Symmetry of correlogram shape suggests common input

The shape of the correlogram has been previously used to characterize the likely nature of the underlying circuitry, whether the cells are mono-synaptically connected or receiving common input (Alonso and Martinez, 1998). Here, we applied the same methods to our data and compare our results against those reported for simple-complex pairs in area 17. (A) Correlogram Asymmetry (CA) is defined as the contrast ratio of the area of the correlogram right (green shading) and left (gray shading) of coincident activity (zero). Peak Asymmetry (PA) is defined as the contrast ratio of the distance along the abscissa from 70% height to the baseline. (B,top) A strong correlation between PA and CA is indicative of monosynaptic connections (+, data from Alonso and Martinez, 1998). However, we measured PA's that were closer to zero (more symmetric) despite comparable shifts in CA. This suggests that common input likely underlies the border-surface correlograms we measured. One possible source of such common input is recurrent connections within cortical regions (e.g. columns), connected by horizontal connections. (B,bottom) The bias in correogram asymmetry (abscissa) coincides with correlograms with taller peaks (i.e. asymmetric correlograms are not limited to weaker peaks).