Inférence de réseaux de régulation de gènes au travers de scores étendus dans les réseaux bayésiens.

Jimmy Vandel, Brigitte Mangin, Matthieu Vignes, Damien Leroux, Olivier Loudet, Marie-Laure Martin-Magniette, Simon de Givry

Abstract: Author Summary The cortex is a highly modular structure with a large number of functionally specialized areas that communicate with each other through long-range cortical connections. It is has been suggested that communication between spiking neuronal networks (SNNs) requires synchronization of spiking activity which is either provided by the flow of neuronal activity across divergent/convergent connections, as suggested by computational models of SNNs, or by local oscillations in the gamma frequency band (30–100 Hz). However, such communication requires unphysiologically dense/strong connectivity, and the mechanisms required to synchronize separated local oscillators remain poorly understood. Here, we present a novel mechanism that alleviates these shortcomings and enables the propagation synchrony across weakly connected SNNs by locally amplifying feeble synchronization through resonance that naturally occurs in oscillating networks of excitatory and inhibitory neurons. We show that oscillatory stimuli at the network resonance frequencies generate a slowly propagating oscillation that is synchronized across the distributed networks. Moreover, communication with such oscillations depends on the dynamical state of the background activity in the SNN. Our results suggest that the emergence of synchronized oscillations can be viewed as a consequence of spiking activity propagation in weakly connected networks that is supported by resonance and modulated by the dynamics of the ongoing activity.