Multiobjective identification of controlling areas in neuronal networks

Area of Science:

• Neuroscience
• Complex Network Theory
• Computational Intelligence

Background:

• Identifying controlling areas in neuronal networks is crucial for understanding brain function.
• Existing methods often struggle with the complexity and multiobjective nature of neuronal control.

Purpose of the Study:

• To develop and evaluate a novel multiobjective optimization method for identifying controlling areas in a cat's brain neuronal network.
• To address the challenge of simultaneously considering multiple controllability measures.

Main Methods:

• Development of a reference-point-based nondominated sorting composite differential evolution (RP-NSCDE) algorithm.
• Utilizing nondominated sorting mechanisms and composite differential evolution (CoDE).
• Comparison with nondominated sorting genetic algorithms II and other statistical/optimization methods.

Main Results:

• The proposed RP-NSCDE demonstrated superior accuracy and convergence speed compared to existing algorithms.
• A tradeoff between minimizing two objectives was identified, visualized through Pareto fronts (PFs).
• Effectiveness and reliability were validated against various benchmark methods.

Conclusions:

• The RP-NSCDE is a promising tool for multiobjective identification of controlling areas in complex neuronal networks.
• The findings highlight the existence of tradeoffs in network control optimization.
• The developed approach has potential applications in coordinating various real-world complex networks, including biological and social systems.