The data here uses the model fitted to cell v4, repeated using the same random synaptic input. Antidromic stimulation (as in [21]) is simulated by adding spikes to the model, at a specified frequency and time. In the left column, spikes are added during the silent period, attempting to trigger a burst. In the right column, spikes are added during the second burst, attempting to terminate the burst. Burst triggering is more likely when stimulated later into the silent period, or using a more intense stimulation. Generally, burst termination requires a more intense stimulation than burst triggering. Successful termination is more likely later into the burst, when there is more dynorphin accumulation, or with a more intense stimulation. The competing effects of spike-triggered increases in [Ca2+]i and dynorphin cause a delay before termination occurs, unless the stimulation is sufficiently intense to trigger a large AHP, which immediately terminates spike firing.