DISENTANGLING THE MOUSE BRAIN FUNCTIONAL CONNECTOME WITH THE GRAPH THEORY

The learning experience shapes the molecular landscape of the brain. The activation pattern of the response to different behavioural paradigms can be imaged using c-Fos protein, the molecular marker of neural activation. The analysis of this process involves imaging, detecting, summarizing activated cells across brain regions, and creating a functional network that could serve as a natural representation of global molecular expression patterns. In this study, mice were trained in the automated group housing cages with individual monitoring (IntelliCage) and exposed to either an appetitive (sucrose) or an aversive (quinine) stimulus. The control animals had access only to water. Subsequently, brains were collected and subjected to optical tissue clearing combined with immunostaining to detect c-Fos. Finally, brains were scanned using a light-sheet microscope, and images were aligned with the Allen Brain Atlas and analysed for c-Fos-positive cell counts. Graphs were created from correlation matrices of activation within experimental groups and across regions. The graph theoretical approach was directly compared with conventional analytical methods to assess whether a network-based perspective can reveal deeper insights into the spatial and functional organization of brain activation. Graph features that differentiate experimental groups were defined, and a bootstrap approach was employed to assess the significance of those differences. Betweenness centrality effectively differentiated the experimental groups and highlighted key nodes that act as critical connectors within the network. This work introduces easy-to-use, graph-oriented analytical pipelines and demonstrates that network-based metrics can differentiate between molecular landscapes. Support: National Science Centre, Poland (UMO-2019/35/B/NZ4/04077).