Systematic generation of biophysically detailed models for diverse cortical neuron types

Area of Science:

• Neuroscience
• Computational Biology
• Biophysics

Background:

• Mammalian neocortical circuits exhibit significant cellular diversity, posing challenges for accurate computational modeling.
• Bridging experimental characterization of cell types with computational model construction is crucial for understanding neural function.

Purpose of the Study:

• To develop a method for generating biophysically detailed computational models of individual neurons.
• To link systematic experimental characterization of cell types with the construction of cortical models.

Main Methods:

• Utilized 3D morphologies and somatic electrophysiological recordings from the Allen Cell Types Database for model building.
• Employed a genetic algorithm to optimize active somatic conductances and other parameters, matching experimental electrophysiological features.
• Validated models by comparing their responses to novel stimuli against experimental data.

Main Results:

• Successfully generated biophysically detailed models for 170 individual neurons.
• Demonstrated that the models preserve distinct intrinsic properties between different neuronal subsets from mouse primary visual cortex.
• Verified model accuracy through comparison with experimental electrophysiological data.

Conclusions:

• The developed approach effectively links experimental cell type characterization to the creation of detailed cortical models.
• The generated neuron models and associated code are openly accessible, facilitating further research in computational neuroscience.
• This work advances the ability to computationally represent the diversity of mammalian neocortical circuits.