Reconstruction of network structures from repeating spike patterns in simulated bursting dynamics

Area of Science:

• Computational Neuroscience
• Network Science
• Systems Biology

Background:

• Neuronal network topology is fundamental to brain function.
• Reconstructing network structure from dynamic activity is a key challenge.
• Repeating spike patterns have been proposed as a potential method for topology reconstruction.

Purpose of the Study:

• To evaluate the effectiveness of a pattern-matching method using repeating spike sequences for reconstructing neuronal network topology.
• To investigate the accuracy of reconstructing global network properties and identifying network hubs.
• To assess the method's utility in physiologically realistic models with diverse connection topologies.

Main Methods:

• Simulated reverberations in a physiologically realistic neuronal network model.
• Employed various physical connection topologies, including random and scale-free networks.
• Utilized a pattern-matching approach based on repeating spike sequences to reconstruct network dynamics.

Main Results:

• Reconstructed functional networks derived from spike patterns often differ significantly from original physical networks.
• Global network properties, such as degree distribution, were not always accurately recovered.
• The pattern-matching method proved effective in identifying network hubs.

Conclusions:

• Reconstructing precise neuronal network topology from spike patterns alone is difficult and may not preserve global properties.
• The presence of hubs can be reliably detected by analyzing reverberation patterns combined with spike pattern reconstruction.
• This approach offers a potential method for identifying hubs in neuronal cultures based on network dynamics.