Complex Systems Seminar (Statistical Physics, Biophysics & Condensed Matter)

Tissue folding is a fundamental process in animal organ development, yet the mechanical mechanisms governing fold stability and unfolding remain incompletely understood. In this seminar, I will discuss the evolution of fold morphology and mechanics during Drosophila wing disc development. I will introduce the techniques we developed to study the apical surface of the tissue and present measurements of fold depth and width that characterize the evolution of folds throughout development. The folding process is closely associated with the presence of a fibrous apical extracellular matrix (aECM) that physically connects opposing sides of the folds. By combining a lateral vertex model incorporating aECM-mediated adhesion with genetic perturbations, we show that the aECM contributes to fold stability and that its removal is necessary for proper unfolding. Furthermore, we demonstrate that perturbations of the aECM during larval development leads to morphological defects in the adult wing. Time permitting, I will also introduce a computational library currently under development that aims to extend the study of folded structures interactively for general surfaces/tissues.

Thermodynamic trade-offs between speed, accuracy, and dissipation have recently attracted considerable interest in the field of stochastic thermodynamics. In mesoscopic systems, stochastic currents are often the most accessible observables, and such trade-off relations --- formulated in terms of current fluctuations and entropy production in the systems --- help probe the nonequilibrium nature of the dynamics by providing lower bounds on the rate of entropy production.

These results are conventionally formulated in terms of current fluctuations at fixed times. However, events of interest in noisy mesoscopic systems often occur at random times. In this talk I introduce a formulation of such trade-off relations in terms of observables at random times --- particularly first-passage times for fluctuating currents --- and show that first-passage observables can encode thermodynamic information equivalently to current fluctuations. This formulation is powerfully general and while complementary to the conventional approach based on current fluctuations, helps paints a unifying picture of existing results.

Face-to-face interactions reveal recurring patterns, such as broad contact distributions, suggesting the possibility of shared underlying mechanisms. To unveil these mechanisms, we present an original modeling framework for generating temporal networks, designed to incorporate the critical role of spatial constraints in shaping interaction patterns. To do this, we study walkers evolving in a two-dimensional space following simple pedestrian dynamics and extract the temporal network of interactions between them, to be compared against empirical time-varying networks of contact. Starting with the simplest case, the two-dimensional random walk, and gradually incorporating layers of complexity, I will show to what extent empirical observations can be explained with simple dynamics and where deeper sociological understanding and complex mechanisms are required.

Quantifying the irreversibility and dissipation of non-equilibrium processes is crucial to understanding their behavior, assessing their possible capabilities, and characterizing their efficiency. 

We introduce a physical quantity that quantifies the irreversibility of stochastic Langevin systems from the observation of individual molecules’ displacements. Categorizing these displacements into a few groups based on their initial and final position allows us to measure irreversibility precisely without the need to know the forces and magnitude of the fluctuations acting on the system. Our model-free estimate of irreversibility is related to entropy production by a conditional fluctuation theorem and provides a lower bound to the average entropy production. 

We validate the method on single-molecule force spectroscopy experiments of proteins subject to force ramps. We show that irreversibility is sensitive to detailed features of the energy landscape underlying the protein folding dynamics. This suggests that our method can be employed to unveil key properties of protein folding processes and to fingerprint mechanosensitive proteins.

How species interact determines whether communities thrive or collapse. Yet we lack a principled way to connect measurable interaction properties to diversity and stability. The Lotka-Volterra model, a cornerstone of theoretical ecology, links community-level outcomes to statistical properties of interaction distributions but typically assumes these interactions are Gaussian-distributed. Using consumer-resource models that mechanistically capture microbial metabolism (competition for shared nutrients and cross-feeding of metabolic byproducts), I will show that species interactions automatically organize into a “many weak, few strong” pattern that deviates strongly from Gaussian assumptions. Remarkably, this same skewed pattern appears across empirical communities spanning microbes, soil food webs, and planktonic networks. I will present an analytical explanation rooted in the combinatorial structure of metabolic interactions, and demonstrate that accounting for this pattern improves predictions of both community stability and diversity, in the model and against experimental data. More broadly, these results highlight how metabolic constraints shape the structure of ecological communities and point toward a tighter integration of metabolism and predictive ecology.

When exposed to a contagion phenomenon, individuals may adapt their behavior in response to the perceived risk of infection, reducing both their exposure and their probability of infecting others. While adaptive behaviors have been studied in pairwise networks, many real-world contagion processes and adaptive mechanisms are driven by group (higher-order) interactions, potentially giving rise to new phenomena.

We consider adaptive behaviors triggered by risk awareness perceived through group and pairwise interactions, and we study their impact on contagion processes using numerical simulations and a mean-field analytical approach. We show that adaptivity can profoundly reshape the nature of contagion dynamics, and that its effectiveness in mitigating contagion depends on the type of information triggering adaptive behaviors. Adaptation can suppress bistability and defuse explosive discontinuous transitions typical of higher-order contagion, effectively transforming the process into a pairwise one and making the phase transition continuous. Moreover, adaptive behaviors based on group information outperform pairwise strategies in mitigating contagion while requiring a lower social cost. By exploiting population heterogeneity, they prevent the spread from reaching highly connected nodes and large groups, which would otherwise sustain the dynamics.

These results highlight the crucial role of group interactions and adaptive mechanisms in shaping both the nature and control of contagion processes, paving the way for the design of new intervention measures.