Cell Type Proportion Inference Over Human Neurodevelopment

Cell distribution is essential for the organization, functionality, and structural integrity of the human brain. The spatial arrangement and differentiation of cells define cytoarchitecture, forming the foundation for neural circuits that support motor and cognitive functions throughout neurodevelopment. Single-nucleus RNA sequencing enables precise quantification of cell proportions in brain samples, revealing how cellular composition influences regional specialization and functionality in the mature brain. However, data on cell-type distribution during neurodevelopment remains limited due to challenges in sample accessibility. To address this gap, we use a machine-learning-based deconvolution tool called CIBERSORTx, to infer cell-type proportions from 15 distinct cell types across multiple brain regions from the BrainSpan dataset. Our analysis captured developmental trends, showing dynamic shifts in cell composition across neurodevelopment without requiring physical cell isolation. These findings provide an enhanced-resolution landscape of brain cell maturation, contributing to a deeper understanding of how cellular populations evolve throughout the human lifespan.