Probing the dynamics of identified neurons with a data-driven modeling approach

Area of Science:

• Neuroscience
• Computational Biology
• Systems Biology

Background:

• Understanding the constraints on neuronal properties for specific behaviors is crucial but remains challenging.
• The nervous system must operate within defined limits for effective animal behavior control.

Purpose of the Study:

• To investigate the degree to which neuronal system properties are constrained at the single-neuron level.
• To develop and utilize data-driven models for analyzing neuronal dynamics and parameter constraints.

Main Methods:

• Collected large datasets of identified invertebrate neuron activity.
• Developed accurate conductance-based neuron models using automated parameter estimation.
• Analyzed autonomous and perturbed neuronal dynamics in both real neurons and models.

Main Results:

• Isolated neurons show higher variability than those within intact circuits.
• Responses to perturbations are more consistent than autonomous behavior.
• Model parameters exhibit a wide range of constraints, with some tightly controlled and others arbitrary.
• Irregular dynamics in the model originate from demonstrable chaoticity.
• Model predictions for ionic current blockade effects were generated.

Conclusions:

• Data-driven models are valuable for in-depth neuronal dynamics analysis.
• Focusing on controlled perturbations, rather than solely autonomous dynamics, offers a new paradigm for studying neural systems.
• The study highlights the role of chaotic dynamics and predicts the impact of channel blockers.