Balanced Active Core in Heterogeneous Neuronal Networks.

Area of Science:

• Computational neuroscience
• Neural network dynamics
• Systems neuroscience

Background:

• Cortical neuronal circuits are hypothesized to operate in a global balanced state.
• Experimental observations show sparse coding, challenging the global balanced state hypothesis.
• Sparse coding involves a small set of active neurons encoding sensory information.

Purpose of the Study:

• To investigate the existence of a balanced state in a small subset of firing neurons within heterogeneous networks.
• To reconcile the concepts of balanced state and sparse coding in neuronal networks.
• To explore the role of network structure, such as scale-free and small-world networks, in maintaining balanced states.

Main Methods:

• Analysis of sparsely connected heterogeneous networks with diverse single-neuron dynamics.
• Investigation of scale-free and small-world network structures.
• Application of the Fokker-Planck equation to predict the dynamics of the active core.

Main Results:

• A small active subnetwork (active core) with high firing rates exists within larger heterogeneous networks.
• The active core exhibits a balanced state, despite the overall network being unbalanced.
• The active core's connectivity resembles a homogeneous Erdös-Rényi network.
• Fokker-Planck equation accurately predicts the dynamics of the active core.

Conclusions:

• A balanced state can be maintained by a small group of spiking neurons within a large, heterogeneous brain network.
• The active core reconciles the balanced state hypothesis with sparse coding observations.
• This finding offers a potential dynamical mechanism for sparse coding in neuronal systems.