Hot-spots of activity are associated with weak topological connectivity.

<p><b>A</b> Raster plots showing the response of model neuronal network to the localized lesion of gap junction connectivity. The extent of lesion is parameterized by the probability to destroy a connection between a pair of model neurons in the predefined area . The timing of lesion is marked with red line. Top panel: . Middle panel: . Bottom panel: . Other parameters are: . <b>B</b> Averaged firing rate (meanS.E.M.) in the lesioned subnetwork vs. the size of the lesioned subnetwork. Closed squares: . Open squares: . In all cases . Firing rate was computed over the time window of 20 seconds. <b>C</b> Averaged firing rate (meanS.E.M.) in the intact subnetwork, vs. the size of the lesioned subnetwork. Closed squares: . Open squares: . In all cases . For comparison, dashed line is the averaged neuronal firing rate in baseline conditions (model network with fully intact connectivity). Firing rate was computed over the time window of 20 seconds. <b>D</b> Raster plot showing the response of a model network with electrical and chemical synapses to the localized breach in gap junction connectivity. Disruption to gap junction connectivity was applied at time T = 2 seconds and is marked with red line. Chemical synaptic connections were added to the model as described in Methods.