Neural assemblies revealed by inferred connectivity-based models of prefrontal cortex recordings

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• Understanding neural coding requires identifying how groups of neurons coordinate activity.
• The prefrontal cortex is crucial for cognitive functions, including memory and decision-making.

Purpose of the Study:

• To develop and apply graphical model-based approaches for analyzing neural activity distributions in the prefrontal cortex.
• To identify and characterize neural cell assemblies, hypothesized units of neural coding and memory.

Main Methods:

• Inferred Ising model distribution and effective connectivity from multi-electrode recordings of neural activity.
• Simulated Ising model with an external drive to reveal multi-neuron configurations and cell assemblies.
• Inferred a Generalized Linear Model (GLM) to integrate spiking events over time using effective connectivity.
• Compared functional connectivity matrices from both Ising and GLM approaches.

Main Results:

• Identified cell assemblies that strongly coactivate in neural spiking data and are specific to the inferred connectivity network.
• The inferred connectivity network provides a sparse representation of neural correlation patterns.
• GLM sampling revealed spatio-temporal activation patterns within cell assemblies, including activation order.
• Activation order prevalence was weakly dependent on average firing rates and effective connection strengths.

Conclusions:

• Two graphical modeling approaches effectively identify neural cell assemblies in the prefrontal cortex.
• These methods provide insights into the structure and dynamics of neural assemblies related to coding and memory.
• The study highlights the utility of computational models in understanding complex neural activity patterns.