Figure 8
Learning effects produced in the model by altering NMDA receptor sensitivity. Graphs show changes (after learning = black; during learning = gray) in (A) Theta and gamma amplitude as a function of stimulus strength (Iapp). (B) Theta/gamma ratio, (C) Coherence between theta phase and gamma amplitude (D) Firing rate of the excitatory output neurons (E) Firing rate of the downstream neuron and (F) Positive correlation between downstream neuron firing rate and magnitude of the theta/gamma ratio (r = 0.34, P < 0.01). NMDA, AMPA and GABA A receptor coefficients after learning are the same as for shallow nested gamma in Figure 8B. For during learning NMee = 0.002 and NMes = 0.0001; after learning NMee = 0.0035 and NMes = 0.00055. Data are means ± sems from 10 averaged runs of the model. Taking an overall average across the different values of Iapp, t-tests revealed significant differences between before and after simulated learning in theta amplitude (A), t18 = 81.5, p < 0.0001; gamma amplitude (B), t18 = -12.1, p < 0.0001; theta/gamma ratio (C), t18 = 32.02, p < 0.0001; theta/gamma coherence (D), t18 = 2.6, p = 0.03; excitatory neuron firing rate (E), t18 = -2.23, p = 0. 04 and the firing rate of the downstream neuron (F), t18 = 13.6, p < 0.0001.