< Back to Article

A Self-Organizing State-Space-Model Approach for Parameter Estimation in Hodgkin-Huxley-Type Models of Single Neurons

Figure 10

Simultaneous estimation of hidden model states, maximal conductances and kinetic parameters in a two-compartment model of a vertebrate motoneuron (I).

Estimation was based on two simulated -long simultaneous recordings of the membrane potential from the soma and dendritic compartment. Only part of this data is illustrated for clarity. Notice the different time scales between the left and right panels. (A) High-fidelity smoothing of the observed voltage at the soma (Ai) and the dendrite (Aii). (B) Inference of unobserved calcium concentrations at the soma (Bi) and dendritic compartment (Bii). (C) Inference of the unobserved activation and inactivation variables for all voltage-gated currents at the soma and the dendrite. Since the kinetics of voltage-gated currents were assumed unknown, the difference between true (black lines) and inferred (red lines) dynamic variables was significant (compare to Fig. 8). The inferred parameters are shown in Fig. 7Ai. In these simulations, , , and the prior interval for was .

doi: https://doi.org/10.1371/journal.pcbi.1002401.g010