Calculated fractional occupancies of fast and slow recovery pathways during AP clamp waveform trains. Strategy for calculation of occupancies follows that in Fig. S1, with identical assumptions with initial state occupancies defined by the steady-state availability curve (Fig. 1 E; 0.958 [−80 mV], 0.890 [−70 mV], −0.719 [−60 mV], 0.400 [−50 mV]), and that initial depolarization leads to even distribution between fast and slow recovery paths. For the impact of the afterhyperpolarization, we assumed a 34.6-ms afterhyperpolarization centered at −50 mV. For 4-Hz stimulation, there is then a 200.2-ms interval at the holding potential; for 10 Hz, there is 49.8 ms; and for 20 Hz, only 1.4 ms. Thus, for 20-Hz stimulation, the recovery is dominated by the tail current period at −50 mV, rather than the holding potential. For the AP clamp waveform, there is also 5 ms at −50 mV that precedes the upswing of the AP voltage waveform. Given the slow rates of onset or recovery from inactivation, this 5-ms interval has negligible impact on changes in occupancy of inactivated states. This procedure essentially defines the expected state occupancies immediately preceding each AP. (A–C) Calculated state occupancies for a 20-Hz AP train for −50, −60, and −70 mV holding potentials are shown. Following the initial AP waveform, at the time of the second AP waveform, there is increased occupancy of both fast- and slow-recovering pathways, but with subsequent Aps, the fraction of channels in slow recovery pathways increases, while those in fast recovery pathways decreases. (D–F) State occupancies are plotted for the 10-Hz cases, qualitatively exhibiting behavior similar to that for 20 Hz. (G–I) State occupancies plotted for the 4 Hz-trains.