Abstract
Conductance-based models of neural activity produce large amounts of data that can be hard to visualize and interpret. Here we introduce two novel visualization methods to display the dynamics of the ionic currents, and to investigate how the contribution of each current changes in response to perturbation. We explored the solutions of a single compartment, conductance-based model of neural activity with seven voltage-gated ionic currents and a leak channel. We employed landscape optimization to find sets of maximal conductances that produce similar target activity and displayed the dynamics of the currents. We examined in detail six examples of a bursting model neuron that differ as much as 3-fold in the conductance densities of each of the 8 currents in the model. The maximal conductance of each current does not simply predict the importance of the current for neuronal dynamics. We then compared the effects of systematically reducing the conductances of each current for neuronal dynamics, and demonstrate that models that appear similar under starting conditions behave dramatically differently to the decreases in conductance densities. These examples provide heuristic insight into why individuals with similar behavior can nonetheless respond widely differently to perturbations.