Computations in Science Seminars
Oct
1
Wed 12:15
Stas Nagy, University of Chicago
e-mail:
Host: Leo Kadanoff ()
Homeostasis in C. elegans sleep.

Sleep may well be universal in the animal kingdom. Yet, fundamental aspects of sleep remain controversial and elusive. Questions under debate include universality, natural history, core function, and even the very definition of sleep. Sleep is widely believed to be essential for neural circuit wiring and maintenance, i.e., synaptic plasticity. That said, synaptic changes also take place independently of the sleep/wakefulness state. Moreover, sleep has been hypothesized to either strengthen or downscale synapses. Experimental evidence supporting both possibilities exists, depending on the model animal, brain region, and type of measurement in question. A key feature distinguishing sleep from other states of decreased activity such as paralysis, comatose, anesthesia, hibernation, or torpor, is its intricate homeostatic regulation. Generally, biological homeostasis invokes modulatory responses aimed at stabilizing internal conditions. In the context of sleep, the most obvious manifestation of homeostatic regulation resembles a spring: the more the period of wakefulness is stretched, the stronger the restoring force or tiredness. More detailed measurements reveal various signatures of homeostatic regulation both in sleep-deprived animals and as the normal period of sleep unfolds. This talk will describe experiments performed on a model system that is the simplest to possess a nervous system and the most phylogenetically ancient – the nematode C. elegans. Specifically, it will focus on lethargus, the sleep-like state of C. elegans. Using tunable photo- and mechano-stimulation, we identified two distinct categories of homeostatic responses during lethargus. Within this state, C. elegans exhibits alternating epochs of motion and complete quiescence. The durations of these epochs typically range between 2 and 100 sec. In the presence of weak or no stimuli, extended epochs of motion were found to cause an extension of the subsequent periods of uninterrupted quiescence. In the presence of strong stimuli, the correlations between motion and quiescence were temporarily disrupted and homeostasis manifested as a global elevation of the time spent in quiescence outside the stimulus. Two mutually exclusive mechanisms – neuropeptidergic and transcriptional regulation – were found to play roles in establishing these distinct responses to weak and strong stimuli respectively. Thus, routine stabilization of lethagus is both behaviorally and mechanistically distinct from the compensation for a strong, stressful disruption. These findings add to the list of similarities between C. elegans lethargus and sleep and highlight the importance of neuropeptides in stabilizing this state.

Oct
8
Wed 12:15
Bob Batterman, University of Pittsburgh
e-mail:
Host: Leo Kadanoff ()
Relative Autonomy and Minimal Modeling: Explaining the Robustness of Theories at Continuum Scales.

Bridging or connecting the descriptions and models of systems across widely separated scales is a deep problem that permeates many areas of scientific investigation. Unfortunately, philosophical discussion of this problem is often contextualized as an ``all or nothing'' dichotomy between reductionism and emergentism. This is much too crude.

This paper will discuss a set of mathematical techniques including the renormalization group and homogenization theory designed to upscale from models of systems that exhibit heterogeneities at small/micro scales to models that are homogeneous at continuum/everyday scales. The focus will be on two aspects of the use of such techniques. On the one hand, they appear to be essential to explain the existence of certain kinds of patterns in nature and the \emph{relative} autonomy of the continuum scale models from the lower scale details. Why, for example, do the equations that govern the scaling behavior of different fluids at criticality work so well when they completely ignore molecular scale details? Why, do the Navier-Cauchy equations for bending elastic beams work so well when they, too, essentially fail to reference any atomic or lower scale details?

On the other hand, we can also sometimes use models (toy models or \emph{minimal models}) to investigate and understand the behavior of real systems. For example, we can employ the Ising model and lattice gas automata to study the behavior of real systems---actual fluids that look absolutely nothing like these models at lower scales. The mathematics of the renormalization group and other techniques provide an account of how such \emph{non-representative} minimal models can be explanatory and can provide understanding. The paper discusses the importance of these mathematical techniques for answering the questions of autonomy, and the role and effectiveness of minimal models.

Oct
15
Wed 12:15
Jane Wang, Cornell
e-mail:
Host: Leo Kadanoff ()
From Macroscopic Laws to Neurons: Predicting Fruit Fly's Sensing Rate and the Role of Steering Muscles

To fly is not to fall. To balance in air, insects not only have to support their weight but also have to make subtle adjustment to their wing movement so not to tumble. What do insects measure to stabilize their flight? How often and how fast must insects adjust their wings to remain stable? Our recent computational analyses of free flight gave theoretical bounds on both the sensing rate and the time delay between sensing and actuation. Interpreting the phase diagram for flight stability together with fruit flies’ reaction time and firing patterns of motor neurons, we conjecture that fruit flies sense their orientation every wing beat (4ms) for stable flight. Such a beat-to-beat controller requires a fast neural pathway. We propose a candidate for the necessary feedback circuitry, which involves fruit fly’s haltere and one of its 17 pairs of steering muscles. With the new neuro-genetics tools, it's now possible to test our conjectures.

Oct
21
Tue 4:00 PM
Andy Ruina, Cornell
Host: Leo Kadanoff ()
Gliders, bicycles, toys and walking robots
JFI Seminar - Room W301 - 4:00 PM

Many airplanes can, or nearly can, glide stably without control. So it seems natural that the first successful powered flight followed from mastery of gliding. Many bicycles can, or nearly can, balance themselves when in motion. Bicycle design seems to have evolved to gain this feature. Also, we can make toys and 'robots' that, like a stable glider or coasting bicycle, stably walk without motors or control in a remarkably human-like way. So it makes sense to use `passive-dynamics' as a core for developing the control of walking robots and to gain understanding of the control of walking people. That's what I used to think. But, so far, this has not led to robust walking robots. What about human evolution? We didn't evolve dynamic bodies and then learn to control them. Rather, people had elaborate control systems way back when we were fish and even worms. But if control is paramount, why is it that uncontrolled passive-dynamic walkers can walk so much like humans? It seems that energy optimal control, perhaps a proxy for evolutionary development, arrives at solutions that have features in common with passive-dynamics. Rather than thinking of good powered walking as passive walking with a small amount of control added, I now think of powered walking as highly controlled, but with much of the motor action titrated out.

Oct
22
Wed 12:15
Andy Ruina, Cornell
e-mail:
Host: Leo Kadanoff ()
Non-holonomic stability and rotation with zero angular momentum: Demonstrations of stability and of the falling cat phenomenon go sour.

This talk is about two classes of (interesting, at least to me) physical behavior that follow from them impossibility of integrating some formulas that involve derivatives. First, systems with wheels or ice skates can be conservative yet have asymptotic stability. This is relevant to braking cars, flying arrows and the balance of skateboards and bicycles. Second, is the well known possibility that a system with zero angular momentum can, by appropriate deformations, rotate without any external torque. This effect explains how a cat that is dropped while upside down can turn over and of how various gymnastic maneuvers are performed. Both rolling contact and constancy of angular momentum are examples of the "non-integrability" of a "non-holonomic" equation. There are various simple demonstrations of these phenomena that can go bad. Cars can crash, bikes fall over and, in terrestrial experiments, various effects can swamp that which one wants to demonstrate. The talk describes the basic theory and then a collection of simple experiments that fail various ways for various reasons.

Oct
23
Thu 5:30 PM
Francesca Casadio, Art Institute of Chicago
e-mail:
Special seminar at the Department of Art History, 166 Cochrane Woods Art Center
Oct
29
Wed 12:15
Kerry Emanuel, MIT
e-mail:
Host: Leo Kadanoff ()
Radiative-Convective Instability: Implications for Tropical Weather and Climate

The concept of radiative-convective equilibrium (RCE) is the simplest and arguably the most elegant model of a climate system, regarding it as a statistically one-dimensional balance between radiative and convective heat transfer. In spite of this, RCE is seldom studied and poorly understood today. Recent advances in cloud-system-resolving numerical models have made it possible to explicitly simulate such states, simulating the convective plumes themselves rather than representing them parametrically. The simulations reveal a startling phenomenon: Above a critical surface temperature, moist convection spontaneously aggregates into a single cluster, in a non-rotating system, or into multiple tropical cyclones on a rotating planet. I will show that this results from a linear instability of the RCE state, and this this instability migrates the RCE state toward one of the two stable equilibria. This instability represents a subcritical bifurcation of the ordinary RCE state, leading to either a dry state with large-scale descent, or to a moist state with mean ascent; these states may be accessed by finite amplitude perturbations to ordinary RCE in the subcritical state, or spontaneously in the supercritical state.

Nov
5
Wed 12:15
Guenter Ahlers, UC Santa Barbara
e-mail:
Host: Leo Kadanoff ()
Nov
12
Wed 12:15
OPEN
Nov
19
Wed 12:15
Emil Martinec, University of Chicago
e-mail:
Dec
3
Wed 12:15
Susan Coppersmith, University of Wisconsin
e-mail:
Host: Leo Kadanoff ()
Dec
10
Wed 12:15
Igor Aronson, Argonne
e-mail:
Host: Leo Kadanoff ()
Jan 2015
7
Wed 12:15
Henry Cohn, Microsoft
e-mail:
Host: Leo Kadanoff ()
Jan 2015
14
Wed 12:15
OPEN
Jan 2015
21
Wed 12:15
OPEN
Jan 2015
28
Wed 12:15
Seth Lloyd, MIT
e-mail:
Host: Leo Kadanoff ()
Feb 2015
4
Wed 12:15
Zheng-Tian Lu, Argonne
Host: Daniel Holz ()
Feb 2015
11
Wed 12:15
OPEN
Feb 2015
18
Wed 12:15
Heinrich Jaeger, University of Chicago
e-mail:
Host: Leo Kadanoff ()
Feb 2015
25
Wed 12:15
OPEN
Mar 2015
4
Wed 12:15
OPEN
Mar 2015
11
Wed 12:15
OPEN
Mar 2015
18
Wed 12:15
OPEN
Mar 2015
25
Wed 12:15
OPEN
Apr 2015
1
Wed 12:15
Andrea Bertozzi, UCLA
e-mail:
Host: Leo Kadanoff ()
Apr 2015
8
Wed 12:15
OPEN
Apr 2015
15
Wed 12:15
OPEN
Apr 2015
22
Wed 12:15
OPEN
Apr 2015
29
Wed 12:15
Michael Rubenstein, Harvard
e-mail:
Host: Leo Kadanoff ()
May 2015
6
Wed 12:15
OPEN
May 2015
13
Wed 12:15
OPEN
May 2015
20
Wed 12:15
OPEN
May 2015
27
Wed 12:15
OPEN
Jun 2015
3
Wed 12:15
Alisa Bokulich, Boston University
e-mail:
Host: Leo Kadanoff ()
Jun 2015
10
Wed 12:15
OPEN
Jun 2015
17
Wed 12:15
OPEN
Jun 2015
24
Wed 12:15
OPEN