Figure 3
Cluster’s activity and network coherence.
(A) Difference in cluster’s firing rate between the perturbed (φP) and the unperturbed (φ0) activities and comparing a control cavity (left, 32 clusters) with one targeted with M-Aβ complex (right, 36 clusters). Data corresponds to the experiments shown in Fig. 2A. Each bar represents a cluster of the network. Clusters are color coded and ordered in the horizontal axis according to their participation in a co-activated group. White bars depict clusters that fire independently. Bars marked with asterisks indicate the clusters that became silent after perturbation and the ones marked with arrowheads highlight those that boosted in activity. The top horizontal color boxes show the structure of co-activations before perturbation and color coded according to the sequences shown in Fig. 2A. Grey boxes are sequences that were not indicated in Fig. 2A. (B) Distributions of the normalized firing rate differences 
for 15 experimental realizations upon action of the different chemical agents. The red curve shows a Gaussian fit to the distributions, with mean μ and standard deviation σ. The value γ provides the skewness of the distributions.
