Invited talk by G. Volpe at BRC Day “Biomaterials meets AI”, Gothenburg, Sweden, 12 November 2019

Machine learning as a tool for the natural sciences: Opportunities and challenges
Giovanni Volpe
Invited Talk at BRC Day “Biomaterials meets AI”, University of Gothenburg, Gothenburg, Sweden, 12 November 2019

Abstract: Data-driven machine-learning methods are more and more widely used in the natural sciences. Machine learning offers unprecedented opportunities, but it also poses unexpected practical and fundamental challenges. Most importantly, machine-learning methods often work as black boxes, and therefore it can be difficult to understand and interpret their results. Here, we present an overview of the current state of the art of the adoption of machine learning in active-matter research. Finally, we discuss the opportunities and challenges that are emerging, highlighting how active matter and machine learning can work together for mutual benefit.

Bio: Giovanni Volpe is Associate Professor at the University of Gothenburg, where he leads the Soft Matter Lab (http://www.softmatterlab.org/).
He has published more than 80 articles on diverse topics including optical trapping, active matter, neurosciences, and machine learning.
He has co-authored the book “Optical Tweezers: Principles and Applications” (Cambridge University Press, 2015).
He is the recipient the ERC Starting Grant ComplexSwimmers, coordinator of the MSCA Innovative Training Networks ActiveMatter, and the KAW research grant “Active Matter Goes Smart”.
He is one of the chairs of the Conference Emerging Topics in Artificial Intelligence at the SPIE Optics & Photonics Meeting held annually in San Diego (CA).

Active Colloids with Position-Dependent Rotational Diffusivity on ArXiv

Active Colloids with Position-Dependent Rotational Diffusivity

Active Colloids with Position-Dependent Rotational Diffusivity
Miguel Angel Fernandez-Rodriguez, Fabio Grillo, Laura Alvarez, Marco Rathlef, Ivo Buttinoni, Giovanni Volpe & Lucio Isa
arXiv: 1911.02291

The non-thermal nature of self-propelling colloids offers new insights into non-equilibrium physics. The central mathematical model to describe their trajectories is active Brownian motion, where a particle moves with a constant speed, while randomly changing direction due to rotational diffu- sion. While several feedback strategies exist to achieve position-dependent velocity, the possibility of spatial and temporal control over rotational diffusion, which is inherently dictated by thermal fluctuations, remains untapped. Here, we decouple rotational diffusion from thermal noise. Using external magnetic fields and discrete-time feedback loops, we tune the rotational diffusivity of active colloids above and below its thermal value at will and explore a rich range of phenomena including anomalous diffusion, directed transport, and localization. These findings add a new dimension to the control of active matter, with implications for a broad range of disciplines, from optimal transport to smart materials.

Introductory Talk at CECAM Workshop “Active Matter and Artificial Intelligence” by G. Volpe, Lausanne, Switzerland, 1 October 2019

Machine learning for active matter
Giovanni Volpe
Introductory Talk at CECAM Workshop “Active Matter and Artificial Intelligence”
CECAM-HQ-EPFL, Lausanne, Switzerland
30 September – 2 October, 2019

Data-driven machine-learning methods are more and more widely used in the natural sciences. Active-matter research is no exception and has recently started experiment- ing machine-learning approaches. Machine learning offers unprecedented opportunities, but it also poses unexpected practical and fundamental challenges. Most importantly, machine-learning methods often work as black boxes, and therefore it can be difficult to understand and interpret their results. Here, we present an overview of the current state of the art of the adoption of machine learning in active-matter research. Finally, we discuss the opportunities and challenges that are emerging, highlighting how active matter and machine learning can work together for mutual benefit.

Invited talk by G. Volpe at RIAO/Optilas 2019, Cancun, Mexico, 23 Sep 2019

Deep Learning Applications in Digital Video Microscopy and Optical Micromanipulation
Saga Helgadottir, Aykut Argun, Giovanni Volpe
Invited talk at RIAO/Optilas 2019, Cancun, Mexico, 23-27 September 2019

Since its introduction in the mid 90s, digital video microscopy has become a staple for the analysis of data in optical trapping and optical manipulation experiments [1]. Current methods are able to predict the location of the center of a particle in ideal condition with high accuracy. However, these methods fail as the signal-to-noise ratio (SNR) of the images decreases or if there are non-uniform distortions present in the images. Both these conditions are commonly encountered in experiments. In addition, all these methods require considerable user input in terms of analysis parameters, which introduces user bias. In order to automatize the tracking process algorithms using deep learning have been successfully introduced but have not proved to be usable for practical applications.

Here, we provide a fully automated deep learning tracking algorithm with sub-pixel precision in localizing single particle and multiple particles’ positions from image data [2]. We have developed a convolutional neural network that is pre-trained on simulated single particle images in varying conditions of, for example, particle intensity, image contrast and SNR.

We test the pre-trained network on an optically trapped particle both in ideal condition and challenged condition with low SNR and non-uniform distortions [3]. This pre-trained network accurately predicts the location the trapped particle and a comparison of detected trajectories, the distribution of the particle position and the power spectral density of the particle trajectory clearly shows that our algorithm outperforms tracking by radial symmetry [4]. Our algorithm is also able to track non-ideal images with multiple Brownian particles as well as swimming bacteria that are problematic for traditional methods.

In conclusion, our algorithm outperforms current methods in precision and speed of tracking non-ideal images, while eliminating the need for user supervision and therefore the introduction of user biases. 

References

[1] John C Crocker, David G Grier, Journal of Colloid and Interface Science 179, 298–310 (1996).

[2] Saga Helgadottir, Aykut Argun, Giovanni Volpe, Optica 6, 506–513 (2019).

[3] Philip H Jones, Onofrio M Maragò, Giovanni Volpe, Optical tweezers: Principles and applications. Cambridge University Press, 2015.

[4] Raghuveer Parthasarathy. Nature Methods 9724 (2012).

Seminar by G. Volpe at the Department of Chemistry, University of Gothenburg, 19 Sep 2019

Soft Matter Meets Deep Learning
Giovanni Volpe
Department of Chemistry, University of Gothenburg, Sweden
19 September 2019

After a brief overview of artificial intelligence, machine learning and deep learning, I will present a series of recent works in which we have employed deep learning for applications in photonics and active matter. In particular, I will explain how we employed deep learning to enhance digital video microscopy [1], to estimate the properties of anomalous diffusion [2], and to improve the calculation of optical forces. Finally, I will provide an outlook for the application of deep learning in photonics and active matter.

References

[1] S. Helgadottir, A. Argun and G. Volpe, Digital video microscopy enhanced by deep learning. Optica 6(4), 506—513 (2019)
doi: 10.1364/OPTICA.6.000506

[2] S. Bo, F Schmidt, R Eichborn and G. Volpe, Measurement of Anomalous Diffusion Using Recurrent Neural Networks. arXiv: 1905.02038

Intercellular Communication Induces Glycolytic Synchronisation Waves on ArXiv

Intercellular communication induces glycolytic synchronization waves between individually oscillating cells

Intercellular communication induces glycolytic synchronization waves between individually oscillating cells
Martin Mojica-Benavides, David D. van Niekerk, Mite Mijalkov, Jacky L. Snoep, Bernhard Mehlig, Giovanni Volpe, Caroline B. Adiels, and Mattias Goksör
arXiv: 1909.05187

Metabolic oscillations in single cells underlie the mechanisms behind cell synchronization and cell-cell communication. For example, glycolytic oscillations mediated by biochemical communication between cells may synchronize the pulsatile insulin secretion by pancreatic tissue, and a link between glycolytic synchronization anomalies and type-2 diabetes has been hypotesized. Cultures of yeast cells have provided an ideal model system to study synchronization and propagation waves of glycolytic oscillations in large populations. However, the mechanism by which synchronization occurs at individual cell-cell level and overcome local chemical concentrations and heterogenic biological clocks, is still an open question because of experimental limitations in sensitive and specific handling of single cells. Here, we show how the coupling of intercellular diffusion with the phase regulation of individual oscillating cells induce glycolytic synchronization waves. We directly measure the single-cell metabolic responses from yeast cells in a microfluidic environment and characterize a discretized cell-cell communication using graph theory. We corroborate our findings with simulations based on a kinetic detailed model for individual yeast cells. These findings can provide insight into the roles cellular synchronization play in biomedical applications, such as insulin secretion regulation at the cellular level.

CECAM Workshop “Active Matter and Artificial Intelligence Location”, Lausanne, Switzerland 30 September – 2 October 2019

Active Matter and Artificial Intelligence
Location : CECAM-HQ-EPFL, Lausanne, Switzerland
September 30, 2019 – October 2, 2019

Organizers:
Frank Cichos (Universität Leipzig, Germany)
Klaus Kroy (Universität Leipzig, Germany)
Fernando Peruani (Université Nice Sophia Antipolis, France)
Giovanni Volpe (University of Gothenburg, Sweden)

Website:
https://www.cecam.org/workshp1750/

Description:
Biological active matter is composed of self-propelling agents, such as molecular motors, cells, bacteria and animals [1,2], which can perform tasks and feature emergent collective behaviors thanks to their capability of sensing their environment, processing this information and exploiting it through feedback cycles [3]. These processes are intrinsically noisy [4] both at the microscale (e.g. thermal noise [5]) and at the macroscale (e.g. turbulence [6]). Therefore, through millions of years, biological systems have evolved powerful strategies to accomplish specific tasks and thrive in their environment – strategies that are encoded in their shape, biophysical properties, and signal processing networks [13].

Artificial active matter is now being explored as a powerful means to address the big challenges that our society is facing [7]: from new strategies for targeted drug delivery, to the decontamination of polluted soils, to the extraction of energy from naturally occurring out-of-equilibrium conditions. In this context, biological active matter provides an ideal source of tested ideas and approaches [8,9], which we are now trying to exploit to develop artificial systems [10,11].

However, in biological systems, there is only a limited possibility to reduce complexity and introduce controllable perturbations. Therefore, the development of computational models and of proof-of-principle experiments provides an ideal test bench to explore the origin of complexity in biological systems and to harness it for the development of new applications. For example, tuning of sensorial delays yield different behaviors in gradient fields relevant for cellular systems [12], and, inspired by neuronal networks, relevant past experience is harnessed to predict the evolution of complex systems.

In this process, we have been led to the application of machine learning to active matter. Machine learning is an abstraction of the adaption processes found in biological active matter and researchers have recently started to explore these algorithms in active matter in some pioneering works. For example, reinforcement learning [14], which reflects a type of learning based on rewards, has been used to steer the motion of microscopic particles [15,16], to understand how birds can exploit turbulent thermal air flows to soar [6], to control the motion of artificial microswimmers in complex flow patterns [17] as well as in collective field taxis [18].

We are now at a critical crossroad in the development of active matter research where biological and artificial active matter are meeting with machine learning. The specific aim of this workshop is to bring together researchers from the fields of physics, biology, mathematics and machine learning to lay the groundwork of a scientific network to address the pressing questions:

1. What can machine learning do for biological active matter? Can we gain new insight into how powerful strategies have evolved? Can we understand the role of information processing, feedback cycles and sensorial delay in these strategies?

2. What can machine learning do for artificial active matter? Can we learn new approaches towards high-impact applications? For example, how can signaling and feedback be introduced into artificial active matter?

3. What insights can active matter provide for machine learning? Can we get some insight from the natural strategies optimized by evolution?

References

[1] Ramaswamy, S., The mechanics and statistics of active matter. Annu. Rev. Condens. Matter Phys. 1, 323–345 (2010).

[2] Marchetti, M. C. et al., Hydrodynamics of soft active matter. Rev. Mod. Phys. 85, 1143–1189 (2013).

[3] Katz, Y., Tunstrøm, K., Ioannou, C. C., Huepe, C., Couzin, I. D., Inferring the structure and dynamics of interactions in schooling fish. Proc. Natl. Acad. Sci. USA 108, 18720–18725 (2011).

[4] Yates, C. A. et al., Inherent noise can facilitate coherence in collective swarm motion. Proc. Natl. Acad. Sci. USA 106, 5464–5469 (2009).

[5] Kromer, J. A., Märcker, S., Lange, S., Baier, C., Friedrich, B. M., Decision making improves sperm chemotaxis in the presence of noise. PLoS Comput. Biol. 14, e1006109–15 (2018).

[6] Reddy, G., Celani, A., Sejnowski, T. J., Vergassola, M., Learning to soar in turbulent environments. Proc. Natl. Acad. Sci. USA 113, E4877–84 (2016).

[7] Bechinger, C. et al., Active particles in complex and crowded environments. Rev. Mod. Phys. 88, 045006 (2016).

[8] Pearce, D. J. G., Miller, A. M., Rowlands, G., Turner, M. S., Role of projection in the control of bird flocks. Proc. Natl. Acad. Sci. USA 111, 10422–10426 (2014).

[9] Bierbach, D. et al., Insights into the social behavior of surface and cave-dwelling fish (Poecilia mexicana) in light and darkness through the use of a biomimetic robot. Front. Robot. AI 5, 15 (2018).

[10] Buttinoni, I. et al., Dynamical clustering and phase separation in suspensions of self-propelled colloidal particles (2017).

[11] Qian, B., Montiel, D., Bregulla, A., Cichos, F., Yang, H., Harnessing thermal fluctuations for purposeful activities: the manipulation of single micro-swimmers by adaptive photon nudging. Chem. Sci. 4, 1420–1429 (2013).

[12] Mijalkov, M., McDaniel, A., Wehr, J., Volpe, G., Engineering sensorial delay to control phototaxis and emergent collective behaviors. Phys. Rev. X 6, 011008 (2016).

[13] Palmer, S. E., Marre, O., Berry, M. J., Bialek, W., Predictive information in a sensory population. Proc. Natl. Acad. Sci. USA 112, 6908–6913 (2015).

[14] Sutton, R. S., Barto, A. G., Reinforcement learning: an introduction. MIT Press, Cambridge (1998).

[15] Colabrese, S., Gustavsson, K., Celani, A., Biferale, L., Flow navigation by smart microswimmers via reinforcement learning. Phys. Rev. Lett. 118, 158004 (2017).

[16] Muiños-Landin, S., Ghazi-Zahedi, K., Cichos, F., Reinforcement learning of artificial microswimmers. arXiv 1803.06425v2 (2018).

[17] Gustavsson, K., Biferale, L., Celani, A., Colabrese, S., Finding efficient swimming strategies in a three-dimensional chaotic flow by reinforcement learning. Eur. Phys. J. E Soft Matter 40, 313–7 (2017).

[18] Palmer, G., Yaida, S., Optimizing collective fieldtaxis of swarming agents through reinforcement learning. arXiv 1709.02379 (2017).

Seminar by G. Volpe at MTL BrainHack School 2019, Montreal, Canada, 22 August 2019

Be friendly to your users:
Add comments and tutorials to your code
Giovanni Volpe
MTL BrainHack School 2019, Montreal, 22 August 2019
https://brainhackmtl.github.io/school2019/

When releasing a software package, it is critical to provide potential users with all the information they need to help them using it.
Using the example of Braph — a software we recently developed to study brain connectivity http://braph.org/ —, I’ll illustrate how we have commented the code, created a website and off-line documentation, and recoded video tutorials.
I’ll conclude with some practical advice and some best practices.

Talk by G. Volpe at SPIE OTOM XVI, San Diego, 14 Aug 2019

FORMA: a high-performance algorithm for the calibration of optical tweezers
Laura Pérez-García, Alejandro V. Arzola, Jaime Donlucas Pérez, Giorgio Volpe  & Giovanni Volpe
SPIE Nanoscience + Engineering, Optical trapping and Optical Manipulation XV, San Diego (CA), USA
11-15 August 2019

We introduce a powerful algorithm (FORMA) for the calibration of optical tweezers. FORMA estimates accurately the conservative and non-conservative components of the force field with important advantages over established techniques, being parameter-free, requiring ten-fold less data and executing orders-of-magnitude faster. We demonstrate FORMA performance using optical tweezers, showing how, outperforming other available techniques, it can identify and characterise stable and unstable equilibrium points in generic force fields.

Reference: Pérez-García et al., Nature Communications 9, 5166 (2018)
doi: 10.1038/s41467-018-07437-x