Creating a Cross-Species Spinal Cord Taxonomy

Why It Matters

A unified spinal‑cord cell‑type map bridges model‑animal research and human neuroscience, accelerating disease modeling and precision therapeutics development.

Key Takeaways

• Cross-species spinal cord atlas integrates human, macaque, mouse data.
• Alignment algorithms remove species-specific batch effects before clustering.
• Multiomics profiling combines transcriptomics, epigenomics, spatial data for insight.
• Hierarchical taxonomy built from broad neuronal classes to fine subtypes.
• AI and annotation transfer accelerate cell-type labeling and discovery.

Summary

The webinar presented a new Allen Institute effort to build a cross‑species spinal cord taxonomy, leveraging multi‑omics data from human donors, macaques and mice. By generating 10x multi‑ome profiles, spatial transcriptomics and epigenetic maps, the team assembled a comprehensive cellular atlas that spans dorsal and ventral spinal cord regions. Key technical steps include removing species‑specific batch effects with alignment tools such as scVI, followed by robust clustering using the Scratch‑Hi algorithm. Annotation transfer from existing mouse and human lumbar studies provides an initial label set, which is refined with marker‑gene analysis and AI‑assisted literature mining to resolve fine‑grained subtypes, especially among motor neurons. The presenters highlighted that, despite the spinal cord’s conserved dorsal‑ventral architecture across vertebrates, cellular complexity differs markedly—mice exhibit many more interneuron subtypes than zebrafish. They also discussed quality‑control challenges like doublets and low‑read nuclei, and demonstrated how spatial mapping links transcriptomic clusters to precise anatomical locations. The resulting hierarchical taxonomy—from broad glutamatergic, GABAergic, cholinergic classes down to specific dorsal‑mid and ventral motor neuron subpopulations—offers a reference framework for translating discoveries in model organisms to human biology and for designing targeted viral tools.

Original Description

This webinar will focus on the creation of a new cross-species spinal cord taxonomy, featuring transcriptomic data from human, macaque, and mouse.

To learn more about the project: https://www.biorxiv.org/content/10.64898/2026.02.04.703852v1

To view the data: https://celltype.info/project/609