"Physical Principles in Biological Networks"
Program Organizers: William Bialek, Curtis Callan, Ned Wingreen, Aleksandra Walczak
Program format, detailed description and registration:
The program is organized into six full day workshops. The first four will be focused on theoretical ideas that cut across multiple systems, and the last two will look at particular systems, with strong input from our experimental friends. On each of the four theory days, we will have a small number of longer talks that set the stage for informal discussion. The two experimental days will have a more traditional conference format.
If you are planning to attend any of the events listed, please REGISTER HERE.
Physical principles in biological networks
Almost everything that happens in living systems is the result of interactions among many elements -- a network. Our perceptions and thoughts result from dynamics in networks of neurons, during embryonic development decisions about cell fate and spatial patterning are shaped by networks of interacting genes, networks of interactions among amino acids stabilize the structure of proteins, and there are many other examples. Recent progress in experimental technique has focused attention on how one can infer aspects of the underlying network from the observable phenomenology, but there is a stream of work in the theoretical physics community which is asking for more than this: Can we find principles that govern the network dynamics, principles which would allow us to predict (rather than just fit) the rich phenomenology of these systems? Importantly, recent work has seen the successful application of similar principles (e.g., maximizing robustness to parameter variation, or optimizing the transmission of information) to very different biological systems. This theoretical work has engaged with the details of real experiments, bridging a huge cultural gap. Even so, deep questions as to goals, methods and feasibility of this approach to biological networks remain open. This program will bring together theoretical physicists who have driven recent progress in this area to engage in an organized, critical discussion of these issues.
Tuesday 24 March: Robustness vs fine tuning.
Biological networks have many parameters, most of which can be adjusted (by evolution, by learning, by changing levels of gene expression, by covalent modifications of proteins, ...). Are there mechanisms that tune these parameters to their "correct" values? Or do real biological networks have specific architectures that render their function invariant to large variations in some of these parameters? Is maximizing the functional volume in parameter space a quantitative version of this intuition about robustness? How can such invariance against parameter changes coexist with exquisite sensitivity to meaningful signals?
Provocateurs: Larry Abbott (Columbia), Chao Tang (UCSF); John Hopfield (Princeton)
Tuesday 31 March: Modularity.
In many systems we know much more about the topology of network connectivity than we do about the actual dynamics of the network. Does this topology itself have a meaning? What are the minimal features of this topology (degree distributions, motifs ...) that capture what is essential for biological function? Do observed network topologies support the intuition that complex systems are built from modules with separable functions? What is the alternative to modularity?
Provocateurs: Chris Wiggins (Columbia) & Eduardo Sontag (Rutgers)
Thursday 9 April: Finding the right operating point.
The mechanisms that we see functioning in organisms today are selected from the range of biological possibilities by many processes; on different time scales we think of these as regulation, adaptation, learning and evolution. To what extent is this need to be "selectable" (learnable, evolvable, ...) itself a principle that can help us to understand the structure and dynamics of biological networks? Do we understand the landscape in which these adaptive dynamics are occurring? What is the relation between the elementary steps in these dynamics and the functional changes that we can observe macroscopically?
Provocateurs: Paul Francois (Rockefeller), Eric Siggia (Rockefeller), Daniel Fisher (Stanford)
Thursday 16 April: Signals, noise and information.
While the macroscopic function of biological networks seems reliable and reproducible, many of the elements operate in a regime where noise (including noise of fundamental physical origin) is not negligible. Has Nature chosen mechanisms that optimize performance in the presence of this noise? Are there situations in which noise can be a creative force, allowing for more efficient exploration, or hedging of behavioral strategies? Can we connect general measures of performance in the presence of noise (e.g., the number of bits of information that a system transmits) with more practical biological measures of fitness?
Provocateurs: Anirvan Sengupta (Rutgers), William Bialek (Princeton)
Tuesday 21 April: Collective behavior in networks of real neurons.
The possibility of using ideas from statistical physics to describe neural networks has been with us for more than twenty-five years. Recently, there have been great advances in the ability of experimentalists to record simultaneously from many (2, 10, 100, ...) neurons; in some cases one can record from essentially an entire network with single cell resolution. Such experiments provide an unprecedented opportunity to connect theory to real data, and the goal of this discussion is to explore these opportunities.
Speakers: Gaspar Tkacik (U. of Penn), Michael J. Berry II (Princeton), Carlos Brody (Princeton), David Tank (Princeton).
See PROGRAM .
Thursday 30 April: Early events in the Drosophila embryo
The initial events of fruit fly development are controlled by several networks of genetic and biochemical interactions. Although the end result of these interactions is the formation of a spatial pattern that provides the blueprint for the body of the fully developed organism, many of the issues which arise are much more general. Indeed, almost all of the theoretical questions that will occupy us in the first four days have been touched upon in the context of this system. The goal for this day of discussion is to explore both the current state of experiments and new opportunities for theory/experiment interaction.
Speakers: Elizabeth Gavis (Princeton), Thomas Gregor (Princeton), Stanislav Shvartsman (Princeton), Eric F. Wieschaus (Princeton)
See PROGRAM .
Note: The state of the field is such that meaningful engagement between theory and experiment often requires pushing the technical envelope, at least in terms of data analysis and often in the experiments themselves. Although our theoretical goals are lofty, participants in these workshops will be expected to roll up their sleeves and wrestle with technical issues rather than waiting for the experimentalists to 'test their theories.'