FIG. 5.

Example of a good prediction. A: PRCs obtained in open-loop configuration from the model (blue) and biological (red) neurons. Raw data points from the biological neuron's PRC are shown, along with a polynomial fit and top and bottom envelopes at ±2σ (dashed lines). Colored arrows indicate stimulus phases for each neuron where the stable mode of synchrony is predicted (see arrow in B). B: graphical method for determining existence of a 1:1 phase-locked mode (ts*–tr* plot). Parametric curves of ts* vs. tr* for model (blue) and biological (red) neurons are shown. Dashed lines indicate an envelope around the biological curve of ±2σ. Intersections reveal points where 1:1 phase-locked modes exist; out of 2 such points, one was calculated to be stable (arrow). C: schematic specifying the hybrid network. Model neuron 3 was coupled to biological neuron 1 using artificial synapses with maximal conductances (g_syn) as shown. PRCs were measured in open-loop configuration using stimuli scaled by g_syn as described in methods. D: uncoupled (D1) and coupled (D2) network activity. E: circular statistics. Dots along the perimeter of the circle represent instantaneous network phase, defined post hoc as ts_bio/P_network (see Circular statistics). We predicted a mode of phase-locking at μ_network = 0.77 (dotted arrow) and we observed a mode of phase-locking in this hybrid network at μ_network = 0.71 (solid arrow).