One system, two perspectives: The role of modeling assumptions ininterconnected brain and epidemic systems

ABSTRACT
We will examine two methodological contributions in the field of complex networks, oriented towards two different problems regarding the betterment of human health.
First, we will explore two representations of multilayer networks (i.e. networks with several types of interactions), arising from the node-layer duality. Through extensive computations and analytical results, we will show that their structural properties are complementary and irreducible from one another. Then, we will examine the potential of this dual representation in deriving biomarkers for Alzheimer’s disease using patients data, multilayer networks being a convenient formalism in brain studies.
Second, we will explore how the modeling assumptions on human mobility impacts mosquito-borne epidemic risk estimates in the particular context of non-endemic areas, at risk of climate change-induced mosquito-borne outbreaks. For that, we present two alternative SEIR-SEI metapopulation network models where local human mobility is represented either through force-of-infection where movements of individuals are implicit or through physical diffusion where those movements are explicit. We examine the mechanisms in the models that could yield divergent risk and timing of epidemic spread in real-world networks, taking Italy as a real-world example.
Together, these studies present how different perspectives on interconnected systems shape our understanding of their properties.