Seminars are held on Tuesdays from 4:00 p.m. at 114 Dana Research Center, with refreshments served beforehand at 3:45. All are welcome to attend!

---

May 29th, 2003 SPECIAL THURSDAY CIRCS/COLLOQUIUM
TITLE: "Precision, reliability and the neural code"
By: Professor Paul Tiesinga
Physics & Astronomy
University of North Carolina, Chapel Hill

ABSTRACT: Neuroscientists believe that neurons encode information in their output spike trains. Unfortunately, it is unclear what spike-train features are informative. A common hypothesis is that neurons produce Poisson spike trains with a time-varying rate. In that case, the temporal dynamics of the rate contains all the information. I will show that results from computational models, in vitro and in vivo experiments are inconsistent with the Poisson assumption, instead spike trains have non-renewal statistics. I will introduce a new framework, "spike-time attractors", to account for these results and discuss consequences for neural coding.

---

May 27th, 2003
TITLE: "Probing Single Enzyme Molecules at Work"
By: Professor XiaoLiang Sunney Xie
Department of Physics and Chemistry and Chemical biology
Harvard University

---

May 20th, 2003
TITLE: "Some Quantitative Issues in Evolution"
By: Professor Daniel Fisher
Department of Physics
Harvard University

---

May 6, 2003
TITLE: "Darwinian evolution as front propagation on the fitness
landscape""
By: Professor Herbert Levine
Department of Physics
University of California at San Diego

---

April 15th, 2003
TITLE: "Judging the quality of gene clusters using information theory"
By: Frank Gibbons, Ph.D.
Computational Biologist, Harvard Medical School
Harvard University

ABSTRACT: Humans are currently believed to have somewhat less than 30,000 genes, only a small fraction of which have a known function. The challenge for the modern biologist is to catalog function as efficiently as possible, in order to home in on the most interesting genes. It will take the combined efforts of both the hard and soft sciences to find effective ways of doing this.

Novel Microarray ('gene-chip') technology makes it possible to perform experiments quickly and easily on many genes at once. Clustering is an exploratory data-analysis tool that has seen widespread adoption in the microarray community, in which genes are grouped according to the similarity of their expression pattern across a range of conditions, in the hope of establishing 'guilt by association' among genes. However, it is somewhat of a black art, in which most decisions are justified a posteriori. After introducing the basic biological and technological concepts behind microarrays, I will illustrate ways in which information-theoretic concepts applied to current databases of gene function can be used to gauge quality among various clustering techniques. This talk is aimed principally at an audience who may be interested in learning more about the exploding interdisciplinary field of bioinformatics/computational biology.

---

April 1, 2003
TITLE: "THEMATICS: Using Physical Science Methodologies for Functional Genomics"
By: Professor Mary Jo Ondrechen
Department of Chemistry and Chemical Biology
Northeastern University

ABSTRACT: New applications of methodologies from physics and chemistry for the prediction of protein function are discussed. THEMATICS, a method for the location and characterization of the active sites of enzymes, is featured. THEMATICS, for Theoretical Microscopic Titration Curves, is based on
well-established finite-difference Poisson-Boltzmann methods for computing the electric field function of a protein. THEMATICS requires only the structure of the subject protein and thus may be applied to proteins that bear no similarity in structure or sequence to any previously characterized protein. The unique features of catalytic sites in proteins are discussed. Discussion of the chemical basis for the predictive powers of THEMATICS is featured. Some results are given for illustrative examples of importance in biology and medicine; implications for genomics are discussed.

---

special circs seminar: note the day
WEDNESDAY February 13th, 2003
TITLE: " Fluctuating manifolds with type specific interactions: from a toy problem to recognition in immunology"
By: Yuko Hori
Chemistry Department
University of California at Berkeley

ABSTRACT: Transfer of information encoded in one fluctuating manifold to another by reorganization processes is important in many contexts. We first consider a disordered heteropolymer interacting with colloidal particles in a 3-dimensional medium. Specifically, we study how these particles reorganize in response to interactions with the polymer segments, and how the resulting spatial pattern depends upon the polymer sequence. Cellular recognition in the immune system is a key feature of complex organisms.
Here, receptors and complementary ligands embedded in 2-dimensional cellular membranes bind and form complex spatial patterns in order to trigger an immune response. The patterned collection of receptors that forms in intercellular junction is called the immunological synapse. We will discuss the mechanisms that underlie the formation of different synapse patterns at different stages of the life cycle of T cells. The functional consequences of synapse formation will also be explored.

---

February 4th, 2003
TITLE: "Dynamic Instability of Calcium Cycling under Rapid Pacing "
By: Dr. Johannes Shiferaw
CIRCS and Department of Physics
Northeastern University

ABSTRACT: It is well-established that rapidly paced cardiac myocytes can exhibit
unstable electrical responses, such as repolarization alternans, which have been linked to life-threatening arrythmias and fibrillation. The role of intracellular calcium cycling in unstable dynamics, however, remains poorly understood in both isolated cells and tissue. Because of the bi-directional coupling of the calcium and the electrical systems, instability of the calcium system is potentially an important mechanism for generating dynamic heterogeneity, independent of, or in conjunction with, electrical restitution. We present a new mathematical model of calcium cycling that distinguishes itself from previous models in two important ways. Firstly, the model takes into account explicitly the spatially localized nature of release events that correspond to experimentally observed calcium sparks. Hence, it naturally incorporates graded release by making the rate at which calcium sparks are recruited proportional to the whole cell L-type calcium current, with the total release of calcium from the sarcoplasmic reticulum (SR) being just the sum of local releases. Secondly, the model remains quantitatively valid for fast pacing rates. Therefore, it can be coupled to existing membrane ionic models to probe systematically the relative importance and the interaction of the electrical and calcium systems in the generation of dynamic heterogeneity in tissues. The model is validated by comparing its predictions to experimental measurements of calcium alternans in
electrically stimulated single myocytes. The physiological origin of the instability leading to alternans is further elucidated by a mathematical reduction of the model to a system of two coupled discrete maps. These maps describe the relationship between the calcium concentration in the myoplasm and the SR at one beat as function of the concentrations at the previous beat. A stability analysis of these maps reveals that calcium alternans can be due to a combination of calcium induced inactivation of the L-type calcium current and a steep SR load dependence of calcium release, with the degree of nonlinearity in the latter being strongly dependent on the spark lifetime.

---

January 28th, 2003
TITLE: "Towards an Anatomical, Physiological and Behavioral Model of Vertebrate Descending Motor Control: Raw Materials "
By: Professor Don O'Malley
Department of Biology
Northeastern University

ABSTRACT:

Control of locomotion is a fundamental task that, in vertebrate animals, is implemented by neurons descending from brain into spinal cord. These neurons govern the functioning of spinal motoneurons, interneurons and CPGs (central pattern generators). The organization of descending controls signals is poorly understood because of (1) the great diversity of descending cell types and (2) the intermingling of distinct cell types within the brainstem reticular formation. Even in lower vertebrate animals, which lack a cortical-spinal tract, the mechanisms by which locomotion is controlled remain elusive. Our lab has focused on a maximally simple vertebrate descending motor control system (DMCS), that of the larval zebrafish. Because it is transparent, we are able to image neural activity in populations of neurons inside the intact animal. Confocal imagingrevealed physiological events associated with different kinds of locomotion and also enabled other kinds of experiments. Laser ablation experiments, for example, allow us to optically dissect neural circuitry in intact animals, after which high-speed behavioral recordings are used to determine the nature and extent of ensuing behavioral deficits. Because the descending neurons in larval zebrafish can be individually identified, information obtained in other kinds of experiment (e.g. neuroanatomical) can be combined with the physiological and behavioral data. Of particular importance is the ability of optical methods to systematically survey cellular elements of the descending control system. Many researchers are attempting to understand vertebrate neural architectures in instances where they might know of only 50%, 20% or even less than 10% of the cell types involved. In contrast, application of optical methods to the larval CNS is leading us towards a comprehensive physiological, anatomical and behavioral description of the DMCS. Data obtained thus far indicate that these free-living animals possess an extremely sophisticated DMCS. Our most immediate challenge is to create a model that satisfactorily incorporates these various kinds of data. A meaningful computational model of this system is expected to be 1 to 2 rders of magnitude more complex than existing models of vertebrate locomotor control.

 

---

January 21, 2003
TITLE: " Engineering Josephson junction circuits for quantum information tasks"
By: Dr. Caspar van der Wal
Department of Physics Harvard University previously at Delft University of Technology

ABSTRACT:

Superconducting circuits with small high-quality Josephson tunnel junctions are currently investigated with the goal to apply these circuits in quantum-information processing tasks. I will review some of the first
experiments that show that well-defined two-level quantum systems can be realized with Josephson junction technology. My review will emphasize the experiments from the Delft group (collaboration with MIT). The Delft group works with a basic unit that is a small loop containing three Josephson junctions. A classical version of the loop has two stable states: a clockwise or counter-clockwise persistent current in the loop. A quantum version of the circuit can be in a superposition of these two states. The quantum circuit operates in an electronic and solid state environment that will cause decoherence of the quantum system. The decoherence sources that dominate in current experiments will be discussed, as well as engineering strategies for long decoherence times.

---

January 14th, 2003
TITLE: " Oscillatory dynamics of the vortex lattice in superconductors"
By: Dr. Sergio O. Valenzuela
Department of Physics
Harvard University

ABSTRACT:

Vortices in high temperature superconductors comprise a model system for the study of thermodynamic and dynamic phase transitions. The interplay of vortex-vortex interactions and pinning by defects in the superconducting matrix both determines the electromagnetic response of type-II superconductors and provides an ideal system to study the effect of quenched disorder on elastic media. It has been recently observed that the vortex system can be in different configurations, with variable degrees of order, which arise in response to different system histories. In particular, the pressure exerted by an oscillating sinusoidal field assists the vortex system in ordering in the penetrated outer zone of the sample (where the ac currents flow). This causes a local increase in the vortex mobility that gives rise to strong memory effects. I will describe experiments and numerical simulations that help to elucidate the mechanism of the ordering process. By measuring ac susceptibility in YBa2Cu3O7 single crystals, we show that when vortices
are shaken by a temporarily symmetric (e.g. sinusoidal) ac field they are driven into an easy-to-move, ordered structure. On the contrary, when the ac field is temporarily asymmetric (e.g. saw-tooth), they are driven into a more pinned disordered state. This is characteristic of tearing of the vortex lattice and it suggests that the ordering mechanism due to symmetric ac fields is essentially different from both an equilibration process and a dynamical crystallization that is expected to occur at high driving currents. The connection between the macroscopic response of the superconductor and the microscopic behavior of vortices is investigated by means of overdamped molecular dynamics simulations.

---

December 3rd, 2002
TITLE: "Theoretical Micromechanics of DNA and DNA-protein Complexes"
By: Dr. Abhijit Sarkar
Department of Physics and Chemistry and Chemical biology
Harvard University

ABSTRACT:

We theoretically analyze three problems in single-molecule micromechanics. These are: (1) the elastic response of torsionally constrained DNA, (2) the effect of applied torque on the stability of binding of proteins to DNA and (3) the force response of DNA with designed sequence inhomogeneities or bound proteins undergoing mechanical strand separation. Statistical mechanics and polymer
physics are the main tools in our study. We find that the high-force elasticity of topologically-constrained
DNA can be explained in terms of transitions between novel DNA structural states. We find static torques of a few k_BT are sufficient to disrupt the binding of proteins to DNA. Dynamical torque generation through transient twist distortions in the DNA are found to propagate up to 100 bp with sufficient strength to dislodge bound proteins. Unzipping DNA with sequence variations or bound proteins is found to generate distinct force signals with large amplitude. We discuss the biological relevance of our work and point to future experiments that may be able to test our predictions.

---

November 19th, 2002
TITLE:The information content of spontaneous activity in the developing visual system and its role in brain development
By: Dr. D. Butts
Department of Neurobiology
Harvard Medical School

ABSTRACT:

Spontaneous activity is present in the mammalian retina before the retina is responsive to light, and is known to be necessary for several aspects of the development of the visual system. Though much is known about the spontaneous activity (called "retinal waves") and the development that it drives, it is not known about the mechanisms responsible. Since retinal waves instruct development, they must convey information, and this information can be measured to determine constrains on the developmental mechanisms that use it. Using both multi-electrode and calcium imaging experiments of the spontaneous retinal activity, I measure the information content of the retinal waves, and determine both its time scales and what features of retinal activity are important in conveying information. This leads to clear predictions about the mechanisms that govern brain development, which are currently being tested experimentally. More generally, this new implementation of information theory demonstrates how the properties of neuronal systems might be inferred from the statistics of their input.

---

November 12th, 2002
TITLE: Wiring a brain: Theoretical physics meets neurobiology
By: Dmitri "Mitya" Chklovskii
Assistant Professor
Cold Spring Harbor Laboratory

ABSTRACT:

Human brain contains a hundred billion neurons, each making thousands of connections. These connections require physical wiring which uses valuable resources. We explore how evolution has optimized the wiring in the brain while preserving a potential for plasticity needed for adaptation. By using theoretical physics tools we attempt to understand essential brain functions such as learning and memory.

---

October 15th, 2002
TITLE: DNA translocation across protein channels: How does a polymer worm through a hole?
By: Professor M. Muthukumar
Polymer Science and Engineering, University of Massachusetts, Amherst

Conformational entropy of a polymer is reduced when the polymer confronts a narrow path. The reduction in entropy and the consequent free energy barriers control the polymer's translocation through narrow channels. Based on an analogy with the classical nucleation and growth process, we have identified
three stages in translocation: (1) a very slow process of placing an initial monomer at the entrance of the channel, (2) attempts to make a stable nucleus of enough monomers across the channel, and (3) the eventual escape of the polymer. Theoretical results on the dependencies of the nature and duration of
these three stages on the length, stiffness, and sequence of the polymer, solution conditions, and the strength of the driving electrochemical potential gradient will be presented. Our predictions will be compared with known experimental results and prospects of fabricating single-molecule devices to read sequences of DNA/RNA and proteins will be addressed.

---

October 1st, 2002
TITLE: Attention and visual perception
By:Prof. A. Reeves
Psychology Department
Northeastern University

ABSTRACT:

Paying attention to a simple visual stimulus, say a patch of color, has no effect on its appearance, but does decrease the 'reaction time', the latency of one's response. That attention speeds visual processing has been shown both behaviorally and electrophysiologically, at the level of the visual cortex (V1 and V2). The operational definition of 'attention' I adopt involves selection between two competing tasks, performance in
both being measurable. Defining attention narrowly as selection by-passes issues of consciousness and brain state but makes behavioral measurements of attention possible. Using a two-competing-visual-task procedure, I have been studying the effect on speed of processing visual stimuli in complex spatio-temporal displays. I will discuss the finding that fluctuations in attention can re-order perceptions. That is, attended stimuli may be processed so much faster than less-attended ones, that the former can be seen before the latter even if they are presented later. I will also use this method to estimate the time of an attention shift,
and demonstrate that after considerable practice, it is possible to shift your attention from left to right while simultaneously shifting your eyes from right to left, with no cost to either. But don't try this at home !

---

Sep. 10th, 2002
Title:""
By: I. Webman
Department of Physics
Barilan University

ABSTRACT:

---

May 28th, 2002
Title:"How do we learn? Learning theory in the 21st Century."
By: Donna Qualters, PhD
Director for Center for Effective University Teaching
Northeastern University

ABSTRACT:

Is the lecture dead? Who's responsible for learning? How can I possibly cover material AND do active learning in a quarter? Modern science has opened the window to the brain and in doing so has revealed some interesting facts about why and how people learn. This seminar will explore the most recent learning theories such as social cognition, brain based/neuroscience, multiple intelligences, and constructivism and explore the implications and impact of these theories on a modern day science classroom.

---

May 14 th, 2002
Title:"Localized States and Biophysics"
By: Professor David Nelson
Department Of Physics
Harvard University

ABSTRACT:

Localized and extended states play important roles in a number of biological problems. We discuss in some detail the delocalization in DNA when the two strands are pulled apart by a constant force. When the force exceeds a critical threshold, a remarkable unzipping transition occurs which is strongly influenced by randomness in the base pair sequence. The DNA unzips via a series of discrete jumps which allow it to reach successively deeper energy minima. Above the threshold force, the dynamics of the unzipping force is related to that of a particle diffusing in a random force field. Time permitting, we will describe related problems involving bacterial growth in an inhomogeneous medium with diffusion and convection and the stalling out of detachable motor proteins, such as DNA and RNA polymerase, while moving along linear
filaments with quenched randomness.

---

---

y:

---

y:

---

y:

---

y:

February 11th, 2002
Title:"Biochemical studies on recombinant prion protein-nucleic acid interaction"
By: P.K. Nandi
Pathologie Infectieuse et Immunologie, Institut National de la Recherche Agronomique, FranceBy:

ABSTRACT:

The fatal neurodegenerative prion disease is both genetic and infectious and can inflict on humans and other animals. Unlike in viruses and bacteria, where nucleic acids carry the infection, a structural change in the a-helix rich normal cellular prion protein, PrPC to its b-sheet rich, scrapie isoform, PrPSC has been considered as an obligatory step of the occurrence and propagation of the prion disease (C, normal cellular prion protein, Sc stands for scrapie, the prion disease in sheep). PrPSC-amyloid as well as intermediate oligomers are associated with the neuropathology of the prion disease. The process of conversion to PrPSC form has been postulated to require the binding of a still unidentified cofactor 'protein X' to PrPC . Nucleic acid can induce structural change of recombinant prion protein to its b-sheet rich form which results in polymerization of the protein to amyloid. The protein in turn induces stepwise association of nucleic acid molecules to condensed ordered aggregates or nucleoprotein complex globules which spontaneously dissociate. This has led to the demonstration of functional properties of the nucleoprotein complex. Our results may indicate that nucleic acid can be the much sought for cofactor for the conversion of prion protein to amyloid related to the prion disease.

---

February 5th, 2002
Title: "DNA Topology "
By: Professor Maxim Frank-Kamenetskii
Center for Advanced Biotechnology and Department of Biomedical Engineering
Boston University By:

ABSTRACT:
By the virtue of DNA chemical structure, its termini can be easily locked with each other forming the circular molecule. As a result, numerous topological states are inherent in DNA. The DNA molecule can form knots. The two strands in closed circular duplex DNA form links of a high order (high linking numbers). Topological invariants of knots and links play very important role in the field of DNA. DNA topology is crucial for DNA functioning. Special enzymes, DNA topoisomerases, resolve the DNA topological problems in the cell. Very recently, various artificial DNA topological (and pseudotopological) nanostructures have been elaborated. They include relatively short single-stranded DNA circles assembled sequence specifically on single-stranded and double-stranded target DNA. Padlock and earring (pseudo)topological probes are most promising for DNA diagnostics.

---

January 22th, 2002
Title: "Fast computation based on spike rate"
By: Dr. Mark van Rossum
Dept. of Biology and Center for Complex Systems, Brandeis University
Waltham, MA 02454

ABSTRACT:
One of the classic neural coding schemes is rate coding. In rate coding the information is coded in the firing rates of the cells. This idea is been based on the experimental observation that presenting the preferred stimulus often leads to an increase in the firing rate of a neuron. However, rate coding has been criticized because given the typical firing rate (up to 100 Hz) and given the Poisson-like variability in typical cortical spike trains, substantial averaging time would be needed for an accurate estimate of the signal. Experimental data, however, suggests there is only very little time for such averaging, as humans can categorize stimuli within 150 ms (Thorpe et al., 1996).
In this paper we show how firing rate can be coded in a population of neurons. More precisely, we study how activity propagates through a layered feed-forward network with some 50 integrate-and-fire neurons per layer. All neurons are primed with an independent noise current and a net excitatory current. Time varying stimuli is rapidly transmitted through many layers (10 layers within 50 ms, not taking axonal delays into account), whereas the temporal shape of the stimulus is highly conserved.
Next, we show how this mode of operation can be used for motion detection. We build a Reichardt-type motion detector, which calculates motion by cross correlating the image with a translated, time-delayed version of the image (Reichardt, 1957). Instead, of relying on detailed biophysics to perform such calculation, we perform motion detection in a simple feed-forward circuit.:

January 15th, 2002
Title: "Fluctuating energy landscapes and enzymes, molecular motors, and ion pumps"
By: Professor Dean Astumian

Department of Physics and Astronomy, University of Maine,

ABSTRACT:

Non-equilibrium Fluctuations, whether imposed externally or driven by an energy releasing chemical reaction, can cause a protein to cycle through several conformations. This cycling can drive a process thermodynamically uphill even though any one conformation considered independently catalyzes the process in the downhill direction. The results apply equally to driving a biochemical reaction away from equilibrium by an enzyme, to formation of an osmotic gradient across a membrane by a molecular pump, or to motion and generation of force by a molecular motor.

December 10th, 2001
Title: "Attractor model of working memory in the neocortex: theory meets experiment"
By: Professor Xiao-Jing Wang
Center for Complex Systems, Brandeis University By:

ABSTRACT:
In this talk, I will give a survey of experiments and theoretical work on cortical networks that show attractor-type dynamics underlying short-term memory. I will discuss cellular and synaptic mechanisms and collective network dynamics that may underlie short-term memory. I will also show how such networks can perform `cognitive computations' such as perceptual decision making.

---

November 6th, 2001
Title: "Is the micro-rheology of living cells governed by a glass transition?"
By: Dr. B. Fabry
Harvard School of Public Health, Physiology Program, Boston, MA 02115

ABSTRACT:
Current descriptions view cell mechanics as an interaction of distinct elastic and viscous components expressing a limited range of characteristic relaxation times, with elasticity thought to be regulated through a sol-gel transition. We developed a micro-rheometer based on magnetically twisted beads that were bound to integrin-receptors to measure the mechanical properties of isolated adherent cells (airway smooth muscle cells, neutrophils, bronchial epithelial cells and alveolar macrophages) over frequencies spanning 5 orders of magnitude. Elastic stresses dominated at frequencies below 300 Hz, increased only weakly with frequency and followed a power law. Frictional stresses were also weakly dependent on frequency below 30 Hz but approached a viscous limit at higher frequencies. Surprisingly, data for all cell types, frequencies and interventions studied could be scaled onto universal master curves. This scaling identifies these cells as soft glassy materials existing close to a glass transition, with an effective noise temperature, x, of about 1.2. These results contradict current models in that 1) relaxation processes exhibited no intrinsic time-scale, implying stress relaxation proportional to t^(1-x), and 2) frictional stresses seemed to reside within solid cytoskeletal structures and did not correspond to a viscous friction. These findings point to the hypothesis that cytoskeletal proteins regulate cell mechanical properties mainly by modulating the effective noise temperature.

October 30th, 2001
Title: "Microstructure evolution in polycrystals"
By: Dr. Alexander Lobkovsky
Physics Department and CIRCS, Northeastern University

ABSTRACT:
Most material properties are strongly influenced by the structure on the microscopic length scales. I will review types of microstructures observed in casting of metals and focus on the crystalline grains. To predict the formation and evolution of the crystal grain structure, it is essential to understand the way grain boundaries move and interact with impurities. I will present a summary of previous modeling efforts in that direction and focus on the phase field method for capturing the physics of grain boundaries.

October 23, 2001, 2001
Title: Tuning DNA "Strings": Controlling the Speed (and Direction) of Molecular Engines that Replicate DNA
By: Anita Goel
Department of Physics, Div. of Health,Sciences and Technology (HST), Harvard University and MIT and Harvard Medical School

ABSTRACT:
Recent single molecule experiments are elucidating new information about the dynamics of motor enzymes that move along polymer substrates,includingtheir sensitivity to external control parameters such as mechanical tension on the template. Meanwhile, X-ray crystallographers have taken high-resolution snapshots of the intricate machinery of these molecular motors (DNA polymerases) involved in the act of replication. Drawing upon both recent structural and single molecule data, we discuss how mechanical forces might couple into the chemical reaction of DNA replication. Time permitting, we might speculate on some possible relevances for biophysics and nanotechnology.

---