Realistic single-neuron modeling

ALL NEUROPHYSIOLOGISTS are modelers . They may explain the bursting behavior of a neuron in terms of voltage-sensitive calcium currents and a calcium-activated potassium conductance . They may conclude from a complex spike waveform that the cell has excitable dendrites . They may choose to voltage-clamp the cell body to study synaptic currents. Behind each of their experimental designs and interpretations of results lies an implicit hypothesis, a model, of how the neuron works . Here we describe a tool with which these mental models can be transformed into precise and explicit computer simulations. This transformation is in itself a valuable exercise, because it requires a complete list of the relevant biophysical parameters . Compiling this list may immediately reveal important gaps in knowledge. Of far greater value, however, is the expansion of intuition that comes with running many simulations over wide ranges of parameter values . Realistic models at the single-neuron level are primarily concerned with local changes in membrane current and voltage, caused by such things as synaptic input and voltage-sensitive conductances, and with the propagation of these local changes over spatially extensive dendrites and axons . With recent