Information flow drives localized morphological differences across neuronal and glial cell types.

Area of Science:

• Neuroscience
• Biophysics
• Computational Biology

Background:

• Neurons and glia exhibit distinct, functionally relevant branching patterns.
• Previous work linked neuronal branching to biophysical constraints using mathematical models.

Purpose of the Study:

• To apply machine learning to neuron and glia morphology using parameters derived from a biophysical model.
• To differentiate between cell types and identify pathological changes based on structural features.

Main Methods:

• Extracted functionally relevant structural parameters from a unifying mathematical model of branching.
• Utilized machine-learning classification with features including asymmetric branching and relative junction location.
• Analyzed how information flow parameters vary with cell position (soma/synapse proximity).

Main Results:

• Machine learning models effectively distinguished between different neuron and glia types.
• Incorporating relative branching junction location significantly improved classification accuracy for specific cell types (e.g., medium spiny neuron dendrites).
• Information flow parameters correlate with cell position, indicating localized functional adaptations.

Conclusions:

• Morphological differences, particularly related to information flow and branching location, are key discriminators of cell types and states.
• The developed methods offer a foundation for classifying neuronal and glial cells based on pathology.
• Asymmetric scale factors and relative branching junction location show potential as biomarkers for disease identification.