arXiv subjects

Carlo Manzo

Publications and source records attributed to Carlo Manzo.

At least 19 recordsLinked to original sources

arXiv2025

Global graph features unveiled by unsupervised geometric deep learning

Graphs provide a powerful framework for modeling complex systems, but their structural variability poses significant challenges for analysis and classification. To address these challenges, we introduce GAUDI (Graph Autoencoder Uncovering Descriptive Information), a novel unsupervised geometric deep learning framework designed to capture both local details and global structure. GAUDI employs an innovative hourglass architecture with hierarchical pooling and upsampling layers linked through skip connections, which preserve essential connectivity information throughout the encoding-decoding process. Even though identical or highly similar underlying parameters describing a system's state can lead to significant variability in graph realizations, GAUDI consistently maps them into nearby regions of a structured and continuous latent space, effectively disentangling invariant process-level features from stochastic noise. We demonstrate GAUDI's versatility across multiple applications, including small-world networks modeling, characterization of protein assemblies from super-resolution microscopy, analysis of collective motion in the Vicsek model, and identification of age-related changes in brain connectivity. Comparison with related approaches highlights GAUDI's superior performance in analyzing complex graphs, providing new insights into emergent phenomena across diverse scientific domains.

cs.LG↗

arXiv2024

Enhanced Spatial Clustering of Single-Molecule Localizations with Graph Neural Networks

Single-molecule localization microscopy generates point clouds corresponding to fluorophore localizations. Spatial cluster identification and analysis of these point clouds are crucial for extracting insights about molecular organization. However, this task becomes challenging in the presence of localization noise, high point density, or complex biological structures. Here, we introduce MIRO (Multifunctional Integration through Relational Optimization), an algorithm that uses recurrent graph neural networks to transform the point clouds in order to improve clustering efficiency when applying conventional clustering techniques. We show that MIRO supports simultaneous processing of clusters of different shapes and at multiple scales, demonstrating improved performance across varied datasets. Our comprehensive evaluation demonstrates MIRO's transformative potential for single-molecule localization applications, showcasing its capability to revolutionize cluster analysis and provide accurate, reliable details of molecular architecture. In addition, MIRO's robust clustering capabilities hold promise for applications in various fields such as neuroscience, for the analysis of neural connectivity patterns, and environmental science, for studying spatial distributions of ecological data.

cs.LG↗

arXiv2023

Quantitative evaluation of methods to analyze motion changes in single-particle experiments

The analysis of live-cell single-molecule imaging experiments can reveal valuable information about the heterogeneity of transport processes and interactions between cell components. These characteristics are seen as motion changes in the particle trajectories. Despite the existence of multiple approaches to carry out this type of analysis, no objective assessment of these methods has been performed so far. Here, we report the results of a competition to characterize and rank the performance of these methods when analyzing the dynamic behavior of single molecules. To run this competition, we implemented a software library that simulates realistic data corresponding to widespread diffusion and interaction models, both in the form of trajectories and videos obtained in typical experimental conditions. The competition constitutes the first assessment of these methods, providing insights into the current limitations of the field, fostering the development of new approaches, and guiding researchers to identify optimal tools for analyzing their experiments.

cond-mat.soft↗

arXiv2023

Learning minimal representations of stochastic processes with variational autoencoders

Stochastic processes have found numerous applications in science, as they are broadly used to model a variety of natural phenomena. Due to their intrinsic randomness and uncertainty, they are, however, difficult to characterize. Here, we introduce an unsupervised machine learning approach to determine the minimal set of parameters required to effectively describe the dynamics of a stochastic process. Our method builds upon an extended $\beta$-variational autoencoder architecture. By means of simulated datasets corresponding to paradigmatic diffusion models, we showcase its effectiveness in extracting the minimal relevant parameters that accurately describe these dynamics. Furthermore, the method enables the generation of new trajectories that faithfully replicate the expected stochastic behavior. Overall, our approach enables the autonomous discovery of unknown parameters describing stochastic processes, hence enhancing our comprehension of complex phenomena across various fields.

cond-mat.soft↗

arXiv2023

High-density single-molecule maps reveal transient membrane receptor interactions within a dynamically varying environment

Over recent years, super-resolution and single-molecule imaging methods have delivered unprecedented details on the nanoscale organization and dynamics of individual molecules in different contexts. Yet, visualizing single-molecule processes in living cells with the required spatial and temporal resolution remains highly challenging. Here, we report on an analytical approach that extracts such information from live-cell single-molecule imaging at high-labeling densities using standard fluorescence probes. Our high-density-mapping (HiDenMap) methodology provides single-molecule nanometric localization accuracy together with millisecond temporal resolution over extended observation times, delivering multi-scale spatiotemporal data that report on the interaction of individual molecules with their dynamic environment. We validated HiDenMaps by simulations of Brownian trajectories in the presence of patterns that restrict free diffusion with different probabilities. We further generated and analyzed HiDenMaps from single-molecule images of transmembrane proteins having different interaction strengths to cortical actin, including the transmembrane receptor CD44. HiDenMaps uncovered a highly heterogenous and multi-scale spatiotemporal organization for all the proteins that interact with the actin cytoskeleton. Notably, CD44 alternated between periods of random diffusion and transient trapping, likely resulting from actin-dependent CD44 nanoclustering. Whereas receptor trapping was dynamic and lasted for hundreds of milliseconds, actin remodeling occurred at the timescale of tens of seconds, coordinating the assembly and disassembly of CD44 nanoclusters rich regions. Together, our data demonstrate the power of HiDenMaps to explore how individual molecules interact with and are organized by their environment in a dynamic fashion.

physics.bio-ph↗

arXiv2023

Roadmap on Deep Learning for Microscopy

Through digital imaging, microscopy has evolved from primarily being a means for visual observation of life at the micro- and nano-scale, to a quantitative tool with ever-increasing resolution and throughput. Artificial intelligence, deep neural networks, and machine learning are all niche terms describing computational methods that have gained a pivotal role in microscopy-based research over the past decade. This Roadmap is written collectively by prominent researchers and encompasses selected aspects of how machine learning is applied to microscopy image data, with the aim of gaining scientific knowledge by improved image quality, automated detection, segmentation, classification and tracking of objects, and efficient merging of information from multiple imaging modalities. We aim to give the reader an overview of the key developments and an understanding of possibilities and limitations of machine learning for microscopy. It will be of interest to a wide cross-disciplinary audience in the physical sciences and life sciences.

physics.optics↗

arXiv2023

Inferring pointwise diffusion properties of single trajectories with deep learning

In order to characterize the mechanisms governing the diffusion of particles in biological scenarios, it is essential to accurately determine their diffusive properties. To do so, we propose a machine learning method to characterize diffusion processes with time-dependent properties at the experimental time resolution. Our approach operates at the single-trajectory level predicting the properties of interest, such as the diffusion coefficient or the anomalous diffusion exponent, at every time step of the trajectory. In this way, changes in the diffusive properties occurring along the trajectory emerge naturally in the prediction, and thus allow the characterization without any prior knowledge or assumption about the system. We first benchmark the method on synthetic trajectories simulated under several conditions. We show that our approach can successfully characterize both abrupt and continuous changes in the diffusion coefficient or the anomalous diffusion exponent. Finally, we leverage the method to analyze experiments of single-molecule diffusion of two membrane proteins in living cells: the pathogen-recognition receptor DC-SIGN and the integrin $\alpha5\beta1$. The analysis allows us to characterize physical parameters and diffusive states with unprecedented accuracy, shedding new light on the underlying mechanisms.

cond-mat.soft↗

arXiv2023

Broadband plasmonic nanoantennas for multi-color nanoscale dynamics in living cells

Recently, the implementation of plasmonic nanoantennas has opened new possibilities to investigate the nanoscale dynamics of individual biomolecules in living cell. However, studies have yet been restricted to single molecular species as the narrow wavelength resonance of gold-based nanostructures precludes the simultaneous interrogation of different fluorescently labeled molecules. Here we exploited broadband aluminum-based nanoantennas carved at the apex of near-field probes to resolve nanoscale-dynamic molecular interactions on intact living cell membranes. Through multicolor excitation, we simultaneously recorded fluorescence fluctuations of dual-color labeled transmembrane receptors known to form nanoclusters in living cells. Fluorescence cross-correlation studies revealed transient interactions between individual receptors in regions of ~60 nm. Moreover, the high signal-to-background ratio provided by the antenna illumination allowed us to directly detect fluorescent bursts arising from the passage of individual receptors underneath the antenna. Remarkably, by reducing the illumination volume below the characteristic receptor nanocluster sizes, we resolved molecular diffusion within nanoclusters and distinguished it from nanocluster diffusion. Spatiotemporal characterization of transient interactions between molecules is crucial to understand how they communicate with each other to regulate cell function. Our work demonstrates the potential of broadband photonic antennas to study multi-molecular events and interactions in living cell membranes with unprecedented spatiotemporal resolution.

physics.optics↗

arXiv2023

Preface: Characterisation of Physical Processes from Anomalous Diffusion Data

Preface to the special issue "Characterisation of Physical Processes from Anomalous Diffusion Data" associated with the Anomalous Diffusion Challenge ( https://andi-challenge.org ) and published in Journal of Physics A: Mathematical and Theoretical. The list of articles included in the special issue can be accessed at https://iopscience.iop.org/journal/1751-8121/page/Characterisation-of-Physical-Processes-from-Anomalous-Diffusion-Data .

cond-mat.stat-mech↗

arXiv2022

Geometric deep learning reveals the spatiotemporal fingerprint of microscopic motion

The characterization of dynamical processes in living systems provides important clues for their mechanistic interpretation and link to biological functions. Thanks to recent advances in microscopy techniques, it is now possible to routinely record the motion of cells, organelles, and individual molecules at multiple spatiotemporal scales in physiological conditions. However, the automated analysis of dynamics occurring in crowded and complex environments still lags behind the acquisition of microscopic image sequences. Here, we present a framework based on geometric deep learning that achieves the accurate estimation of dynamical properties in various biologically-relevant scenarios. This deep-learning approach relies on a graph neural network enhanced by attention-based components. By processing object features with geometric priors, the network is capable of performing multiple tasks, from linking coordinates into trajectories to inferring local and global dynamic properties. We demonstrate the flexibility and reliability of this approach by applying it to real and simulated data corresponding to a broad range of biological experiments.

physics.data-an↗

arXiv2021

Objective comparison of methods to decode anomalous diffusion

Deviations from Brownian motion leading to anomalous diffusion are found in transport dynamics from quantum physics to life sciences. The characterization of anomalous diffusion from the measurement of an individual trajectory is a challenging task, which traditionally relies on calculating the trajectory mean squared displacement. However, this approach breaks down for cases of practical interest, e.g., short or noisy trajectories, heterogeneous behaviour, or non-ergodic processes. Recently, several new approaches have been proposed, mostly building on the ongoing machine-learning revolution. To perform an objective comparison of methods, we gathered the community and organized an open competition, the Anomalous Diffusion challenge (AnDi). Participating teams applied their algorithms to a commonly-defined dataset including diverse conditions. Although no single method performed best across all scenarios, machine-learning-based approaches achieved superior performance for all tasks. The discussion of the challenge results provides practical advice for users and a benchmark for developers.

physics.data-an↗

arXiv2021

Extreme Learning Machine for the Characterization of Anomalous Diffusion from Single Trajectories (AnDi-ELM)

The study of the dynamics of natural and artificial systems has provided several examples of deviations from Brownian behavior, generally defined as anomalous diffusion. The investigation of these dynamics can provide a better understanding of diffusing objects and their surrounding media, but a quantitative characterization from individual trajectories is often challenging. Efforts devoted to improving anomalous diffusion detection using classical statistics and machine learning have produced several new methods. Recently, the anomalous diffusion challenge (AnDi, www.andi-challenge.org) was launched to objectively assess these approaches on a common dataset, focusing on three aspects of anomalous diffusion: the inference of the anomalous diffusion exponent; the classification of the diffusion model; and the segmentation of trajectories. In this article, I describe a simple approach to tackle the tasks of the AnDi challenge by combining extreme learning machine and feature engineering (AnDi-ELM). The method reaches satisfactory performance while offering a straightforward implementation and fast training time with limited computing resources, making it a suitable tool for fast preliminary screening of anomalous diffusion.

physics.data-an↗

arXiv2020

AnDi: The Anomalous Diffusion Challenge

The deviation from pure Brownian motion generally referred to as anomalous diffusion has received large attention in the scientific literature to describe many physical scenarios. Several methods, based on classical statistics and machine learning approaches, have been developed to characterize anomalous diffusion from experimental data, which are usually acquired as particle trajectories. With the aim to assess and compare the available methods to characterize anomalous diffusion, we have organized the Anomalous Diffusion (AnDi) Challenge (\url{http://www.andi-challenge.org/}). Specifically, the AnDi Challenge will address three different aspects of anomalous diffusion characterization, namely: (i) Inference of the anomalous diffusion exponent. (ii) Identification of the underlying diffusion model. (iii) Segmentation of trajectories. Each problem includes sub-tasks for different number of dimensions (1D, 2D and 3D). In order to compare the various methods, we have developed a dedicated open-source framework for the simulation of the anomalous diffusion trajectories that are used for the training and test datasets. The challenge was launched on March 1, 2020, and consists of three phases. Currently, the participation to the first phase is open. Submissions will be automatically evaluated and the performance of the top-scoring methods will be thoroughly analyzed and compared in an upcoming article.

cond-mat.stat-mech↗

arXiv2020

Machine learning method for single trajectory characterization

In order to study transport in complex environments, it is extremely important to determine the physical mechanism underlying diffusion, and precisely characterize its nature and parameters. Often, this task is strongly impacted by data consisting of trajectories with short length and limited localization precision. In this paper, we propose a machine learning method based on a random forest architecture, which is able to associate even very short trajectories to the underlying diffusion mechanism with a high accuracy. In addition, the method is able to classify the motion according to normal or anomalous diffusion, and determine its anomalous exponent with a small error. The method provides highly accurate outputs even when working with very short trajectories and in the presence of experimental noise. We further demonstrate the application of transfer learning to experimental and simulated data not included in the training/testing dataset. This allows for a full, high-accuracy characterization of experimental trajectories without the need of any prior information.

cond-mat.stat-mech↗

arXiv2019

Bayesian analysis of data from segmented super-resolution images for quantifying protein clustering

Super-resolution imaging techniques have largely improved our capabilities to visualize nanometric structures in biological systems. Their application further enables one to potentially quantitate relevant parameters to determine the molecular organization and stoichiometry in cells. However, the inherently stochastic nature of the fluorescence emission and labeling strategies imposes the use of dedicated methods to accurately measure these parameters. Here, we describe a Bayesian approach to precisely quantitate the relative abundance of molecular oligomers from segmented images. The distribution of proxies for the number of molecules in a cluster -- such as the number of localizations or the fluorescence intensity -- is fitted via a nested sampling algorithm to compare mixture models of increasing complexity and determine the optimal number of mixture components and their weights. We test the performance of the algorithm on {\it in silico} data as a function of the number of data points, threshold, and distribution shape. We compare these results to those obtained with other statistical methods, showing the improved performance of our approach. Our method provides a robust tool for model selection in fitting data extracted from fluorescence imaging, thus improving the precision of parameter determination. Importantly, the largest benefit of this method occurs for small-statistics or incomplete datasets, enabling accurate analysis at the single image level. We further present the results of its application to experimental data obtained from the super-resolution imaging of dynein in HeLa cells, confirming the presence of a mixed population of cytoplasmatic single motors and higher-order structures.

physics.bio-ph↗

arXiv2018

Diffusion through a network of compartments separated by partially-transmitting boundaries

We study the random walk of a particle in a compartmentalized environment, as realized in biological samples or solid state compounds. Each compartment is characterized by its length $L$ and the boundaries transmittance $T$. We identify two relevant spatio-temporal scales that provide alternative descriptions of the dynamics: i) the microscale, in which the particle position is monitored at constant time intervals; and ii) the mesoscale, in which it is monitored only when the particle crosses a boundary between compartments. Both descriptions provide --by construction-- the same long time behavior. The analytical description obtained at the proposed mesoscale allows for a complete characterization of the complex movement at the microscale, thus representing a fruitful approach for this kind of systems. We show that the presence of disorder in the transmittances is a necessary condition to induce anomalous diffusion, whereas the spatial heterogeneity reduces the degree of subdiffusion and, in some cases, can even compensate for the disorder induced by the stochastic transmittance.

cond-mat.stat-mech↗

arXiv2017

Planar Optical Nanoantennas Resolve Cholesterol-Dependent Nanoscale Heterogeneities in the Plasma Membrane of Living Cells

Optical nanoantennas can efficiently confine light into nanoscopic hotspots, enabling single-molecule detection sensitivity at biological relevant conditions. This innovative approach to breach the diffraction limit offers a versatile platform to investigate the dynamics of individual biomolecules in living cell membranes and their partitioning into cholesterol-dependent lipid nanodomains. Here, we present optical nanoantenna arrays with accessible surface hotspots to study the characteristic diffusion dynamics of phosphoethanolamine (PE) and sphingomyelin (SM) in the plasma membrane of living cells at the nanoscale. Fluorescence burst analysis and fluorescence correlation spectroscopy performed on nanoantennas of different gap sizes show that, unlike PE, SM is transiently trapped in cholesterol-enriched nanodomains of 10 nm diameter with short characteristic times around 100 μs. The removal of cholesterol led to the free diffusion of SM, consistent with the dispersion of nanodomains. Our results are consistent with the existence of highly transient and fluctuating nanoscale assemblies enriched by cholesterol and sphingolipids in living cell membranes, also known as lipid rafts. Quantitative data on sphingolipids partitioning into lipid rafts is crucial to understand the spatiotemporal heterogeneous organization of transient molecular complexes on the membrane of living cells at the nanoscale. The proposed technique is fully biocompatible and thus provides various opportunities for biophysics and live cell research to reveal details that remain hidden in confocal diffraction-limited measurements.

physics.bio-ph↗

arXiv2016

Fractional kinetics emerging from ergodicity breaking in random media

We present a modelling approach for diffusion in a complex medium characterized by a random length scale. The resulting stochastic process shows subdiffusion with a behavior in qualitative agreement with single particle tracking experiments in living cells, such as ergodicity breaking, p-variation and aging. In particular, this approach recapitulates characteristic features previously described in part by the fractional Brownian motion and in part by the continuous-time random walk. Moreover, for a proper distribution of the length scale, a single parameter controls the ergodic-to-nonergodic transition and, remarkably, also drives the transition of the diffusion equation of the process from non-fractional to fractional, thus demonstrating that fractional kinetics emerges from ergodicity breaking.

cond-mat.stat-mech↗