PuSH - Publication Server of Helmholtz Zentrum München

Export: TXT Text RIS Endnote (RIS) BIB BibTeX
Tejada Lapuerta, A. ; Schaar, A. ; Gutgesell, R.M. ; Palla, G. ; Halle, L. ; Minaeva, M. ; Vornholz, L. ; Dony, L. ; Drummer, F. ; Richter, T. ; Bahrami, M. ; Theis, F.J.

Nicheformer: A foundation model for single-cell and spatial omics.

Nat. Methods, DOI: 10.1038/s41592-025-02814-z (2025)
Publ. Version/Full Text Research data DOI PMC
Open Access Hybrid
Creative Commons Lizenzvertrag
Tissue makeup depends on the local cellular microenvironment. Spatial single-cell genomics enables scalable and unbiased interrogation of these interactions. Here we introduce Nicheformer, a transformer-based foundation model trained on both human and mouse dissociated single-cell and targeted spatial transcriptomics data. Pretrained on SpatialCorpus-110M, a curated collection of over 57 million dissociated and 53 million spatially resolved cells across 73 tissues on cellular reconstruction, Nicheformer learns cell representations that capture spatial context. It excels in linear-probing and fine-tuning scenarios for a newly designed set of downstream tasks, in particular spatial composition prediction and spatial label prediction. Critically, we show that models trained only on dissociated data fail to recover the complexity of spatial microenvironments, underscoring the need for multiscale integration. Nicheformer enables the prediction of the spatial context of dissociated cells, allowing the transfer of rich spatial information to scRNA-seq datasets. Overall, Nicheformer sets the stage for the next generation of machine-learning models in spatial single-cell analysis.
Impact Factor
Scopus SNIP
Altmetric
32.100
0.000
Tags
Annotations
Special Publikation
Hide on homepage

Edit extra information
Edit own tags
Private
Edit own annotation
Private
Hide on publication lists
on hompage
Mark as special
publikation
Publication type Article: Journal article
Document type Scientific Article
Language english
Publication Year 2025
HGF-reported in Year 2025
ISSN (print) / ISBN 1548-7091
e-ISSN 1548-7105
Journal Nature Methods
Publisher Nature Publishing Group
Publishing Place New York, NY
Reviewing status Peer reviewed
Institute(s) Institute of Computational Biology (ICB)
Institute of Diabetes and Obesity (IDO)
POF-Topic(s) 30205 - Bioengineering and Digital Health
30201 - Metabolic Health
Research field(s) Enabling and Novel Technologies
Helmholtz Diabetes Center
PSP Element(s) G-503800-001
G-502200-001
DOI 10.1038/s41592-025-02814-z
PubMed ID 41168487
Erfassungsdatum 2025-11-03
[➜Report correction]