Plenary Speakers - Abstracts

Helen Byrne
Oxford University | UK

Antonio DeSimone


MathLab@SISSA-International School for Advanced Studies, Italy

Abstract: Cell motility is key to many biological functions, and it is accomplished by coordinated shape changes. Locomotion strategies employed by unicellular organisms are particularly interesting because they are invisible to the naked eye, and offer surprising new solutions to the question of how shape can be programmed. In recent years, we have studied locomotion by shape control using a variety of methods: modeling, theory, and numerical simulation, observations at the microscope, manufacturing of prototypes. A concrete case study is provided by our results on Euglena gracilis, a unicellular protist that is able to move both by agellar propulsion and by highly coordinated changes of the shape of the whole cell body [1, 2]. We will survey the most recent ndings within this stream of research, and point out to current directions and challenges for the future.

Keywords: cell motility, shape control, mechano-biology, active matter.

Acknowledgements: This work has been supported by the ERC Advanced Grant 340685-MicroMotility.


[1] M. Rossi, G. Cicconofri, A. Beran, G. Noselli, A. DeSimone (2017). Kinematics of agellar swimming in Euglena gracilis: Helical trajectories and agellar shapes, Proc Nat Acad Sci USA 114,13085{13090

[2] G. Noselli, A. Beran, M. Arroyo, A. DeSimone (2018). Experimental and theoretical study of metaboly in Euglena gracilis. Preprint.

Eva Kisdi


Department of Mathematics and Statistics, University of Helsinki, Finland

Abstract: Natural selection is usually paraphrased as the survival of the fittest - or the demise of all others. How can natural selection explain the enormous diversity of variants living together in Nature? Adaptive dynamics nds the answer in deriving fitness explicitly from models of population dynamics. This leads to an ever-changing
fitness landscape, which facilitates not only the coexistence of multiple species but also the formation of new lineages through a process called evolutionary branching.
After a brief introduction to the mathematical framework of adaptive dynamics, I consider three questions relating to diversity. First, is there an upper bound to the
number of species, and if so, how does a "saturated" community evolve? Second, can natural selection lead to extinction? Third, when diversity evolves, it may be
just variation but not speciation. Will natural selection lead to the origin of new species?

Keywords: adaptive dynamics, evolutionary branching, evolutionary suicide, speciation.


Samuel Kou


Department of Statistics, Harvard University, USA

Abstract: Big data collected from the Internet have generated signi cant interest in not only the academic community but also industry and government agencies. They bring great potential in tracking and predicting massive social activities. We focus on tracking disease epidemics in this talk. We will discuss the applications, in particular, Google Flu Trends, some of the fallacy and the statistical implications. We will propose a new model that utilizes publicly available online data to estimate disease
epidemics. Our model outperforms all previous real-time tracking models for influenza epidemics at the national level of the US. An extended version of the model gives accurate tracking of Dengue fever in Asian and South American countries. We will also draw some lessons for big data applications.

Keywords: influenza, dengue fever, Internet search, forecasting.

Mirjam Kretzschmar


Julius Center for Health Sciences and Primary Care, University Medical Center, Utrecht, The Netherlands

Joint work with Odo Diekmann (Utrecht University), Wilfred de Graaf (Utrecht University) and Peter Teunis (RIVM).

Abstract: The immune status of an individual host is determined by the increase of immunity during infection, waning of immunity after clearing the infection, and boosting
by renewed exposure to the pathogen. The process of boosting, the rate at which immunity wanes, and the level of protection it confers, all influence the transmission
dynamics of the pathogen. Information about the immune status of a population is often available from serological studies, but it may be unclear what this means
for the level of protection against infection or symptomatic disease. We would like to understand how an intervention changes a population's immune status and the
incidence of symptomatic infection for an infectious disease with waning immunity. In this talk I will introduce a mathematical model for the waning and boosting of
immunity. The model is defined on two levels. On the within-host level we defined a model that distinguishes between episodes of infection and time periods of waning
of immunity (De Graaf et al. 2014). During infection, a simple 2-dimensional system of ODE's describes the time evolution of pathogens and immunity within
the host. Between infection episodes immunity wanes until a new exposure triggers the next infection episode. We then lift the model to the population level by
studying the distribution of immune states in a population under the assumption of a constant force of infection. The events of exposure and infection are described
by a time-homogeneous Poisson process, between exposures immune status wanes deterministically. This model can be formulated in terms of a renewal equation, for
which a stable stationary distribution can be derived. The modelling framework will be illustrated with applications to pertussis epidemiology. For pertussis, longitudinal and cross-sectional serological data are available, which can be used to parameterize the model. We were interested in obtaining estimates for the incidence of symptomatic infections, the ratio of symptomatic to asymptomatic infections, and the immune level at which protection from symptomatic infection occurs. We found remarkable correspondence between predictions of the within-host model with observations reported in the literature concerning the serological correlate of protection. The modelling framework has strong links with a statistical model used for estimating incidence from serological data.

Keywords: Mathematical model, Waning immunity.


[1] de Graaf WF, Kretzschmar ME, Teunis PF, Diekmann O. (2014). A two-phase within-host model for immune response and its application to serological profiles of
pertussis, Epidemics 9:1-7.

Eva Löcherbach


Laboratoire AGM UMR CNRS 8088 Université de Cergy-Pontoise, France

Abstract: A class of recently introduced models to describe networks of neurons as stochastic processes with memory of variable length will be presented. These are non-
Markovian processes in high or infinite dimension in which the past dependence of transition probabilities or intensities has a range that is finite but depends on the
particular history. Starting from existence results, we study related mean-field models in continuous time and their large population limits, and discuss the relation with associated Piecewise Deterministic Markov Processes (PDMP's) and state results concerning their longtime behavior. Finally, two important problems of statistical inference in such models will be considered: estimation of the spiking rate function and estimation of the neuronal interaction graph. The talk is based on joint work with Susanne Ditlevsen, Aline Duarte, Antonio Galves and Guilherme Ost.

Keywords: Multivariate nonlinear Hawkes processes, Mean-field approximations, Piecewise deterministic Markov processes, Multi-class systems, Oscillations.

Andrea Pugliese
Trento University | Italy

Eörs Szathmáry


Evolutionary Systems Research Group, MTA Ecologogical Research Center, Tihany, Hungary, and Parmenides Centerfor the Conceptual Foundations of Science, Pullach, Germany

Abstract: In the past several scholars have noted some relationship between learning and evolution and various levels of abstraction. William James was wondering about the possible role of a process analogous to evolution of natural selection in the brain, whereby adaptive answers to complex problems might arise. Changeux and Edelman were considering selectionist approaches to brain dynamics during its development: while their approach has been experimentally validated, replicator dynamics has not been entertained by them. The first question is then whether true evolutionary dynamics can unfold in the brain (evolution in learning). On the flip side of the coin Richard Watson has raised the idea whether associative, reinforcement and deep learning dynamics could play a role in the evolution of ecosystems, developmental genetic regulatory networks and evolutionary transitions in individuality (learning in evolution). A third, potentially unifying theme is the analogy between Bayesian inference and the discrete-time replicator equation: here the question is whether similar algorithms could realize either of them in some natural systems. I shall review the relevant concepts and mathematical formulations behind these ideas. Open questions will be raised that, if answered positively, could entail that there will ultimately be only one uni ed theory including evolution and learning as subcases.

Keywords: replicator equation, Bayesian models, Hebb synapse, reinforcement, adaptation.

Acknowledgements: This project is supported by the Templeton World Charity Foundation (Learning in evolution, evolution in learning) and by funded by National
Research, Development, and Innovation Office Grants NKFI-K119347 and GINOP-2.3.2-15-2016-00057.

Kees Weijer
Dundee University | UK