A Population Coupling Model Identifies Reduced Propagation from V1 to Higher Visual Areas During Locomotion.

Area of Science:

• Systems Neuroscience
• Computational Neuroscience

Background:

• Generalized linear models (GLMs) are standard for analyzing neural population activity.
• Modeling individual neuron interactions can be complex and redundant when neurons share similar firing patterns.
• Population-level statistics offer a more efficient approach for capturing cross-population interactions.

Purpose of the Study:

• To reformulate the GLM framework for direct population-level analysis.
• To develop a computationally efficient method for estimating coupling between neuronal populations.
• To improve sensitivity and robustness in detecting functional connectivity.

Main Methods:

• Developed the population generalized linear model (pop-GLM).
• Applied pop-GLM to simulated data to assess sensitivity and bias correction.
• Utilized pop-GLM on real neural data from visual cortex.

Main Results:

• pop-GLM demonstrated greater sensitivity in detecting coupling effects compared to single-neuron GLMs.
• The model effectively accounts for trial-to-trial variations in stimulus drive, reducing bias.
• A decrease in functional connectivity from V1 to a higher visual area during locomotion was identified, which single-neuron GLMs missed.

Conclusions:

• The pop-GLM offers a computationally efficient and sensitive method for analyzing neural population interactions.
• This population-level approach reveals functional connectivity changes not detectable by traditional single-neuron models.
• pop-GLM advances the understanding of coordinated circuit dynamics in the brain.