Evolving Simple Models of Diverse Intrinsic Dynamics in Hippocampal Neuron Types

Area of Science:

• Computational Neuroscience
• Neural Modeling
• Systems Neuroscience

Background:

• Neuronal intrinsic dynamics enrich circuit computations and emergent network properties like synchronization.
• Accurate large-scale spiking network models are crucial for testing neural computation theories.
• Existing methods struggle to quantitatively match experimental voltage traces using the Izhikevich model.

Purpose of the Study:

• To develop an automated pipeline for quantitatively reproducing neuronal spike pattern features using the Izhikevich model.
• To accurately capture the biological diversity of individual neurons in network simulations.
• To create multiple, suitable Izhikevich models for each neuron type, reflecting feature variability.

Main Methods:

• An automated pipeline based on evolutionary algorithms was developed.
• Experimental data from Hippocampome.org was used to define target spike pattern classes.
• The pipeline quantitatively fits Izhikevich models to diverse, experimentally recorded spike patterns.

Main Results:

• The approach reliably fitted Izhikevich models to nine distinct classes of hippocampal neuron spike patterns.
• Multiple suitable models were generated for each neuron type, capturing feature variability.
• Both single-point and multi-compartment Izhikevich models were successfully created.

Conclusions:

• The developed pipeline offers a robust method for quantitatively modeling neuronal diversity.
• This approach enhances the accuracy of large-scale neural network simulations.
• The methodology is broadly applicable to various neural systems beyond the hippocampus.