Human Brain Gene Activity Mirrors Neural Wiring in Cross-Species Model
Comparing human gene expression with mouse neural connections reveals that protein interaction networks and genomic gradients track brain wiring.
A cross-species brain model linking human genetic expression with mouse neural wiring reveals that gene activity profiles mirror the architecture of mammalian neural connections. Researchers constructed a cross-species digital xenocortication framework to test whether brain regions in humans and mice share a conserved connectional plan. By pairing human gene expression measurements from the Allen Human Brain Atlas with mouse projection strengths across 73 to 92 anatomical nodes, the authors report significant statistical associations between regional transcriptomes and neural connectivity.
The study, available on the arXiv preprint server, investigated how endogenous multiprotein complexes and transposable elements shape the molecular landscape of mammalian brains. Endogenous multiprotein complexes serve as fundamental structural building blocks in all living cells, but cataloging them typically ignores cell-type specificity and intracellular protein availability. To overcome this limitation, the investigator examined hybrid assembly pathways, where specific DNA sequences act as templates for protein interactions.
How Transposable Elements Shape Neural Connections
Transposable elements are repetitive DNA sequences that make up roughly half of the human genome. While many are silenced in differentiated cells, certain families remain active and provide regulatory sequences like promoters and enhancers. The author previously showed that these elements can seed transcription factor binding sites and initiate protein interaction cascades that generate cell-type-specific pools of protein dimers.
In the cross-species model, the analysis tracked four families of protein interaction cascade constituents: principal initiator transcription factors, complementary initiators, protein-protein interaction branch targets, and hub proteins. Per-node availability analyses demonstrated that these components exhibit ubiquitous expression and universal availability across brain regions. Notably, hub proteins linked to synaptic integrity displayed consistent expression across all evaluated human brain nodes, supporting their role in maintaining synapses and post-synaptic densities.
What Variance Decomposition Reveals About Brain Organization
To determine how different factors influence brain connectivity, the author performed variance decomposition across multiple model resolutions. The results indicate that a steep rostrocaudal, meaning front-to-back, gradient and hybrid assembly pathway constituency act as major unique contributors to regional connectome associations. In human models, unique components accounted for over 54 percent of the explained variance in certain parameter configurations, whereas mouse models showed a higher proportion of shared variance.
When evaluating protein complexes catalogued in the CORUM database, the fraction of engaged subunits tracked interregional connectivity even more closely than raw marker panels. Complexes involved in synaptic signaling and structural maintenance showed robust associations with how strongly brain nodes connected to one another. These associations remained statistically significant even after controlling for spatial distance between brain regions.
Limitations and Future Directions
While the findings provide new insights into the molecular correlates of brain wiring, several limitations should be noted. The model relies on anatomical correspondences between human and mouse brains, and certain brain regions lacked sufficient data coverage in one or both species. Furthermore, gene expression measurements derived from in situ hybridization in mice differ methodologically from human microarray and RNA sequencing datasets.
The author emphasizes that these models serve as descriptive frameworks rather than definitive proofs of species-specific regulatory differences. Future research will require direct experimental validation of protein dimer repertoires and finer spatial mapping to fully elucidate how genomic regulatory networks govern structural and functional features of the mammalian brain connectome.
The study was posted to the arXiv preprint server on November 15, 2026.
This article has been fact checked for accuracy, with information verified against reputable sources. Learn more about us and our editorial process.
Last reviewed on .
Article history
- Latest version
Reference(s)
- Glinsky, Gennadi. “Cross-species digital xenocortication model delineates molecular correlates of human brain connectome.” arXiv.org <http://arxiv.org/abs/2611.19713v1>.
Cite this page:
- Posted by Hassan Raza