Generalized Leaky Integrate and Fire (GLIF) models are a simple point neuron models that have a striking ability to reproduce the spike times of real neurons. To explore how different phenomenological mechanisms contribute to spiking behavior, GLIF models of varying complexity are fit to electrophysiology data. In addition, optimized model parameters are useful for neuron classification. Models alongside electrophysiology data and more complex biophysical models are available in the Allen Cell Types Database.
Simplified neuron models are useful for both understanding the basic principles that recreate biological spiking behavior and populating network models. Detailed information about GLIF models at five different levels of abstraction have been described by Teeter, et al., (2018) “Generalized leaky integrate-and-fire models classify multiple neuron types.” This publication also contains a list of exemplary models for potential integration into network models (the Supplementary Material of the article contains a more detailed list). The article also describes how optimized GLIF parameters can classify cell types, and has code available via Github.
All data, along with models and documentation, are included in the Allen Cell Types Database, in the Cell Feature Search application. Simply filter data using “Has GLIF Model”, and information present on the page will contain only neurons that have at least one GLIF model. Note that other filters can simultaneously be selected to further refine data of interest. Individual models and neuron data can be viewed by clicking on one of the Cell Results panels, displayed below the search tool. Descriptions about the models and methods are available in the Help section, and Technical Whitepapers.