A multicolour polymer model facilitating prediction of 3D structure and transcriptional activity of human chromosomes

Giuseppe Negro, Università degli studi di Bari and INFN

Abstract: Within each human cell, different kinds of RNA polymerases and a panoply of transcription factors bind chromatin to simultaneously determine 3D chromosome structure and transcriptional programme. Experiments show that, in some cases, different proteins segregate to form specialised transcription factories; in others they mix together, binding promiscuously the same chromatin stretch. We use molecular dynamics simulations to study a polymer model for chromosomes accounting for multiple types (“colours”) of chromatin-binding proteins. Our multi-colour model shows the spontaneous emergence of both segregated and mixed clusters of chromatin-bound proteins, depending mainly on their size, thereby reconciling the previous experimental observations.  Additionally, the model leads to non-trivial spatio-temporal correlations between different transcription units, dependent on the underlying chromosome  structure. Small-world networks emerge; in these, positive and negative correlations in activities of transcription units provide simple explanations of why adjacent units in large domains are co-transcribed so often. We also explain how local genome edits induce distant omnigenic and pangenomic effects, and develop ways to predict activities of all transcription units on human chromosomes. 

[1] 3D polymer simulations of genome organisation and transcription across different chromosomes and cell types. M Semeraro, G Negro, A Suma, G Gonnella, D Marenduzzo. Physica A: Statistical Mechanics and its Applications 625, 129013, (2023).

 [2] A unified-field theory of genome organization and gene regulation.  G. Negro, M. Semeraro, P. R. Cook, D. Marenduzzo. arXiv preprint arXiv:2308.02861 (2023).

 [3] A multicolour polymer model for the prediction of 3D structure and transcription in human chromatin. M. Semeraro, G. Negro, A. Suma, G. Gonnella, P. R. Cook, D. Marenduzzo bioRxiv 2023.01.16.524198, (2023).

 14th September 2023 - 16.30 Rome Time

Training neural networks with structured noise improves classification and generalization

Enrico Ventura, Sapienza Rome University - École Normale Supérieure  Paris

Abstract:Attractor neural networks have been extensively used for modeling associative memory, as biologically inspired devices that learn concepts by optimizing their connectivity structure. Various algorithms have been developed to modify the interactions between neurons, aiming to achieve memory retrieval and generalization capabilities, which are manifested as finite basins of attraction containing the memories. The training-with-noise algorithm, initially proposed by Gardner and collaborators, introduces noise injection into the training dataset, contributing to understand the positive impact of noise on enhancing the system's generalization capabilities. This same idea is at the basis of a whole class of regularization techniques that are widely applied in the field of artificial neural networks today. In this talk, we present how the memory performance of attractor neural networks can be significantly enhanced by adding internal dependencies into the noisy training data, allowing to achieve perfect classification and approach maximal basins of attraction. The criterion for finding the optimal structure of training noise will be derived and used in practice. We will also prove that the so-called unlearning rule coincides with the training-with-noise algorithm when noise is maximal and data are fixed points of the network dynamics.

References:

E. Gardner, D.J. Wallace, and N. Stroud. J. Phys. A, 22: 2019-2030, 1989.

M. Benedetti and E. Ventura. arXiv:2302.13417, Under review, 2023.

M. Benedetti, E. Ventura, E. Marinari, G. Ruocco and F. Zamponi. J. Chem. Phys., 156: 104107, 2022.

K.Y.M. Wong, D. Sherrington. Phys. Rev. E, 47(6): 4465–4482, 1993.

 13th July 2023 - 16.30 Rome Time

Response Theory, Reaction Coordinates and Critical Phenomena for Noisy Interacting Systems

Niccolò Zagli, Nordic Institute for Theoretical Physics

Abstract:In this talk, I will present our latest results on the close link between response theory, reaction coordinates and critical phenomena for noisy systems with mean field interactions.

Such systems are routinely used to model collective emergent behaviours in multiple areas of social and natural sciences as they exhibit, in the thermodynamic limit, continuous and discontinuous phase transitions.

I will show that the perspective of Response Theory, that aims at establishing a link between natural and forced variability of general physical systems, is particularly useful to understand the physical mechanisms establishing critical phenomena in interacting systems. 

Firstly, I will show how to define a set of reaction coordinates for the system starting from the coupling structure among the microscopic agents. 

Secondly, I will show that such reaction coordinates prove to be proper nonequilibrium thermodynamic variables as they carry information on correlation, memory properties and resilience properties of the system. 

In particular, the investigation of response properties of the reaction coordinates allows to identify and characterise phase transitions of the system as they manifest as singular values of the susceptibility associated to such thermodynamic variables.

 8th June 2023 - 16.30 Rome Time

Diversity of information pathways drives scaling and sparsity in real-world networks

Arsham Ghavasieh, Trento University

Abstract: Empirical complex systems must differentially respond to external perturbations and, at the same time, internally distribute information to coordinate their components. While networked backbones help with the latter, they limit the components' individual degrees of freedom and reduce their collective dynamical range. Here, we show that real-world networks are formed to optimize the gain in information flow and loss in response diversity.

Encoding network states as density matrices, we demonstrate that such a trade-off mathematically resembles the thermodynamic efficiency characterized by heat and work in physical systems. Our findings explain, analytically and numerically, the sparsity and the empirical scaling law observed in hundreds of real-world networks across multiple domains. We show, through numerical experiments in synthetic and biological networks, that ubiquitous topological features such as modularity and small-worldness emerge to optimize the above trade-off for middle- to large-scale information exchange between system's units. Our results highlight that the emergence of some of the most prevalent topological features of real-world networks have a thermodynamic origin.

  11th May 2023 - 16.30 Rome Time

Glass and pseudo-localization transitions in the Mode-Locked p-spin model for Random Lasers

Jacopo Niedda, Università di Roma "Sapienza"

Abstract: Optical waves in active disordered media display the typical phenomenology of complex systems. Several spectral shots taken from the same piece of material in the lasing regime display strong fluctuations in the position of the intensity peaks, suggesting that there is no specific mode which is preferred in the amplification, but depending on the initial state, with the disorder kept fixed, the modes gaining the highest intensity change every time. In order to explain this behaviour, a spin-glass model has been developed, where the light modes are described as non-linearly interacting phasors on the so-called mode-locked diluted graph [1]. 

The specific mode-coupling selection rule, which naturally emerges in the study of lasing modes dynamics, impairs the analytical solution of the model out of the narrow bandwidth limit, where the interaction network is fully connected. In this talk we present recent results from numerical simulations of the mode-locked glassy random laser. A phenomenology compatible with a glass transition is revealed from the divergence of the specific heat and the non-trivial structure of the Parisi overlap distribution function. By means of a refined finite-size scaling analysis of the critical region, the transition is assessed to be compatible with a mean-field universality class [2]. 

A pseudo-localization transition to a phase where the intensity of light is neither properly localized on a single mode nor equiparted among all the modes is revealed from the measure of the inverse participation ratio and of the spectral entropy [3]. The two transitions occur at the same temperature as different manifestations of the same underlying phenomenon, the breaking of ergodicity.

[1] F. Antenucci, C. Conti, A. Crisanti and L. Leuzzi, Phys. Rev. Lett. 114, 043901 (2015).

[2] J. Niedda, G. Gradenigo, L. Leuzzi and G. Parisi, arXiv:2210.04362, (2022).

[3] J. Niedda, L. Leuzzi and G. Gradenigo, arXiv:2212.05106 (2022).

  13rd April 2023 - 16.30 Rome Time

Optimal policies for Bayesian olfactory search

Robin Heinonen, University of Rome - Tor Vergata

Abstract: Many flying insects and animals depend on a remarkable ability to track the source of an odor which is advected by a turbulent flow. This olfactory search problem is rendered especially difficult by the intermittency intrinsic to turbulence, and it requires complex search strategies which properly leverage infrequent odor detections. In this seminar, I will briefly outline the physics underlying the olfactory search problem and why naive gradient-based strategies will not succeed. Then I will introduce the partially-observable Markov decision process (POMDP) formalism and discuss how to obtain policies which are optimal with respect to the search time. These will be compared to well-known heuristic strategies on data from direct numerical simulations. Finally, I will discuss current and future directions for research.

   9th March 2023 - 16.30 Rome Time

Ion channels in critical membranes: clustering, cooperativity, and memory effects

Antonio Suma, Università degli Studi di Bari 

Abstract:The response and conformational changes of voltage-gated ion channels are well understood at a single-molecule level, but there is limited knowledge about how these channels interact with the lipids and with one another in their native environment. In this study, we propose that interactions between ion channels mediated by the surrounding lipid environment can explain the observed behaviors of cooperativity and hysteresis. These have been suggested to give rise to multistability of membrane potentials and thus to ”cellular memory”. We develop a theoretical model based on the statistical mechanics of interacting, diffusing agents with internal degrees of freedom that are subject to an external field. This model allows us to understand several poorly understood aspects of ion channels behavior, including the non-Markovian character of single channel currents.

Date:  9th February  2023 - 17.00 Rome Time

Polymer physics of chromosome spatial organization

Mattia Conte, Università di Napoli "Federico II"

Abstract: In the nucleus of cells, chromosomes have been discovered to self-organize into a complex spatial architecture that serves vital functional purposes as, for instance, genes have to establish specific physical contacts with their distal DNA regulators to control transcriptional activities. Those discoveries have changed our view of the genome, which is no longer considered to be just a 1D linear sequence of nucleotides to be deciphered, but rather a complex 3D structure whose arrangement defines the fate of the cell. However, how the system self-assembles to shape the folding of our genome and its functions is only poorly understood. In this talk, I discuss principled models of interacting polymers from statistical mechanics to investigate the mechanisms whereby distal DNA sequences recognize and interact with each other. By combining polymer physics, machine learning and computer simulations, I show that chromosome spatial organization is controlled at the single-molecule level by thermodynamic mechanisms of phase transitions, which spontaneously establish contact or segregation between specific genomic sites, such as genes and their regulators [1]. Those theories are validated against independent experiments (e.g., Hi-C, SPRITE, GAM and super-resolution microscopy) [2,3], opening to new tools for real-world applications, such as the predictions of the effects of disease-associated mutations, linked to congenital disorders or cancer, on genome 3D structure [4,5].

[1] Conte et al. Nature Communications, 11, 3289 (2020)

[2] Conte et al. Nature Communications 13, 4070 (2022)

[3] Fiorillo, Musella, Conte et al. Nature Methods, 18, 482-490 (2021)

[4] Huang et al. Nature Genetics, 53, 1064–1074 (2021)

[5] Bianco et al. Nature Genetics 50, 662 (2018)

Date:  12th January 2023 - 16.30 Rome Time

The role of higher-order interactions in the dynamics of social contagion and norm change

Iacopo Iacopini, Central European University, (Wien)

Abstract:  Dynamical processes that emulate human behavior have been the focus of many studies, where social relationships and interactions are typically considered as an underlying structure. As shown recently in various contexts, it is moreover important to take into account the fact that individuals do not interact only in pairs but also in larger groups [1]. Social interactions are indeed a natural testing ground for higher-order approaches. In this talk, I will discuss the effects of considering group interactions on the dynamics of social systems. In particular, I will generalize two models of social contagion and norm evolution, initially introduced and studied on graphs, and now extended as dynamical processes on hypergraphs. Leveraging real-world interaction data and analytical insights, I will show the emergence of novel phenomena such as discontinuous transitions and critical mass effects induced by higher-order interactions [2, 3]. After allowing for the presence of a committed minority, I will show that the ability of committed individuals to overturn an existing norm is non-monotonic in the number of participants in higher-order interactions [4]. These results provide theoretical support to the observation that extremely small minorities can overcome the opinion of the corresponding large majority of the population. Overall, the findings confirm the relevance of exploring network representations beyond pairwise interactions when modeling social phenomena [5].

References:

[1] Battiston, F., Cencetti, G., Iacopini, I., Latora, V., ... & Petri, G. (2020). Networks beyond pairwise interactions: structure and dynamics. Physics Reports, 874, 1-92.       [2] Iacopini, I., Petri, G., Barrat, A., & Latora, V. (2019). Simplicial models of social contagion. Nature Communications, 10(1), 1-9.                                             [3] St-Onge, G., Iacopini, I., Latora, V., Barrat, A., Petri, G., Allard, A., & Hébert-Dufresne, L. (2022). Influential groups for seeding and sustaining nonlinear contagion in heterogeneous hypergraphs. Communications Physics, 5(1), 1-16.           [4] Iacopini, I., Petri, G., Baronchelli, A., & Barrat, A. (2022). Group interactions modulate critical mass dynamics in social convention. Communications Physics, 5(1), 1-10.                       [5] Battiston, F., Amico, E., Barrat, A., Bianconi G., Ferraz de Arruda, G., Franceschiello, B., Iacopini, I., ... & Petri, G. (2021). The physics of higher-order interactions in complex systems. Nature Physics, 17(10), 1093-1098.

Date:  15th December 2022 - 16.30 Rome Time

Optimal transport problem in complex systems

Dario Mazzilli, CREF (Centro Ricerche Enrico Fermi, Rome)

Abstract:  In general, optimal transport (OT) theory can be used in systems where there is an optimization problem under constraints. Indeed, it is not hard to find examples in many complex systems where agents are maximizing/minimizing some gain/cost function given limited resources or time. Following [1], a simple application of OT to Economics is represented by the worker-job assignment. The solution finds the answer to the question relating the best partitioning of a mass (a.k.a. the number) of workers knowing their skills, with respect to the job requirements. This assignment can be formulated by an OT problem. A well known property of OT solution is its network topology. In facts, it is easy to prove, see for example [2], that the optimal mass distribution is a sparse matrix representing a tree of the bipartite (e.g. worker-job) network. Diversification and nestedness are considered important properties of many complex systems such as ecological and economic systems [3]. Usually, they are linked to the resilience of the systems against shocks or external attacks. While the binary structure of such networks shows indeed nestedness and many redundant paths between nodes, the weighted matrices (export flows in world trade web or frequencies of interaction in plant-pollinator systems) are much sparser in terms of concentration of weights, hinting an OT-like solution. We explore the idea that many of these systems have indeed an OT problem running onto a binary network of potential interactions.

References:

[1] Alfred Galichon. Optimal Transport Methods in Economics. Princeton University Press, 2016.

[2] Richard A Brualdi. Combinatorial matrix classes, volume 13. Cambridge University Press, 2006.

[3] Manuel Sebastian Mariani, Zhuo-Ming Ren, Jordi Bascompte, and Claudio Juan Tessone. Nestedness in complex networks: observation, emergence, and implications. Physics Reports, 813:1–90, 2019. 

Date:  10th November 2022 - 16.30 Rome Time

Approaching optimal memory retrieval in Hopfield-like networks

Marco Benedetti, "La Sapienza" University, Rome

Abstract:  One of the most important tasks performed by the brain is Associative Memory, namely the ability to recognize similarities between slightly distorted versions of a given stimulus, tracing them back to one single archetype. Hopfield-like neural networks are a form of Artificial Intelligence designed to mimic this ability. They consist of a collection of binary spins endowed with a stochastic dynamics, which must be such as to create finite basins of attraction around a predetermined set of configurations (memories). When initialized to a configuration similar to one of the memories, the network will be driven towards the corresponding attractor, resulting in error correcting performance. Among the possible strategies to build such a dynamics, the linear perceptron algorithm and Hebbian unlearning are two of the most influential. The first is a supervised learning algorithm, meaning that it needs to be fed explicitly the information about the patterns to be learned. The second is unsupervised, and exploits the information contained in the attractor landscape generated by Hebb's biologically inspired learning rule. In this talk, I will introduce both paradigms, illustrate some unexpected similarities in the performance of these apparently very different algorithms, and speculate on a possible explanation for them.

Main references:
- Benedetti, Ventura, Marinari, Ruocco, Zamponi, Supervised perceptron learning vs unsupervised Hebbian unlearning: approaching optimal memory retrieval in Hopfield-like networks, J. Chem. Phys. 156, 104107 (2022)
- D. Amit, H. Gutfreund, and H. Sompolinsky, Storing infinite numbers of patterns in a spin-glass model of neuralnetworks, J. Stat. Phys. 15, 1530 (1985)
- J. L. van Hemmen, L. Ioffe, R. K ̈uhn, and M. Vaas, Increasing the efficiency of a neural network through unlearning, Physica A 163, 386 (1990)

Date:  12th May 2022 - 16.30 Rome Time

Tensor Network methods for Lattice Gauge Theories

Giuseppe Magnifico, Università di Padova

Abstract:  Gauge theories are of paramount importance in our understanding of fundamental constituents of matter and their interactions, ranging from high-energy particle physics to low-temperature quantum many-body physics. However, the complete characterization of their phase diagrams and the full understanding of non-perturbative effects are still debated, especially at finite charge density, mostly due to the sign-problem affecting Monte Carlo numerical simulations. In recent years, a complementary numerical approach, Tensor Networks (TN) methods, in strict connection with emerging quantum technologies, have found increasing applications for studying Lattice Gauge Theories (LGTs) in low-dimensional systems. In this talk, I will present some recent results concerning the extension of TN algorithms to high-dimensional LGTs including dynamical matter. In particular, I will focus on their application to a compact Quantum Electrodynamics at zero and finite charge densities, addressing questions such as the characterization of collective phases of the model, the presence of confining phases at large gauge coupling, and the study of charge-screening effects.

Main references:
G. Magnifico, T. Felser, P. Silvi, and S. Montangero, Nature Communications 12, 3600 (2021).
M. Rigobello, S. Notarnicola, G. Magnifico, and S. Montangero, Phys. Rev. D 104, (2021).

Date:  07th April 2022 - 16.30 Rome Time

Resetting in Stochastic Optimal Control 

Francesco Mori, LPTMS, Université Paris-Saclay  

Abstract:  "When in a difficult situation, it is sometimes better to give up and start all over again''. While this empirical truth has been regularly observed in a wide range of circumstances, quantifying the effectiveness of such a heuristic strategy remains an open challenge. In this talk, I will combine ideas from optimal control and stochastic resetting to address this question. The emerging analytical framework allows not only to measure the performance of a given restarting strategy, but also to obtain the optimal policy for a wide class of dynamical systems. This approach, analog to the celebrated Hamilton-Jacobi-Bellman equation, is successfully applied to simple settings and provides the basis to investigate realistic restarting strategies across disciplines.

Reference: B. De Bruyne and F. Mori. "Resetting in Stochastic Optimal Control." arXiv preprint arXiv:2112.11416 (2021).

Date:  10th March 2022 - 16.30 Rome Time

Quantum simulation of lattice gauge theories with Rydberg atoms

Federica Maria Surace, Caltech

Abstract:  Gauge theories are the cornerstone of our understanding of fundamental interactions among particles. Their properties are often probed in dynamical experiments, such as those performed at ion colliders and high-intensity laser facilities. Describing the evolution of these strongly coupled systems is a formidable challenge for classical computers and represents one of the key open quests for quantum simulation approaches to particle physics phenomena. In this talk, I will show how recent experiments done on Rydberg atom chains naturally realize the real-time dynamics of a U(1) lattice gauge theory, at system sizes that are difficult to achieve with classical computational methods.

Date:  10th February 2022 - 16.30 Rome Time

Machine Learning applications in Science

Lorenzo Buffoni, Physics of Information and Quantum Technologies Group, University of Lisbon

Abstract:  Machine learning is a broad field of study, with multifaceted applications of cross-disciplinary breadth that ultimately aims at developing computer algorithms that improve automatically through experience. Recently, scientists have increasingly become interested in the potential of Machine Learning for fundamental research and, to some extent, this is not too surprising, since both Machine Learning algorithms and scientists share some of their methods as well as goals. The two fields are both concerned about the process of gathering and analyzing data to design models that can predict the behavior of complex systems. However, the fields prominently differ in the way their fundamental goals are realized.  Here we will argue, using practical cases and applications from biology, network theory and quantum physics, that the communication between these two fields can be not only beneficial but also necessary for the progress of both of them.

Date:  13th January 2022 - 16.30 Rome Time

Quantum statistics and BKT transition of a shell-shaped superfluid

Andrea Tononi, Laboratoire de Physique Théorique et Modèles Statistiques, Orsay

Abstract:  The development of NASA Cold Atom Laboratory, a space-based facility for ultracold atoms experiments, enabled the routine production of Bose-Einstein condensates in microgravity. The ongoing investigations are focusing on shell-shaped geometries, in which the atoms are confined on a thin ellipsoidal surface with radio frequency-induced adiabatic potentials, and which cannot be obtained in the presence of gravity. We analyze the quantum statistical properties of spherical and ellipsoidal shells, focusing on the phenomena of Bose-Einstein condensation and superfluidity. In particular, we discuss the Berezinskii-Kosterlitz-Thouless transition of a spherical superfluid, driven by the proliferation of quantized vortices, and we analyze the finite-size universal effects that emerge in this system. Our results are a reliable benchmark for the current experimental investigations.

Date:  9th December 2021 - 16.30 Rome Time

Nonequilibrium quantum state preparation with driven systems in engineered baths

Francesco Petiziol, Technical University of Berlin

Abstract:  The progress in the development of quantum simulators has made possible the experimental investigation of phenomena which are hardly, if even, observable in un-controlled quantum systems. The central difficulty in this context is that of finding appropriate control operations which realize, in an effective manner, a Hamiltonian featuring desired properties. Successful techniques in the field, such as Floquet engineering and trotterization, achieve this by using time-periodic modulations and coarse-graining the dynamics. Even when a desired effective Hamiltonian is attained, a fundamental challenge remains in how to efficiently prepare its eigenstates, for instance in order to study ground state physics. In this seminar, I will discuss how interesting nonequilibrium states can be prepared and stabilised by combining periodic driving, on the one hand, with engineered forms of dissipation, on the other. In particular, considering arrays of artificial atoms individually coupled to cavities acting as engineered quantum baths, I will discuss the preparation of Aharonov-Bohm cages, in which quantum interference constrains the dynamics in small subsystems, and of chiral ground state currents. 

Date:  11th November 2021 - 16.30 Rome Time

Unraveling the role of node metadata in network robustness: the feature-based percolation model

Oriol Artime, Fondazione Bruno Kessler (FBK)

Abstract:  Percolation is an emblematic model to assess the robustness of interconnected systems when some of their components are corrupted. It is usually investigated in simple scenarios, such as the removal of the system's units in random order, or sequentially ordered by specific topological descriptors, such as the degree or the betweenness centrality. However, in the vast majority of empirical applications, it is required to dismantle the network following more sophisticated protocols, for instance, by combining topological properties and non-topological node metadata. In this seminar I will introduce a novel mathematical framework that fills this gap: networks are enriched with features and their nodes are removed according to the importance in the feature space. We will discuss features of different nature, from ones related to the network construction to ones related to dynamical processes such as epidemic spreading. In this way, we not only provide a natural generalization of percolation but, more importantly, this framework offers an accurate way to test the robustness of networks in realistic scenarios.
More information can be found at https://www.nature.com/articles/s41467-021-22721-z.

Date:  14th October 2021 - 16.30 Rome Time

Quantum simulation of lattice gauge theories: from models to experimental protocols

Simone Notarnicola, Università degli Studi di Padova

Abstract:  Recent experiments on several experimental platforms, such as ultracold atoms, trapped ions, Rydberg atoms and superconducting circuits, succeeded in realizing quantum many body states at unprecedented sizes, allowing to investigate their static and dynamical properties. These achievements have boosted the search for quantum simulation protocols to experimentally investigate complex quantum models, such as for example lattice gauge theories. Indeed, numerical simulation techniques to investigate models describing fundamental interactions suffers from notorious and long-standing limits, such as the sign problem.

For this reason, developing and exploiting quantum simulation protocols can push forward the investigation of phenomena so far unaccessible via standard classical numerical simulations. In my talk, I will review the state of the art of lattice gauge theories quantum simulation from an experimental and theoretical point of view. I will introduce an Abelian Lattice gauge theory model that can be investigated in realistic quantum simulators and I will conclude by showing a proposal for a Rydberg atom experimental implementation.

Joint project with: Mario Collura, Elisa Ercolessi, Paolo Facchi, Giuseppe Magnifico, Giuseppe Marmo, Simone Montangero, Saverio Pascazio, Francesco Pepe

Date:  9th September 2021 - 16.30 Rome Time

A scaling law describes the spin-glass response in theory, experiments and simulations 

Ilaria Paga, Sapienza Università di Roma

Abstract:  The dynamical arrest found upon cooling glass formers to their glass temperature Tg is a major open problem [1, 2]. In the longstanding description [3], this slowing down is caused by the unbounded expansion of cooperative regions as Tg is approached or as the system is left to age below Tg, which, in turn, leads to growing free-energy barriers. A quantitative description of this process is usually attempted in terms of a correlation length ξ. Unfortunately, in numerical simulations it is extremely dicult to measure the quantities that are easily accessible to experiments (and vice versa), which has led to seemingly irreconcilable approaches to the computation of the correlation length. We were able to solve this dilemma in a framework that completely harmonizes experiments with theory [4]. We conduct a parallel study of non-equilibrium spin-glass dynamics both in an experiment in a CuMn single crystal and in a large-scale simulation of the Ising-Edwards-Anderson (IEA) model carried out on the Janus II custom-built supercomputer [5]. We introduced a scaling law that describes the system's response over its entire natural range of variation.

[1] A. Cavagna, Physics Reports 476, 51 (2009), arXiv:0903.4264.
[2] P. Charbonneau, J. Kurchan, G. Parisi, P. Urbani, and F. Zamponi, Nature Communications 5, 3725 (2014), arXiv:1404.6809.
[3] G. Adam and J. H. Gibbs, The Journal of Chemical Physics 43, 139 (1965).
[4] Q. Zhai et al., "A scaling law describes the spin-glass response in theory, experiments and simulations", (2020), arXiv:2007.03871, Phys. Rev. Lett., submitted for publication, 2020.
[5] M. Baity-Jesi et al., Comp. Phys. Comm 185, 550 (2014), arXiv:1310.1032.
[6] Q. Zhai, V. Martin-Mayor, D. L. Schlagel, G. G. Kenning, and R. L. Orbach, Phys. Rev. B 100, 094202 (2019).

Date:  8th July 2021 - 16.30 Rome Time

Multi-source data and transmission models to fight COVID-19 epidemic in France 

Giulia Pullano, INSERM, Sorbonne Université, Orange S.A.  

Abstract:  On March 17, 2020, French authorities implemented a nationwide lockdown to respond to COVID-19 epidemic emergency. Analyzing multiscale mobility network, reconstructed from mobile phone data, we measured how lockdown altered mobility patterns at both local and country scales. Lockdown caused a 65\% reduction in countrywide number of displacements. Mobility drops were unevenly distributed across regions and they were strongly associated with socio-economic, demographic factors and risk aversion. Major cities largely shrank their pattern of connectivity, reducing it mainly to short-range commuting, despite the persistence of some long-range trips. Our findings indicate that lockdown was very effective in reducing population mobility across scales and help to predicting how and where restrictions will be the most effective. As countries in Europe relaxed lockdown restrictions after the first wave, test–trace–isolate strategies became critical to maintaining the incidence of COVID-19 at low levels. By integrating mobile phone, virological, and surveillance data, we then developed transmission epidemic models, calibrated to French COVID-19 epidemic, to evaluate the performance of the testing system in exit of lockdown. 90,000 symptomatic infections, corresponding to 9 out 10 cases, were not ascertained by the surveillance system from 11 May to 28 June 2020. While the detection rate increased over time, this achievement was likely due to a decreasing epidemic activity. The increase in viral circulation in late summer instead strained the testing system and led to the 2nd wave. Substantially more aggressive, targeted, and efficient testing with easier access is required to act as a tool to control the COVID-19 pandemic. As we are still facing COVID-19 pandemic, and there may be other pandemics, epidemiological and behavioral data should be thus collected and open sources, as they are crucially important to outbreak response.

Date:  10th June 2021 - 16.30 Rome Time

Species coexistence and proteome allocation in competitive microbial communities

Leonardo Pacciani Mori, University of California, San Diego 

Abstract:  Microbial communities are ubiquitous and play crucial roles in many natural processes. Despite their importance for the environment, industry and human health, however, there are still many aspects of microbial communities that we do not fully understand. It is a long-standing problem, for example, the fact that microbial communities are normally much more diverse than what models would allow. Recent experiments, then, have shown that the metabolism of microbial species in a community is intertwined with its structure, suggesting that properties at the intracellular level such as the allocation of cellular proteomic resources must be taken into account when describing microbial communities and species abundances. In this talk I will illustrate the problem of describing biodiversity in purely competitive microbial communities, and how models fail to predict the right number of coexisting species. Then, I will show how we can reconsider one of the most commonly used models to describe population dynamics in competitive ecosystems in light of known experimental results that link the species' growth rate to the allocation of their proteome. This new framework describes microbial communities at an "intermediate" level of complexity, describing the species' population dynamics while also retaining insights on the molecular aspects of growth. The results of the model are also compared to some experimental data.

Date:  13rd May 2021 - 16.30 Rome Time

Evidence of glassy phases in large interacting ecosystems with demographic noise 

Ada Altieri, Laboratoire Matière et Systèmes Complexes (MSC), Université de Paris, France

Abstract:  Many complex systems in Nature, from metabolic networks to ecosystems, appear to be poised at the edge of stability, hence displaying enormous responses to external perturbations. This feature, also known in physics as marginal stability, is often the consequence of the complex underlying interaction network, which can induce large-scale collective dynamics and therefore critical behaviors.  In this seminar, I will present the problem of ecological complexity by focusing on a reference model in theoretical ecology, the high-dimensional Lotka-Volterra model with random symmetric interactions and finite demographic noise [1]. I will show how to obtain a complete characterization of the phase diagrams by means of techniques rooted in mean-field spin-glass theory. Notably, I will relate emerging collective behaviors and slow relaxation dynamics to the appearance of different phases and rough energy landscapes akin to those occurring in glassy systems  [2,3]. I will describe in particular: i) a multiple equilibria phase, which can be proven to be associated with an exponential number of stable equilibria in the system size; ii) a marginally stable amorphous phase (denoted as Gardner phase) as characterized by a hierarchical organization of these equilibria [1].  Finally, I will discuss the wide-ranging applicability of these outcomes to many different contexts, from evolutionary game theory to complex economic systems.

[1] A. Altieri, F. Roy, C. Cammarota, G. Biroli, Properties of equilibria and glassy phases of the random Lotka-Volterra model with demographic noise, arXiv:2009.10565  (2020).

[2] P. Charbonneau, J. Kurchan, G. Parisi, P. Urbani, Fractal free energy landscapes in structural glasses, Nature Communications 5, 3725 (2014).

[3] A. Altieri, Jamming and Glass Transitions: In Mean-Field Theory and Beyond, Springer Nature (2019).

Date:  8th April 2021 - 16.30 Rome Time

Statistical mechanics of interacting polymers explains chromosome folding  

Simona Bianco, Max-Delbrück Center for Molecular Medicine, Berlin 

Abstract:  Chromosomes are folded in complex, non-random three-dimensional conformations within the cell nucleus, as highlighted by novel biochemical and microscopy technologies. Notably, chromosomes architecture and their interaction network are involved in vital cell functions, controlling gene expression, whereas abnormal chromosome folding has been linked to diseases. In this talk, I discuss how massive data on genome architecture, generated thanks to significant experimental advances in the last decade, can be explained in a principled approach based on the statistical mechanics of polymers and some of their underlying molecular mechanisms understood. I also discuss how polymer models can be employed to investigate chromosome structure at the single-molecule level and to predict the effects of pathogenic genomic mutations, as validated by experimental data, opening the way to revolutionary medical applications.

Date:  11th March 2021 - 16.30 Rome Time

Modelling Immune Recognition with Restricted Boltzmann Machines 

Barbara Bravi, Laboratoire de Physique, École Normale Supérieure Paris

Abstract:  The immune response of an organism when it is infected by a pathogen is based on the recognition of small portions of its proteins. This raises two questions: what protein portions are relevant to this process? And what immune cells are able to recognize them? In this talk, I will discuss models to answer those two questions that are based on the machine learning method known as Restricted Boltzmann Machine and that are learned from large protein sequence datasets. These models provide flexible and interpretable frameworks to characterize and predict immune recognition of both cancer and infections.  

Date:  11th February 2021 - 16.30 Rome Time

Spatial patterns in the velocity field of Active Matter systems 

Lorenzo Caprini, University of Camerino

Abstract:  Many systems of biological or technological interest, such as bacterial colonies or cell monolayers, show spatial patterns in their velocity field without displaying a global polarization. In this talk, we investigate this phenomenon through a non-equilibrium stochastic dynamics, the so-called Active Brownian Particles (ABP), which is one of the most popular minimal models to describe the behavior of several experimental active particles. We report the first evidence that pure repulsive spherical ABP, without alignment interactions, spontaneously form large domains of particles with aligned velocities, both in homogeneous dense phases and phase-separated regimes. The size of the velocity domains is measured through the correlation length of the spatial velocity correlations whose shape is analytically predicted. We unveil the non-thermal nature of this collective phenomenon that, instead, is induced by the interplay between steric interactions and active forces, also highlighting the dynamical role played by inertial forces. The results are summarized in a non-equilibrium phase diagram, packing fraction vs persistence time, where the structural properties of the system (distinguishing active liquid, hexatic and solid phases) are superimposed with the velocity correlation lengths. The presence of the almost-translational order typical of hexatic and solid configurations plays a crucial role and reveals an interesting scenario which also involves intermittency phenomena in the time-trajectory of the kinetic energy. 

Date:  14th January 2021 - 16.30 Rome Time