Classifying Neuronal Cell Types Based on Shared Electrophysiological Information from Humans and Mice.

Area of Science:

• Neuroscience
• Computational Biology
• Machine Learning

Background:

• Accurate neuron classification is essential for understanding brain function and neurological diseases.
• Electrophysiological activity is a key characteristic for differentiating neuron types.
• Machine learning offers powerful tools for analyzing complex biological data.

Purpose of the Study:

• To develop a deep-learning framework for classifying neurons based exclusively on electrophysiological data.
• To integrate and classify neuron data from both mouse and human sources.
• To classify mouse neuron subtypes using an interpretable model.

Main Methods:

• Utilized the Allen Cell Types database containing single-cell recordings from mice and humans.
• Developed a joint domain-adaptive model integrating electrophysiological data from both species.
• Employed an interpretable neural network for classifying mouse neuron subtypes based on transgenic line labels.

Main Results:

• Achieved state-of-the-art accuracy and precision in neuron classification.
• Successfully integrated cross-species electrophysiological data for broad neuron type classification.
• Enabled subtype classification for mouse neurons with interpretable predictions.

Conclusions:

• The deep-learning framework provides a robust and accurate method for neuron classification using electrophysiological data.
• This approach advances the understanding of neural diversity and function in health and disease.
• The interpretable nature of the model aids in biological insight and validation.