Hosts interactions and mobility

Understanding the spatiotemporal patterns of an infectious disease requires detailed information on the epidemiological coupling of populations, in addition to the biological and etiological knowledge on the pathogen. Such coupling is typically mediated by the mobility of hosts. In the last few years, thanks to the availability of massive data records and computational resources to analyze them, a large effort has been devoted to the investigation of the fundamental laws of human mobility at all scales. Furthermore, the integration of large mobility datasets in mathematical and computational models used in epidemiology have considerably increased their realism and allowed the explicit simulations of entire populations up to the scale of single individuals.
Our work in this direction is actively focusing on a number of different aspects realated to the role of human mobility and interactions in epidemic spreading.

networks of commuting and air travel in Europe

gravity law definition and test of the fit of the model (M) to the data (D)

First, we work on the collection and analysis of real mobility datasets, ranging from air travel to daily commuting in many different areas of the world, aimed at the understanding of human mobility processes and how they depend on cultural/geographic/socio-economic features. We perform such analyses by integrating methods of network theory, statistical physics and geographic information systems. 



In our analyses, we also adopt more traditional approaches of transportation theory, as, for instance, gravity models, that are based on the assumption that mobility is directly deterred by the costs associated to physical distance. In analogy with the Newton law, gravity models decribe the number of trips between any two given locations, by assuming that this number is proportional to the populations of the locations and decays with the distance between them. The results from our work include the definition of a gravity model able to provide a global description of commuting patterns in 29 countries, worldwide (Balcan et al, PNAS 2009).

  

In collaboration with the SocioPatterns project we are also interested in monitoring, analyzing and modeling human f2f interactions. These contacts are very relevant for the understanding of specific features of human behavior and they represent the means of transmission of respiratory infectious diseases. Thus, at a different scale with respect to human mobility in space, these network too play a crucial role in the epidemic diffusion. Data is collected by means of new sensing devices, and data-driven models can be formulated that allow the assessment of the impact of the various features of human interactions on epidemic spreading (see e.g. Stehle et al, BMC Medicine 2011).