Introduction
Computational neuroscience modeling involves the difficult task of choosing an appropriate model that represents neuronal spiking dynamics. Biologically realistic but complex conductance-based neuronal models were pioneered by Hodgkin and Huxley. These models are usually difficult to develop and analyze, particularly in the case of mammalian neurons which have complex morphologies and involve a large number of voltage- and calcium- dependent conductances of varying time scales. Other possibilities include abstract models such as integrate-and-fire or the Izhikevich model and reduced biophysical models such as FitzHugh-Nagumo and Morris-Lecar, which are conceptually and mathematically simple but their parameters can not be related to measurable biophysical phenomena such as ionic conductances in the cell membrane. The abstract models allow for abstract studies of neuronal information processing but can not address questions such as how blocking particular currents affects brain activity.
The development of neuronal spiking models has accordingly shaped the development of neuronal modeling software. Several simulators exist that are capable of modeling conductance-based models of morphologically detailed neurons, as well as networks constructed from these models. These are exemplified by NEURON and Genesis. Other simulators such as NEST and the NeoCortical Simulator (NCS) are designed for large-scale network modeling and trade complexity for performance by using only integrate-and-fire or similar point neuron models. These rely on Discrete Event Simulation techniques to minimize the communication between nodes during simulation, and thus achieve high computational performance.
However, as computer technology evolves, the choices of modeling software and neuron models need not be binary. A number of studies exist on modeling with reduced cell morphologies and modern hardware is powerful enough to solve even fairly large systems of equations that represent the dynamics of membrane conductances and ionic concentrations.