Realistic Models of Neurons Require Quantitative Information at the Single-cell Level
Detailed modelling of neurons is now a recognised sub-field of neurobiology. Such models rely on accurate and quantitative experimental measurements. For instance, modelling electrophysiology requires morphological reconstructions of identified neurons. Similarly, understanding the biochemical basis of neurotransmission becomes possible if we know about the molecular composition of the connected neurons. In this chapter we will describe the pitfalls of generic models that seek to reproduce common features of groups of neurons, and in particular, the artifacts generated by an excessive abstraction. Instead, we advocate the development of typological models, seeking to describe accurately a given neuron, generic inferences being derived afterwards.