ANTIPODE Provides a Global View of Cell Type Homology and Transcriptomic Divergence in the Developing Mammalian Brain

Area of Science:

• Neuroscience
• Developmental Biology
• Computational Biology

Background:

• Mammalian brain development involves conserved sequences of neuron and glia generation.
• Gene regulatory network divergence impacts brain composition, scaling, timing, and function.
• Resolving gene regulatory divergence necessitates cellular-resolution, cross-species data and robust homology identification.

Purpose of the Study:

• To introduce ANTIPODE, a deep-learning framework for integrating single-cell data across species.
• To identify homologous cell types and analyze differential gene expression.
• To uncover principles of gene regulatory divergence in mammalian brain evolution.

Main Methods:

• Developed ANTIPODE, a deep-learning variational inference framework.
• Applied ANTIPODE to macaque whole brain development data.
• Integrated human, macaque, and mouse brain development data into a meta-atlas.

Main Results:

• ANTIPODE successfully integrates diverse single-cell datasets and identifies homologous cell types.
• Found broad conservation of early neuron classes across species.
• Revealed widespread gene regulatory divergence within homologous cell types, influenced by genomic context, lineage, and timing.

Conclusions:

• ANTIPODE offers a formalized, interpretable framework for cross-species single-cell analysis.
• Demonstrated principles of gene regulatory divergence shaping mammalian brain evolution.
• Highlighted the interplay of conserved structures and regulatory plasticity in brain development.