Interesting Characteristics of Epidemiological Processes Spreading on Bipartite Networks

In bipartite networks nodes are divided into two classes and edges are allowed only between individuals belonging to different families. Until now, bipartite networks were proposed to model and analyze different phenomena. For example [3], bipartite networks are used to describe co-actor networks (having movie and actor nodes), or academic collaboration networks (where researchers are bind with their articles). Bipartite heterosexual contact networks were also proposed to study the spreading of sexually transmitted diseases [2]. Always in epidemiological studies, this class of networks was recently adopted to describe contacts between ticks and hosts [1], an important feature when facing the challenge of studying vector-borne disease transmission dynamics. Tick-borne diseases such as Tick Borne Encephalitis or Lyme Borelliosis are emergent zoonoses of great interest for public health in temperate areas. By observing a pathogen spreading among nodes on scale-free bipartite networks we report two interesting results: on one hand, a lower epidemic threshold with respect to the one obtained on a unipartite network with the same connectivity. On the other hand, a form of synchronization of oscillations of the pathogen is observable. In fact, after a thermalization period for the dynamical system we found that pathogens spring in a family of nodes but not in the other one, while in the next iteration pathogens are among nodes of the second class but not in the first one. Therefore the infection alternates its presence between the two sets of nodes of the bipartite network. These two results are of great interest from the eco-epidemiological point of view. The fact the epidemic threshold is lower in bipartite networks could result in an evolutionary advantage for pathogens spreading over them. We could infer that vector-borne diseases are, from an evolutionary point of view, subject to a lower selection pressure than other pathogen spreading on unipartite contact newtorks. Moreover, the phenomenon of the oscillation synchronization is observable in natural systems. Tick nymphs are active in Spring and susceptible larvae are active during the Summer of the same year. Hence, in order for the pathogen to be passed to a next generation of larvae, hosts are required to keep the pathogen from Spring to Summer, having been first infected by nymphs and then transmitting to larvae. References [1] Bisanzio et al. PLoS ONE, 5(11):e13796, 11 2010. [2] Gomez-Gardenes et al. PNAS, 105(5):1399–1404, 2008. [3] M. E. J. Newman. Networks: An Introduction. Oxford University Press, Oxford, UK, 2010.