Homeostatic structural plasticity increases the efficiency of small-world networks

Area of Science:

• Computational Neuroscience
• Network Science
• Systems Biology

Background:

• Small-world networks exhibit efficient information processing due to high clustering and short path lengths.
• The brain's structure resembles small-world networks, suggesting evolutionary optimization for neural information processing.
• Understanding the developmental mechanisms of efficient small-world network topology in the brain remains a key challenge.

Purpose of the Study:

• To investigate how a growth process favoring short-range connections and homeostatic synapse formation shapes neuronal network topology.
• To determine the impact of homeostasis on the efficiency and small-world properties of developing neuronal networks.
• To elucidate the self-organizing principles underlying the emergence of efficient brain network architectures.

Main Methods:

• Simulated a network growth process prioritizing short-range connections.
• Incorporated a homeostatic rule for synapse formation to maintain post-synaptic firing rates.
• Analyzed network topology, characteristic path length, clustering coefficient, and efficiency during development.

Main Results:

• The combination of growth rules and homeostasis led to the formation of small-world networks.
• Homeostasis significantly increased network efficiency, particularly as electrical activity approached the homeostatic set-point.
• Neurons formed more long-range connections near the homeostatic set-point, enhancing efficiency while preserving small-world properties.
• Small-world characteristics were maintained throughout the network's developmental process.

Conclusions:

• Homeostatic mechanisms play a crucial role in self-organizing efficient small-world network topologies in neuronal systems.
• These findings offer insights into how complex systems like the brain establish optimal network structures for information processing.
• The study suggests potential strategies for constructing large-scale neuronal networks through self-organization.