Consistent estimation of complete neuronal connectivity in large neuronal populations using sparse "shotgun" neuronal...

Area of Science:

• Computational Neuroscience
• Systems Neuroscience
• Network Science

Background:

• Estimating neuronal connectivity is crucial for understanding brain function.
• The common inputs problem presents a significant challenge in functional connectivity estimation.
• Shotgun sampling is a novel approach proposed to address this challenge.

Purpose of the Study:

• To evaluate the properties of the shotgun sampling approach for functional estimation of neuronal connectivity.
• To analyze the asymptotic correctness, convergence speed, and data requirements of the shotgun method.
• To determine the conditions under which complete connectivity matrix inference is feasible.

Main Methods:

• Theoretical analysis of asymptotic correctness and convergence.
• Investigation of data size requirements based on network parameters.
• Development and application of a numerical approach for shotgun estimation.
• Demonstration using simulated synfire and weakly coupled cortical networks.

Main Results:

• The shotgun approach can infer complete connectivity matrices in large neuronal populations under general conditions.
• Posterior error increases with unobserved population size, average connectivity strength, and observation sparseness.
• Significant increases in neuronal activity data are needed for small observed populations.

Conclusions:

• Shotgun sampling offers a promising method for inferring neuronal connectivity, particularly in large networks.
• Careful consideration of data requirements is necessary, especially for smaller or densely connected networks.
• The developed numerical approach facilitates practical application of shotgun estimation in neuroscience.