Skip to main content
. 2012 Apr 19;5:6. doi: 10.3389/fneng.2012.00006

Figure 4.

Figure 4

The proportions of blend interaction types found in 100 model realizations (μ ± SD, blue bars) are compared with the data values obtained from the experimental recordings of the neurons that could be successfully classified as PNs or LNs (N = 29 neurons, red squares). (A) PNs that respond with excitation (N = 12). (B) LNs that respond with excitation (N = 6). (C) PNs that respond with inhibition (N = 9). (D) LNs that respond with inhibition (N = 2). The number of neurons used to compute the proportions represented by the red squares are provided under their corresponding abscissa labels. For the model simulations of PNs that respond with excitation (blue bars in A) the neuron numbers were 397 for suppression, 306 for synergy, 240 for hypoadditivity, and 47 for linear addition. For the LNs that respond with excitation (blue bars in B) the neuron numbers were 504 for suppression, 516 for synergy, 306 for hypoadditivity, and 34 for linear addition. For the PNs that respond with inhibition (blue bars in C) the neuron numbers were 506 for suppression, 457 for synergy, 151 for hypoadditivity, and 42 for linear addition. For the LNs that respond with inhibition (blue bars in D) the neuron numbers were 116 for suppression, 63 for synergy, 28 for hypoadditivity, and 12 for linear addition.