Strategy for fitting neuronal models to dual patch data under multiple stimulation protocols

It is a difficult task to fit a neuronal model to data under multiple stimulation protocols. The protocol of weakest stimulation was used first to fit the model at the initial stage. The strategy of fitting an action potential (AP) consisted of three parts: passive parameter fitting, reduction of spatial complexity, reduction of temporal complexity. The nearly linear response of a neuron in the early depolarization well before AP initiation was used to fit the passive parameters with all the ion channel densities set to zero. We separated the fitting of the intervening part between the two electrodes from that beyond the dendritic electrode to reduce the spatial complexity by using dendritic voltage clamp simulation. The decomposition of stimulation protocols and the time course of an AP proved feasible and successful.