Seminario 5

Identification of coding-structures for cognition and behavior

Eleonora Russo

Neuronal dynamics emerge from a complex interplay of biological processes operating across diverse temporal scales. A significant challenge in neuroscience is to find structure within such variability and discern which components of this dynamics encode information and drive behavior. In this endeavor, unsupervised data mining presents a powerful opportunity. In this talk, I will discuss different unsupervised approaches for analyzing neuronal time series to uncover and model hidden regularities in experimental data. I will exemplify this approach with results from hippocampal, midbrain, and prefrontal regions.

Symmetries and Synchronization in Neuronal Systems

Tommaso Gili

Complex biological networks display a sophisticated interplay between the structural organization of their components’ interactions and the overall functionality of the system. This is especially evident in neuroscience, where the multiscale organisation of neurons generates a wide range of hierarchical functions, posing a longstanding challenge in systems neuroscience: understanding how ‘structure’ relates to ‘function’. The link between functional connectivity patterns in neuronal systems and the underlying structural connectivity architecture that governs their communication dynamics remains poorly understood, leaving the structure-function relationship still constrained by a one-to-many correspondence. In this work, we demonstrate that a novel class of non-global symmetries — known as fibration symmetries — which capture local dynamical regularities such as cluster synchronization, can be leveraged to infer distinct subsets of structural connections from the broader connectome, each associated with specific patterns of local dynamical activity.