Research Political Sciences

The Network Modeling Chip that Fights Influenza

, by Claudio Todesco
Big data assists medicine in Alessia Melegaro's current project

Chips and coding can help us to understand how infectious diseases spread and which vaccination strategies to adopt. This is an ongoing project by Alessia Melegaro (Department of Policy Analysis and Public Management), in collaboration with ISI Foundation, Fondazione Bruno Kessler, and Kenya Medical Research Institute.
The goal of Who contacts whom? Social contact networks in Kilifi is the development of a model to study the human respiratory syncytial virus that affects infants and for which there is still no vaccine available. In Europe, the virus causes flu-like symptoms, in poor Countries it is associated with a high mortality rate.

The research is based on data collected in 2016-2017. Students from two schools in Kenya and a sub-sample of their families brought with them for a week a chip as large as a coin. The chip, which was developed by SocioPatterns, recorded each time two persons went close enough to transmit an infection. "Information on these interactions has enabled us to define a network of relationships between individuals and groups", Melegaro says. Before that, researchers had to perform data cleaning. The number of nightly interactions, for instance, was very high because all family members put their chips on the same table before going to bed. "The preliminary work was very long and tricky. We then used a software to analyze the relationship networks and to study and look at them from different angles". Afterwards, they developed a code for that particular demographic, that is, a code that takes into account information such as age distribution and household size.

The researchers are now integrating the transmission mechanisms of the respiratory syncytial virus into the model and testing its robustness using the existing epidemiological data. "In the past, computer simulation models of the spread of infectious diseases were based on theoretical assumptions about social interactions. Today, realistic network-based models allow us to more accurately identify the most effective practical vaccination strategy".

Read more about this topic:
Riccardo Zecchina. Teaching Machines How to Learn to Improve Business and Life
Carlo Baldassi. Learning Is a Quantum Question
Daniele Durante. How to Study Dynamic Networks
Dirk Hovy. The algorithm that Prevents Suicide
Antonio Lijoi, Igor Pruenster. Algorithms for Analysis of Complicated Phenomena
Raffaella Piccarreta, Marco Bonetti. A (Statistical) Model for Life