News

Improving epidemic testing and containment strategies using machine learning accepted in Machine Learning: Science and Technology

Comparison of different evolution regimes of disease spreading: free evolution (bottom left half) vs network strategy (top right half). Image by L. Natali.
Improving epidemic testing and containment strategies using machine learning
Laura Natali, Saga Helgadottir, Onofrio M. Maragò, Giovanni Volpe
Machine Learning: Science and Technology (2021)
doi: 10.1088/2632-2153/abf0f7
arXiv: 2011.11717

Containment of epidemic outbreaks entails great societal and economic costs. Cost-effective containment strategies rely on efficiently identifying infected individuals, making the best possible use of the available testing resources. Therefore, quickly identifying the optimal testing strategy is of critical importance. Here, we demonstrate that machine learning can be used to identify which individuals are most beneficial to test, automatically and dynamically adapting the testing strategy to the characteristics of the disease outbreak. Specifically, we simulate an outbreak using the archetypal susceptible-infectious-recovered (SIR) model and we use data about the first confirmed cases to train a neural network that learns to make predictions about the rest of the population. Using these prediction, we manage to contain the outbreak more effectively and more quickly than with standard approaches. Furthermore, we demonstrate how this method can be used also when there is a possibility of reinfection (SIRS model) to efficiently eradicate an endemic disease.

Round Table Discussion on: Living Active Matter

As part of the experimental training, a second round-table discussion took place yesterday, 18 March 2021. The event centered around a discussion on the topic of « Living Active Matter » and featured four invited guests, all physicists, who have studied different topics and length scales relevant to living systems. The invited panel was composed of Aidan Brown from University of Edinburgh, Salima Rafai who works at CNRS in Grenoble, Eric Clément from PMMH-ESPCI in Paris and Benjamin Friedrich from TU Dresden, and was conducted by six of the students attending the training: Audrey Nsamela, David Bronte, Jérémie Bertrand, Ojus Satish Bagal, Daniela Peréz Guerrero and Dana Hassan, who first introduced each guest and then asked selected questions. From molecules and cells to tissues, organisms and populations, each guest had a particular expertise which made for a wide-ranging and interesting discussion.

A question on the evolutionary role of self-propulsion was met with an answer from Dr Brown, who, as obvious as it may seem, pointed out that organisms become “active” when whatever they need to survive is not in their immediate surroundings and must be found elsewhere. When Dr Rafai suggested that the micro-swimmers she studied were not converting their energy to motion optimally, Dr Brown pointed out that biological systems are optimized only in the sense that they are versatile and can adapt to a large number of situations or physical parameters, which is not something that can be captured by a single experiment. This goes to show that, when given the same set of facts, physicists and biologists will often interpret their observations differently, and that discussions between the two disciplines can be fruitful.

Dr Friedrich pointed out that the inherent complexity of biology was such that you could sometimes make progress by just looking and writing down how the processes unfold. He went on to explain that one of the bigger challenges biologists face is that many of these processes occurr below the resolution limit of the microscopes (the “diffraction barrier”) and can therefore not be observed by regular optical microscopes. Several panelists are excited about the coming of newly-designed, ground-breaking microscopes; devices that would use entangled photons to break the diffraction barrier. These new technologies could help not only the field of biology, but also encourage physicists and chemists to collaborate and create models of previously undiscovered mechanisms at the smaller scales. Deep learning is another tool that was alluded to by Dr Rafai as something to look forward to for image reconstruction.

Overall, we found the discussion very productive and we would like to thank once more the panelists for their insights and willingness to participate!

Round Table Discussion on: Phoretic Propulsion Mechanism

During the second day of the experimental training, we organised the first round table discussion. The session was chaired by six of the students attending the training: Carolina van Baalen, Danne van Roon, Gülce Bayram, Harshith Bachimanchi, Laura Natali and Sandrine Heijnen.

The topic of the round table was phoretic propulsion mechanisms and we had four panelists – Juliane Simmchen, Frank Cichos, Ivo Buttinoni and Felix Ginot – and a guest speaker, Antoni Homs Corbera. After a brief introduction of the panelists, we had a chance to ask all the questions we collected from the other participants.

The discussion started with the definition of the term “phoresis” and continued with the simulation frameworks for phoretic colloids. It included a brief discussion of the complexity involved in these processes and the typical length scales at which interfacial effects are relevant.

The conclusion was “a common joke at conferences is that the phoresis starts when coffee is about to be served”. The real conclusion was that phoretic interaction needs very large gradients on the macroscopic scale and is hidden by diffusion on a very small scale.

All participants had the possibility to jump in and add upcoming questions. We ended the round table by discussing the possible applications of phoretic colloids, highlighting the environmental aspects like microplastics’ filtration in water.

We thank all the guests and participants for making it a successful discussion moment.

Talk by Chun-Jen Chen at Institute of Physics, Academia Sinica (Taiwan), 25 February 2021

The topic of this talk covered the laser experiments enabling active particle steering (upper left), collective motion of such particles (middle) and connection to social animal behaviours, eg. fish school (lower). (Image by C-J Chen.)
On the 25th of February 2021, Chun-Jen gave a talk at the Institute of Physics, Academia Sinica (Taipei, Taiwan) about his research project at University of Konstanz. He explained how active Janus micro-spheres can be propelled and steered at the indivitual level in his experiment system and how such experiments are linked to studies of collective behaviours in living systems. The talk induced vivid discussions with audience of different backgrounds. Chun-Jen also shared experiences regarding PhD life in Germany with prospective young researchers in Taiwan.

Active noise-driven particles under space-dependent friction in one dimension on arXiv

Sketch of the confining potential U(x) = κ|x|, a linear friction gradient γ(x) = γ0+γ1|x| in arbitrary units. The particle, shown by a blue dot on the x-axis, is activated by noise (indicated in red), under the influence of the potential and the friction gradient. Image by D. Breoni.
Active noise-driven particles under space-dependent friction in one dimension

Davide Breoni, Ralf Blossey, Hartmut Löwen
arxiv: 2102.09944

Abstract: We study a Langevin equation describing the stochastic motion of a particle in one dimension with coordinate x, which is simultaneously exposed to a space-dependent friction coefficient γ (x), a confining potential U(x) and non-equilibrium (i.e., active) noise. Specically, we consider frictions γ (x) = γ0 + γ1|x|p and potentials U(x) ∝ |x|p with exponents p = 1; 2 and n = 0; 1; 2. We provide analytical and numerical results for the particle dynamics for short times and the stationary
probability density functions (PDFs) for long times. The short-time behaviour displays diffusive and ballistic regimes while the stationary PDFs display unique characteristic features depending on the exponent values (p; n). The PDFs interpolate between Laplacian, Gaussian and bimodal distributions, whereby a change between these different behaviours can be achieved by a tuning of the friction strengths ratio
γ0 / γ1. Our model is relevant for molecular motors moving on a
one-dimensional track and can also be realized for confined self-propelled colloidal particles.

Improving epidemic testing and containment strategies using machine learning on ArXiv

Comparison of different evolution regimes of disease spreading: free evolution (bottom left half) vs network strategy (top right half). Image by L. Natali.
Improving epidemic testing and containment strategies using machine learning
Laura Natali, Saga Helgadottir, Onofrio M. Maragò, Giovanni Volpe
arXiv: 2011.11717

Containment of epidemic outbreaks entails great societal and economic costs. Cost-effective containment strategies rely on efficiently identifying infected individuals, making the best possible use of the available testing resources. Therefore, quickly identifying the optimal testing strategy is of critical importance. Here, we demonstrate that machine learning can be used to identify which individuals are most beneficial to test, automatically and dynamically adapting the testing strategy to the characteristics of the disease outbreak. Specifically, we simulate an outbreak using the archetypal susceptible-infectious-recovered (SIR) model and we use data about the first confirmed cases to train a neural network that learns to make predictions about the rest of the population. Using these prediction, we manage to contain the outbreak more effectively and more quickly than with standard approaches. Furthermore, we demonstrate how this method can be used also when there is a possibility of reinfection (SIRS model) to efficiently eradicate an endemic disease.

David Bronte Ciriza visits the Soft Matter Lab, Gothenburg, Sweden

From the 2nd to the 9th of November 2020, David Bronte Ciriza and Alessandro Magazzù (Post Doc at the CNR-IPCF Messina) have visited the Soft Matter Lab at the University of Gothenburg, Sweden. During their visit, they have been working on a project related to the use of neural networks to improve optical forces calculations. This visit has also served as networking opportunity with the researchers at the Gothenburg University, including the ones that participate in the ActiveMatter network.

Alessandro Magazzù and David Bronte Ciriza in Gothenburg. Foto by Aukut Argun

Active Brownian and inertial particles in disordered environments: short-time expansion of the mean-square displacement on ArXiv

Active Brownian and inertial particles in disordered environments: short-time expansion of the mean-square displacement
Davide Breoni, Michael Schmiedeberg, Hartmut Löwen
arXiv: 2010.11076

We consider an active Brownian particle moving in a disordered two-dimensional energy or motility landscape. The averaged mean-square-displacement (MSD) of the particle is calculated analytically within a systematic short-time expansion. As a result, for overdamped particles, both an external random force field and disorder in the self-propulsion speed induce ballistic behaviour adding to the ballistic regime of an active particle with sharp self-propulsion speed. Spatial correlations in the force and motility landscape contribute only to the cubic and higher order powers in time for the MSD. Finally, for inertial particles two superballistic regimes are found where the scaling exponent of the MSD with time is α = 3 and α = 4. We confirm our theoretical predictions by computer simulations. Moreover they are verifiable in experiments on self-propelled colloids in random environments.

On the Morphology of Active Deformable 3D Droplets

It is increasingly becoming apparent that the physical concepts of forces and flows play an important role in understanding biological processes, from the spread of cancers to morphogenesis, thedevelopment of organisms. However, biological systems, such as cells, probe new ideas in that theyoperate out of thermodynamic equilibrium continually taking chemical energy from their surroundings, and using it to move and self-organise.

The term active matter has come to describe models of living systems where such a continuous influx of energy leads to striking collective behaviour like the chaotic flow patterns of active turbulence seen in collections of bacteria and self-propelled topological defects which are now thought to be relevant to some modes of biofilm formation. This paper is a numerical investigation of three-dimensional droplets composed of active matter and the ways in which their shapes change in response to the continuous input of energy. One striking observation is the continuous formation of finger-like protrusions, reminiscent of the collective motion of invading cancer cells. By changing the mechanical properties of the drop or the activity level, we find several different dynamical responses: for example the droplet surface can wrinkle in a way that resembles a walnut or the active forces can drive a dimple in the droplet to grow, leading to a cup-shape: such invagination is reminiscent of patterns seen during morphogenesis.

Understanding the behaviour of model systems, here a continuum model of active material, is an important step towards the goal of understanding the role of physical theories in the life sciences.

Links: https://arxiv.org/abs/2010.10427

Morphology of active deformable 3D droplets

Morphology of active deformable 3D droplets.

Morphology of active deformable 3D droplets on ArXiv

Morphology of active deformable 3D droplets.

Morphology of active deformable 3D droplets
Liam J. Ruske, Julia M. Yeomans
arXiv: 2010.10427

We numerically investigate the morphology and disclination line dynamics of active nematic droplets in three dimensions. Although our model only incorporates the simplest possible form of achiral active stress, active nematic droplets display an unprecedented range of complex morphologies. For extensile activity finger-like protrusions grow at points where disclination lines intersect the droplet surface. For contractile activity, however, the activity field drives cup-shaped droplet invagination, run-and-tumble motion or the formation of surface wrinkles. This diversity of behaviour is explained in terms of an interplay between active anchoring, active flows and the dynamics of the motile dislocation lines. We discuss our findings in the light of biological processes such as morphogenesis, collective cancer invasion and the shape control of biomembranes, suggesting that some biological systems may share the same underlying mechanisms as active nematic droplets.