NOVA: a novel R-package enabling multi-parameter analysis and visualization of neural activity in MEA recordings

Area of Science:

• Neuroscience
• Computational Biology
• Bioinformatics

Background:

• Multielectrode array (MEA) technology records simultaneous electrical signals from neuronal networks, generating complex datasets.
• Current analysis methods often underutilize MEA data by focusing on limited metrics like firing rate and synchronicity.

Purpose of the Study:

• To develop an accessible computational tool, NOVA (Neural Output Visualization and Analysis), for comprehensive interpretation and visualization of MEA data.
• To address the underutilization of complex MEA datasets in neuroscience research.

Main Methods:

• NOVA integrates dimensionality reduction (principal component analysis), hierarchical clustering with heatmap generation, and temporal trajectory mapping.
• The tool provides a user-friendly pipeline and customizable advanced plotting modules for diverse user needs.
• Validation involved primary cortical neurons during development and pharmacological manipulation.

Main Results:

• NOVA detected subtle neuronal activity shifts missed by conventional methods.
• The analysis revealed network burst duration as a more significant contributor to activity variance than firing rate.
• The tool facilitates unbiased pattern discovery and hypothesis generation from MEA data.

Conclusions:

• NOVA offers a powerful and accessible solution for in-depth MEA data analysis.
• The tool enhances the discovery of meaningful patterns and supports data-driven hypothesis generation in neuroscience.
• This approach moves beyond traditional metrics to provide a more comprehensive understanding of neural network dynamics.