Transcriptomic diversity of cell types across the adult human brain
Siletti K, Hodge R, Mossi Albiach A, et al. Science. 2023;382.
doi:10.1126/science.add7046About DigitalBrain
A large-scale human brain atlas and foundation-model project integrating molecular and cellular information into a unified, AI-ready framework.
The atlas harmonizes single-cell and single-nucleus transcriptomic datasets generated by research groups worldwide. We systematically collected, quality-controlled, standardized and integrated these datasets, unified anatomical annotations across studies, organized brain regions within a common hierarchical framework, and harmonized cell-type and donor-level metadata.
This enables genes, cell types, brain regions, ages and disease conditions to be explored and compared within a consistent reference system.
The DigitalBrain Data Explorer provides an interactive interface for querying gene expression across human brain regions and cell types, visualizing anatomical distributions, and supporting downstream computational and AI-assisted analysis.
What DigitalBrain adds
A common analytical layer across previously heterogeneous datasets.
Data curation & metadata
Systematic collection and quality control of human brain single-cell datasets; harmonization of dataset, donor and cell metadata.
Anatomical harmonization
Standardized anatomical annotation across studies; organization of brain regions within a unified hierarchical brain atlas.
Expression & integration
Harmonized gene-expression representations across regions and cell types; integration into a common framework for large-scale computational analysis and AI modeling.
Our goal is to transform fragmented human brain datasets into a unified resource that can be searched, compared and computationally modeled across biological scales.
Cite DigitalBrain
If you use DigitalBrain, the DigitalBrain Data Explorer, or data processed through this resource in your research, please cite:
Multiscale transcriptomic organization of the human brain with DigitalBrain
bioRxiv (2026). doi:10.64898/2026.04.14.718492.
When using data derived from individual datasets, please also cite the relevant primary publications.
Data sources & acknowledgements
DigitalBrain builds upon the extraordinary collective efforts of the global neuroscience community. The underlying datasets were generated by numerous laboratories and consortia worldwide through years of human tissue collection, single-cell and single-nucleus sequencing, data curation and biological analysis.
Adult human whole-brain atlas
The transcriptomic atlas reported by Siletti et al. provided broad cellular coverage across the adult human brain.
Human hippocampal lifespan atlas
The atlas generated by Su, Zhou et al. characterized glial diversity and age-related molecular changes across the postnatal human hippocampus.
These studies, together with many other publicly shared datasets, form the experimental foundation upon which DigitalBrain was constructed.
We are deeply grateful to the investigators who generated and openly shared these datasets, as well as to the research participants, tissue donors and their families whose contributions made this work possible.
DigitalBrain does not replace or supersede the original studies.
Users are encouraged to cite the relevant primary publications when using data derived from individual datasets. For the full list of source datasets and primary publications, please refer to the DigitalBrain manuscript.
Representative references
Selected source studies. The complete list is provided in the DigitalBrain manuscript.
A single-cell transcriptome atlas of glial diversity in the human hippocampus across the postnatal lifespan
Su Y, Zhou Y, Bennett ML, et al. Cell Stem Cell. 2022;29:1594-1610.e8.
doi:10.1016/j.stem.2022.09.010