sensing" in which the cell goes to a particular temperature in stead of tracking the temperature gradient. For E. coli thermotaxis, we have identified a general mechanism and the required conditions for the cell to achieve precision sensing via a gradient sensing pathway. Our recent work also showed that similar "precision sensing" behvaior occurs in pH taxis. We believe it may represent a general navigation strategy in guiding cells to move to paricular environmental conditions.

Precision sensing via gradient sensing pathway HI-resolution image

  • "A mechanism for precision-sensing via a gradient sensing pathway: a model of E. coli thermotaxis", L. Jiang, Q. Ouyang, and Y. Tu, Biophysical Journal, 97, 74-82 (2009).
  • "Precision sensing by two opposing gradient sensors: How does Escherichia coli find its preferred pH level?", Bo Hu and Yuhai Tu, Biophysical Journal, 105(1), 276-285 (2013).


Last update 18.05.11