Example of detection for excitatory and inhibitory links in a network model.

a, NCCH computed between two spike trains related to two neurons linked by an excitatory link in the model (identified by a red asterisk). b. FNCCH of the two neurons of panel a. The “entity peak” allows a better recognition of the excitatory link. c. NCCH computed between two spike trains related to two neurons linked by an inhibitory link in the model. The NCCH might detect a false excitatory peak (blue cross). d, FNCCH of the two neurons of panel c. The filtering procedure allows to recognize the through and to detect the negative peak correspondent to the inhibitory link (blue asterisk). e. Schematic representation and description of the algorithm to obtain the FNCCH. f. FNCCH (dashed line) for two putative neurons connected by a putative excitatory connection. g. FNCCH (dashed line) for two putative neurons connected by a putative inhibitory connection. The continuous lines in panels f and g are a smoothing of the histogram (dashed lines).