Fernando Nóbrega Santos, Research Associate at the Department of Anatomy and Neuroscience of the Amsterdam AMC, will kick off his fellowship at IAS with an online lecture. He will present his research topic and explore possible links with researchers from other fields.
|Date||4 March 2021|
The IAS fellowship allows me to zoom into a central question of my research: how to better understand complex systems, such as the human brain, pandemics or financial markets. Network theory, mobilized to study such complex systems, is usually based on the analysis of pairwise interactions, i.e. relationships between two nodes of a network. This is not realistic for many real-world complex systems, as it does not accurately capture high-order interactions, i.e. between three or more constituents of a network. As an interdisciplinary theoretical physicist, I also want to bring insights from topology, statistical mechanics, geometry and information theory to the analysis of complex systems.
The talk will outline two building blocks of my fellowship: First, I would like to explore the relationship between topology and theoretical physics to further understand complex systems. We showed in earlier work that it is possible to characterize topological phase transitions in complex systems in a reliable manner. Our research also suggested the possibility of finding high-order network analogues to the relationship between topology and theoretical physics that can lead to basic principles in network science and numerous applications. I want to move this research one step further by linking topology, theoretical physics, and information theory. Given IAS’ strong track record in multivariate information theory, and that this is currently one of the most promising strategies to infer high-order interactions in networks, I am particularly excited to discuss these theoretical bridges with IAS fellows.
Second, I would like to apply such advanced analysis to high-order interactions in brain networks. In particular, I want to know how the topology of the human brain network relates to behavioral properties at individual and group level. From a clinical network neuroscience perspective, the connection between brain networks and behavioural traits is not well-understood despite recent brain networks research successes. So far, network neuroscience has been mainly looking at pairwise interactions. In my research, I would like to propose including high-order metrics in brain network analysis as an exciting, realistic route for this quest. With colleagues at VUMC, we will apply such advanced metrics to state-of-the-art clinical neuroscience databases covering diseases such as Glioma and Multiple Sclerosis.