ABSTRACT
Small networks of neurons, particularly those from invertebrate preparations where it is possible to work with identified neurons, are attractive objects for modeling using conductance-based or “realistic” neuron models. Often there is the possibility of using voltage clamp techniques to characterize the membrane currents of the component neurons and their synaptic interactions, so that sensible conductancebased model neurons can be constructed. Moreover, for small networks it is often possible to quantify precisely the network behavior that one wishes to simulate in
the network model so that a benchmark for model performance can be established. The aim of this chapter is to provide some tools useful in modeling synaptic interactions in small network models and to lay out a strategy for using these tools in constructing network models. Particular attention is given to the processes of parameterization and model testing. There is above all an emphasis on the interactive nature of realistic modeling with the model being developed from experimental observations and in turn generating predictions that are experimentally testable. The chapter also presents an example from our own work that we hope illustrates this strategy. The chapter ends with a discussion of how abstract models can inspire, supplement, and illuminate the more realistic models developed with the tools and strategy presented here.
