Network-based Observability and Controllability Analysis of Dynamical Systems: the NOCAD toolbox

Area of Science:

• Dynamical Systems
• Network Science
• Neuroscience

Background:

• Network science methods for driver node identification and sensor placement are increasingly used in dynamical systems.
• These methodologies offer powerful tools for understanding complex systems.
• Their application in life sciences, particularly neuroscience, is a growing area of interest.

Purpose of the Study:

• To introduce and demonstrate the applicability of network science-based methodologies in the life sciences.
• To analyze the neural network of Caenorhabditis elegans using these techniques.
• To present a novel NOCAD toolbox for structural controllability and observability analysis.

Main Methods:

• Application of network science principles to determine driver nodes and optimal sensor placement.
• Analysis of the Caenorhabditis elegans neural network structure.
• Development and utilization of an Octave/MATLAB-compatible NOCAD toolbox.
• Calculation of structural controllability and observability measures for linear/linearized systems.

Main Results:

• Demonstrated the successful application of network science methodologies to the C. elegans neural network.
• The NOCAD toolbox facilitates automated generation and comparison of controllability and observability measures.
• Identified key nodes and potential sensor locations within the neural network.

Conclusions:

• Network science provides a robust framework for analyzing complex biological systems like neural networks.
• The proposed NOCAD toolbox enhances the practical application of these methods in life sciences research.
• This approach aids in understanding system dynamics and optimizing monitoring strategies.