Network Science as the Key for Understanding Complex Problems at Different Spatial Scales

Abstract

Network analysis is an emerging field of science. It offers a unique perspective to tackle complex problems, impenetrable to linear-proportional thinking. Network analysis aims at explaining complex phenomena emerging from simple principles of making network links locally. In this talk, I present three examples concerning very diverse systems: a protein-protein interaction network, an ecological system and the European pipeline network for natural gas distribution. Detection of modules in the protein-protein interaction network (i.e., the identification of cohesive sub-networks) and their significant enrichment with Alzheimer-related genes help to highlight the most relevant biological processes in this disease. Simulations are used to explicitly describe the dynamics of an ecological system at three hierarchical levels (i.e., population, food web and landscape network). They allow identifying how processes that are specific of a hierarchical layer can transgress its boundaries, thus affecting the others. Analysis of the European pipeline network contributes to detect the most strategic countries in natural gas delivery. Exposure and dominance of each country are estimated, showing that future scenarios will shift the relative weight of energy security concerns away from source to transit countries. A common property shared by these case studies is represented by the way network analysis allows shedding light on the whole system functioning: it helps to identify elements that would remain undiscovered with the analysis of small portions of the whole networks. Moreover, these studies borrow methods that were initially developed in other research areas. I argue that recent analytical advances, coupled with the enhanced potential of high-speed computation, have opened up new vistas to take on previously unimaginable challenges. ?