Statistical Mechanics of Biological Cooperativity

22 May 2013, 16:30 → 25 May 2013, 19:00 Europe/Stockholm

Ferry Stockholm-Mariehamn and Hotel Arkipelag, Mariehamn, Åland

Erik Aurell (KTH), Juho Rousu (Aalto University), Mikko Alava (HUT, Espoo, Finland), Ralf Eichhorn (Nordita)

Summary

This meeting, which is now over, addressed applications of statistical mechanics to biological cooperativity, from molecules to populations.

Background to the field

The scientific interface between physics and biology is one of the most active areas of scientific research, as shown by its rapid evolution during the recent years. Biological cooperativity roughly includes all action in biology that cannot be explained linearly, from synergistic interactions on the molecular level up to games and strategic interactions in populations.

The regulation of gene expression is in its simplest form akin to a switch with the binding of transcription factor to DNA playing the role of a relay. However, as has become increasingly clear, the generic picture is more global and more intricate e.g. the rate of transcription of a gene generally depends on transcription factors acting in cooperation, on epigenetic changes and marks and even on the three-dimensional location of chromosomes in the nucleus. All these lead to interesting problems in statistical mechanics ranging from deterministic and stochastic models of gene expression, including Wentzel-Freidlin theory for rare events, to models of nucleosome positioning by entropic effects up to the fractal globule model and other exotic long-lived by only metastable polymer states.

Signalling is the process whereby biological information is transferred from the outside of the cell to regulate internal processes. Statistical mechanics has been used to construct models of chemotaxis, of quorum sensing, and to lymphocyte activation in response to antigens.

It is well known that there is a close analogy between non-equilibrium statistical mechanics and models in evolutionary and population genetics, where selection and mutations play the role of drift, and genetic drift, the (partly) random selection of individuals that survive from one generation to the next plays the role of noise. Over the last decade the fundamental understanding of non-equilibrium statistical mechanics has been revolutionized by the fluctuation relations, which hold also far from equilibrium; these have only very recently been introduced in population genetics, and may there have a much larger impact.

The meeting is generously supported by NORDITA, the Aalto Science Institute, and the National Graduate School in Materials Physics (Finland) NGSMP, and is carried out with the active participation of the European scientific coordination network Evolution, Regulation and signalling (ERS).

The venue is Hotel Arkipelag in downtown Mariehamn, the capital of the province of Åland, Finland. Note however that the program starts on a ferry from Stockholm to Mariehamn and your presence there is important information for the organizers.

There is no workshop fee. Travel on the ferry from Stockholm to Mariehamn, including dinner on the ferry, is free for all participants, as are coffee and lunches at the Workshop. In addition, a number of grants can be offered, to PhD students and to others, to cover accommodation in Mariehamn and, in special cases, travel to Stockholm. If you need such support, indicate so at registration.

Invited speakers

Uri Alon	Weizmann
Ulrich Gerland	LMU Munchen
Vincent Hakim	ENS-Paris
Oskar Hallatschek	MPI, Germany
Sui Huang	ISB, Seattle, USA
Mogens Jensen	Niels Bohr Institute, Copenhagen
Debora Marks	Harvard Medical School
Olivier Martin	U. Paris-Sud
Sergei Maslov	Brookhaven National Laboratory
Ville Mustonen	Sanger Centre
David R Nelson	Harvard
Chris Sander	Memorial Sloan Kettering
Aleksandra Walczak	ENS-Paris
Chris Watkins	Royal Holloway, University of London
Pieter Rein ten Wolde	AMOLF, The Netherlands

Registration for this event is now closed

Practical information about the conference start

The conference begins on the Silja Line ferry sailing from Stockholm harbour to Helsinki on May 22, 2013, at 17.00 Swedish time. The ferry makes a stop-over in Mariehamn (where we get off). The trip to Mariehamn from Stockholm takes about five hours. We will have a lecture room and dinner on the ship.

It is recommended that you be in the ferry terminal at 16.15, at the latest. Someone from the organizing committee (Erik Aurell, Mikko Alava or Ralf Eichhorn) will be in the ferry terminal with the tickets from 15.45, at the latest.

Note that these (large) ferries leave on time, and wait for nobody.

Practical information about the return trip

The return trip from Mariehamn is the responsibility of the individual participants.

However, to simplify things, everyone except those who have explicitly indicated a contrary preference will be booked on Viking Line sailing from Mariehamn on Saturday May 25 at 14.25 (Finnish time), arriving in Stockholm at 18.55 (Swedish time). If you continue elsewhere by air travel from Stockholm that same evening you should count at least one hour from the ferry terminal to the airport (to be on the safe side).

Main sponsors:

Wednesday, 22 May

Ferry session

1 Introduction

Opening of the workshop and practical information for the Stockholm-Mariehamn trip.

Speaker: Mikko Alava (HUT, Espoo, Finland)

2 Theory and Practice of Evolutionary Couplings - Part 1

Attributes of living systems are constrained in evolution. An alternative to the analysis of conserved attributes ('characters') is analysis of functional interactions ('couplings') that cause conservation. A quantitative theory of evolutionary couplings may be widely applicable to biological and technical evolution at different scales of phenomena. In a particularly interesting application, evolutionary couplings in proteins in the form of amino acid pairwise covariation across a protein family, can be used to computationally fold proteins, to predict oligomerization, functional sites and paths, and functionally distinct conformational states. For example, protein residue-residue couplings, used as input to distance geometry and molecular dynamics tools, are sufficient to generate good all-atom models of proteins from different fold classes, ranging in size from 50 to more than 300 residues. The evolutionary couplings in proteins, extracted from the rich evolutionary sequence record, provide insight into essential interactions constraining protein evolution and, with the rapid rise in large-scale sequencing, are likely to facilitate a comprehensive survey of the universe of protein structures by a combination computational and experimental technology. Since March 2013, a web service at www.EVfold.org provides a tool for the analysis of covaration in proteins with respect to functional interactions and structural distance constraints. Project leaders: Debora Marks, Harvard Medical School and Chris Sander, Memorial Sloan-Kettering Cancer Center. See http://bit.ly/tob48p (PDF) and www.evfold.org.

Speaker: Prof. Chris Sander (MSKCC)

17:30 Break

3 Cell state transitions in mammalian cells: From development to cancer

The transition from one cell phenotype to another represents a non-genetic (mutation-less) change of phenotype and plays a central role not only in metazoan development (differentiation) but also in cancer. From a systems dynamics point of view such a (typically quasi-discrete) “switch” of a cell in the stable steady state S(A) to state S(B) can be formalized as a transition between high-dimensional attractor states with respect to the network state S(t)=[x_1(t), x_2(t), ..x_N(t)] where x_i(t) is the expression level gene i of the gene network that consists of N genes and drives the transition. Such elementary state transition processes, however, are not simply first-order probabilistic transition as measurements in living cells reveal, but exhibit interesting kinetics. Notably, tissues and tumors consist of populations of cells which, even if they are clonal (isogenic), exhibit a complex phenotype dynamics that is not described by an ensemble of replicates of identical cells. By contrast, cell population level state transitions are influenced by stochastic cell-cell variability in transition rates, by cell-cell interactions (cooperation) and by differential growth rates between the states. These features give rise to non-trivial, rarely considered dynamics of phenotype change. In this talk I will emphasize the experimental observables and show examples of cell phenotype switching and counterintuitive properties. The goal is to provide physicists a “feel” of the state–of-the-art in experimental biology of mammalian cell dynamics. I will discuss the profound implications of cell state transitions for understanding cancer as a disease of quantitative cell population dynamics rather than qualitative genetics, as commonly thought. This is important at a time when (in the U.S.) broad governmental efforts to involve physicists in cancer research are underway in an attempt to bring in fresh ideas and lead us out of the stalemate in the war on cancer.

Speaker: Prof. Sui Huang (Institute for Systems Biology)

20:00 Dinner

Thursday, 23 May

Thursday morning

4 Clustering and optimal arrangement of enzymes in reaction-diffusion systems

Enzymes within biochemical pathways are often colocalized, yet the consequences of specific spatial enzyme arrangements remain poorly understood. We study the impact of enzyme arrangement on reaction efficiency within a reaction-diffusion model. The optimal arrangement transitions from a cluster to a distributed profile as a single parameter, which controls the probability of reaction versus diffusive loss of pathway intermediates, is varied. We introduce the concept of enzyme exposure to explain how this transition arises from the stochastic nature of molecular reactions and diffusion.

Speaker: Prof. Ulrich Gerland (LMU)

5 In vivo single-molecule kinetics of activation and subsequent activity of bacterial promoters

Recent developments of single molecule detection techniques have made possible to observe transcription and translation, one event at a time, in live cells. Here, we report measurements of time intervals between consecutive transcription events from several promoters in live Escherichia coli cells, which were obtained using a single-RNA detection technique by MS2-GFP tagging. From these, we show that, surprisingly, the kinetics of production of transcripts is consistent with multi-step, sub-Poissonian process for all promoters tested. Next, we report recent measurements of the waiting time for the production of the first RNA under the control of PBAD promoter following induction by arabinose. These provide the first direct estimation of the contribution of intake times of inducers by the cells to the cell to cell diversity in RNA numbers. We find that the kinetics of the arabinose intake system affects mean and diversity in RNA numbers, long after induction. We observed the same effect on Plac/ara-1 promoter, inducible by arabinose or IPTG. Importantly, the distribution of waiting times of Plac/ara-1 is indistinguishable from that of PBAD, if and only if induced by arabinose alone. We conclude that inducer-dependent waiting times affect mean and cell-to-cell diversity in RNA numbers long after induction, suggesting that intake mechanisms have non-negligible effects on the phenotypic diversity of cell populations in natural, fluctuating environments.

Speaker: Prof. Andre S. Ribeiro (TUT)

10:15 Break

6 Sex as Gibbs sampling: a Markov Chain Monte Carlo model of evolution

Evolution by natural selection is a learning algorithm of remarkable power. We propose a simple, general abstract model of evolution for which the mutation-selection equilibrium can be given in closed form for arbitrary fitness functions. The model is a modification of the Moran process for evolution with overlapping generations.
Computational models of evolution -- such as genetic algorithms -- are usually specified as processes, in which breeding with recombination, mutation, and selection is repeatedly applied to produce a sequence of populations: this sequence of populations is a Markov chain, and the stationary distribution of this Markov chain is the mutation-selection equilibrium distribution over populations. Unfortunately, these Markov chains are typically not easy to analyse because they do not satisfy the detailed balance condition and their stationary distributions do not in general have a simple closed form.
In contrast, we specify a probability model explicitly by considering the population of genomes as a Markov random field. We then observe that a standard Markov chain Monte Carlo (MCMC) sampling method -- blocked Gibbs sampling within Metropolis Hastings -- can naturally be regarded as a genetic algorithm. The Markov chain of populations satisfies detailed balance. Although this result is rather simple, we have so far been unable to find it in the literature.
The implications seem quite deep. The stationary distribution factorises as the product of two terms: the first term is the probability of generating the population by pure genetic drift with no selection; the second term is the product of the fitnesses of the genomes. This expression is analogous to that for a Bayesian posterior distribution, and a population in equilibrium is analogous to a sample from a Bayes posterior distribution.
A practical computational consequence is that for evolutionary simulations, it is possible to apply alternative MCMC sampling methods to achieve faster convergence to the same stationary distribution. In other words, one can do 'evolution' with non-biological sampling methods that may be much faster for some interesting cases.

Speaker: Prof. Chris Watkins (Royal Holloway)

7 Collective cell motion

In different biological processes, cells move in a coordinated way. Several experiments have quantitatively investigated this phenomenon. We will describe a simple model of interacting random walkers that we have developed to phenomenologically describe these results, based on data obtained in P Silberzan's lab (Institut Curie, Paris). The model serves to suggest some biological mechanisms underlying collective cell motion and its guidance by leaders cells. It also helps to analyze other observed phenomena, such as pulsatile behavior in confined cell assemblies.

Speaker: Prof. Vincent Hakim (ENS)

12:15 Lunch & afternoon break

Thursday afternoon

16:00 Coffee & tea

8 Acceleration of epidemic outbreaks by long-range dispersal

The spreading of evolutionary novelties across populations is the central element of adaptation. Unless population are well-mixed (like bacteria in a shaken test tube), the spreading dynamics not only depends on fitness differences but also on the dispersal behavior of the species. Spreading at a constant speed is generally predicted when dispersal is sufficiently short-ranged. However, the case of long-range dispersal is unresolved: While it is clear that even rare long-range jumps can lead to a drastic speed-up, it has been difficult to quantify the ensuing stochastic growth process. Yet, such knowledge is indispensable to reveal general laws for the spread of modern human epidemics, which is greatly accelerated by the human aviation. We present a simple self-consistent argument supported by simulations that accurately predicts evolutionary spread for broad distributions of long distance dispersal. In contrast to the exponential laws predicted by deterministic 'mean-field' models, we show that growth is either according to a power-law or a stretched exponential, depending on the tails of the dispersal kernel. We also find that the actual fitness advantage of the mutants has a surprisingly small impact on the spreading dynamics. This conflicts with the paradigm that the rapidity of a selective sweep is a good measure for the selective advantage of the spreading variant. Due to the simplicity of our model, which lacks any complex interactions between individuals, we expect our results to be applicable to a wide range of spreading processes.

Speaker: Prof. Oskar Hallatschek (MPI)

9 Coupled Oscillators and Arnold Tongues in Cell Dynamics

Oscillating genetic patterns have been observed in networks related to the transcription factors NFkB, p53 and Hes1 [1]. We identify the central feed-back loops and found oscillations when time delays due to saturated degradation are present. By applying an external periodic signal, it is sometimes possible to lock the internal oscillation to the external signal. For the NF-kB systems in single cells we have observed that the two signals lock when the ration between the two frequencies is close to basic rational numbers [2]. The resulting response of the cell can be mapped out as Arnold tongues. When the tongues start to overlap we observe a chaotic dynamics of the concentration in NF-kB [2]. Oscillations in some genetic systems can be triggered by noise, i.e. a linearly stable system might oscillate due to a noise induced instability. By applying an external oscillating signal to such systems we predict that it is possible to distinguish a noise induced linear system from a system which oscillates via a limit cycle. In the first case Arnold tongues will not appear, while in the second subharmonic mode-locking and Arnold tongues are likely [3].
[1] B. Mengel, A. Hunziker, L. Pedersen, A. Trusina, M.H. Jensen and S. Krishna, "Modeling oscillatory control in NF-kB, p53 and Wnt signaling", Current Opinion in Genetics and Development 20, 656-664 (2010).
[2] M.H. Jensen and S. Krishna, "Inducing phase-locking and chaos in cellular oscillators by modulating the driving stimuli", FEBS Letters 586, 1664-1668 (2012).
[3] N. Mitarai, U. Alon and M.H. Jensen, "Entrainment of linear and non-linear systems under noise", Chaos, to appear (2013).

Speaker: Prof. Mogens Hogh Jensen (NBI)

18:00 Break

10 Using laboratory evolution experiments to study the genetic basis of cancer drug resistance in yeast

Understanding the molecular basis of the adaptive evolution of a population has relevance for important biological questions. For example, the problem of identifying genetic variants which underlie drug resistance, a question of importance for the treatment of pathogens, and of cancer, can be understood as a matter of inferring selection. A key problem in discovering variants under positive selection is the complexity of the underlying evolutionary dynamics, which may involve an interplay between several contributing processes, including mutation, recombination and genetic drift. Fortunately, technological advances driven by next generation sequencing are making it possible to systematically follow across time how the genomic composition of a population evolves. However, such data needs new quantitative methods to fulfil its potential. We here present our ongoing work on how to use time-resolved sequence data to draw inferences about the evolutionary dynamics of a population under study. More specifically, we describe an analysis of a laboratory evolution experiment where a yeast cross was exposed to a number of cancer drugs to study the genetic basis how populations respond to such external stresses. This work is a collaboration project with Gianni Liti Lab (Institute of Research on Cancer and Ageing of Nice).

Speaker: Prof. Ville Mustonen (Sanger Centre)

Friday, 24 May

Friday morning

11 Theory and Practice of Evolutionary Couplings - Part 2

Attributes of living systems are constrained in evolution. An alternative to the analysis of conserved attributes ('characters') is analysis of functional interactions ('couplings') that cause conservation. A quantitative theory of evolutionary couplings may be widely applicable to biological and technical evolution at different scales of phenomena. In a particularly interesting application, evolutionary couplings in proteins in the form of amino acid pairwise covariation across a protein family, can be used to computationally fold proteins, to predict oligomerization, functional sites and paths, and functionally distinct conformational states. For example, protein residue-residue couplings, used as input to distance geometry and molecular dynamics tools, are sufficient to generate good all-atom models of proteins from different fold classes, ranging in size from 50 to more than 300 residues. The evolutionary couplings in proteins, extracted from the rich evolutionary sequence record, provide insight into essential interactions constraining protein evolution and, with the rapid rise in large-scale sequencing, are likely to facilitate a comprehensive survey of the universe of protein structures by a combination computational and experimental technology. Since March 2013, a web service at www.EVfold.org provides a tool for the analysis of covaration in proteins with respect to functional interactions and structural distance constraints. Project leaders: Debora Marks, Harvard Medical School and Chris Sander, Memorial Sloan-Kettering Cancer Center. See http://bit.ly/tob48p (PDF) and www.evfold.org.

Speaker: Prof. Debora Marks (Harvard)

12 Solvable model for template coexistence in protocells

Compartmentalization of self-replicating molecules (templates) in protocells is a necessary step towards the evolution of modern cells. However, coexistence between distinct template types inside a protocell can be achieved only if there is a selective pressure favoring protocells with a mixed template composition. Here we study analytically a group selection model for the coexistence between two template types using the diffusion approximation of population genetics. The model combines competition at the template and protocell levels as well as genetic drift inside protocells. At the steady state, we find a continuous phase transition separating the coexistence and segregation regimes, with the order parameter vanishing linearly with the distance to the critical point. In addition, we derive explicit analytical expressions for the critical steady-state probability density of protocell compositions.

Speaker: Prof. Maurizio Serva (University of Rio Grande do Norte - Brazil)

10:15 Coffee & tea

13 Quantifying immune receptor diversity

Recognition of pathogens relies on the diversity of immune receptor proteins. Recent experiments that sequence the entire immune cell repertoires provide a new opportunity for quantitative insight into naturally occurring diversity and how it is generated. The generation process is implemented via a series of stochastic molecular events involving gene choices and random nucleotide insertions between, and deletions from, genes. I will describe how we can attempt to quantify the diversity of the receptors formed in this complex process and point to the origins of diversity in these sequences.

Speaker: Prof. Aleksandra Walczak (ENS)

14 Dynamical Regulatory Networks: from function to structure and back

Even though many vital cellular process are known to be tightly controlled, the associated regulatory network can vary significantly from species to species. To explore the space of all possible networks or circuits that implement a specified control or function, we use Markov Chain Monte Carlo (MCMC) and biophysical modeling. First, we show that the molecular encoding of the genetic interactions naturally leads to sparse networks. Second, we consider different types of cellular functions; our MCMC sampling then reveals that the structure of the functional networks is shaped by these functions. Finally, by examining the emergent structural features found such as edge usage or network motifs, we uncover pieces of regulatory logic responsible for functional capabilities in these networks.

Speaker: Prof. Olivier Martin (Universite Paris Sud - Orsay)

12:15 Lunch & afternoon break

Friday afternoon

16:00 Coffee & tea

15 A space odyssey in cell signaling

Experiments in recent years have vividly demonstrated that the membrane is a highly heterogenous environment. A key example is the partitioning or clustering of proteins via lipid domain formation or cytoskeleton-induced corralling. In this talk, I will show using theory and computer simulations that protein clustering can enhance biochemical information transmission by removing correlations in the signal and by linearizing the response. Yet, protein partitioning can also impede signaling when the partitions become too small. This trade-off leads to an optimal protein cluster size that agrees quantitatively with experiment. Our results suggest that molecular partitioning or clustering is not merely a consequence of the complexity of subcellular structures, but also plays an important functional role in cell signaling.

Speaker: Prof. Pieter Rein ten Wolde (AMOLF)

16 Competing endogenous RNAs: A novel microRNA-based mechanism of gene regulation

It has been recently proposed that competitive endogenous RNAs (ceRNAs) sequester microRNAs to regulate mRNA transcripts containing common microRNA recognition elements (MREs). The ceRNA hypothesis stems from the observation that RNA transcripts can communicate with each other through a recently discovered mechanism. MicroRNAs are tiny snippets of RNA (~22nt long) which negatively regulate target gene expression via translational inhibition or transcript cleavage. A microRNA may target many different transcripts, and conversely, individual transcripts may be bound by multiple different microRNAs. In addition to this conventional microRNA → RNA function, it has recently been established that a reversed RNA → microRNA dimension exists, whereby ceRNAs regulate transcript expression via competition for common microRNAs, with microRNA response elements (abbreviated MREs) as the building blocks of this ‘RNA language’. Both phenomenological and theoretical aspects will be addressed.

Speaker: Prof. Andrea Pagnani (Human Genetics Foundation (HuGeF))

Poster session

Saturday, 25 May

Saturday morning

17 Dislocation Mediated Elongation of Bacteria

Recent experiments have revealed a remarkable growth mechanism for rod-shaped bacteria: specialized proteins associated with cell wall elongation move at constant velocity in clockwise and counterclockwise directions on circles around the cell circumference. We argue that this machinery attaches to dislocations in the ordered peptidoglycan cell wall, and study theoretically the dynamics of these interacting defects on the surface of a cylinder. Unlike the dislocations typical in materials science, the motion is predominantly climb (glycan strand extension) instead of glide. The activated motion of these dislocations and the resulting dynamics within a simple kinetic model show surprising effects arising from the cylindrical geometry, with important implications for bacterial growth. Recent experiments revealing plastic deformation of bacterial cell walls in a hydrodynamic flow will be presented as well.

Speaker: Prof. David R. Nelson (Harvard University and Niels Bohr Institute)

18 Diffusion of public goods in bacterial colonies, and its impact on cooperation

The maintenance of cooperation in populations where public goods are equally accessible to all, but inflict a fitness cost on its sole producers, is a long-standing puzzle of evolutionary biology. An example of such a scenario is the secretion of siderophores by bacteria into their environment in order to fetch soluble iron. In a well-mixed culture, these molecules diffuse rapidly, such that all bacteria experience the same concentration, giving an advantage to potential cheaters—bacteria that do not produce the public good but benefit from it. However, on solid substrates, bacteria form dense and packed colonies, which may alter the diffusion dynamics through cell-cell contact interactions. Based on fluorescence microscopy data tracking the concentration of pyoverdine in P. aeruginosa microcolonies, we propose a model of local exchange of the public good between neighboring cells. The model is equivalent to a model of diffusion on the network of adjacent cells. The model quantitatively explains the formation of a concentration gradient, and reproduces the observed variability of concentration in the population, as well as its spatial and temporal correlation functions, with only two parameters. In addition, we show that this local trafficking modulates the growth rate of individual cells. Using computer simulations of population dynamics, we show that this modulation suffices to maintain cooperation against the invasion of cheaters. Our results give a physical basis that explains the stability of public goods production in packed colonies.

Speaker: Prof. Thierry Mora (ENS)

10:15 Coffee & tea

19 Quantifying contributions of mutations and homologous recombination to E. coli genomic diversity

Understanding evolutionary dynamics of bacterial genomes is of great importance in microbiology, microbial ecology, and epidemiology. Collective effects play an important role in shaping this dynamics due to ubiquitous horizontal gene transfer between different members of bacterial population. We recently interpreted the variability of 37 fully sequenced genomes of E. coli and Shigella strains within a neutral evolutionary framework and quantified contributions of random mutations in the clonal frame and homologous recombination to the observed diversity of the basic genome of this species. Distinct signatures of spatial SNP distributions allowed us to separate clonal, and recombined chromosomal regions when comparing genomes of different strains. We estimated that in E. coli for every SNP brought by clonal frame point mutations recombination brings on average 5-8 SNPs. Our methods also allowed us to estimate the number of strains frequently exchanging genetic material with each other to be around 3 x 10^8. This is consistent with previously published estimates of the effective population size of E. coli as well as our own analysis of the MLST data. The observed correlations between SNPs within recombined regions indicate phage-mediated gene transduction as the likely mechanism of exchange of genomic segments between strains. While the vast majority of the basic genome diversity is consistent with purely neutral model in which mutations and recombination events are random we also identified around 30 out of nearly 3900 1kb basic genome where the observed diversity is significantly higher than expected this model. Some of these regions contain genes encoding proteins under positive selection such as biosynthetic enzymes in the O-antigen region, fimbrial like adhesins, etc., while other regions have not been studied before.

Speaker: Prof. Sergei Maslov (Brookhaven National Laboratory)

20 Pareto optimality in evolution and the geometry of phenotype space

NO ABSTRACT

Speaker: Prof. Alon Uri (Weizmann Institute of Science)

12:15 Lunch