Dimension reduction is a cornerstone of exploratory data analysis; however, traditional methods fail to preserve the spatial context of spatial genomics data. In this work, we develop a nonnegative spatial factorization (NSF) model that allows interpretable, parts-based decomposition of spatial single-cell count data. NSF allows label-free annotation of regions of interest in spatial genomics data and identifies genes and cells that can be used to define those regions.
This is a preview of subscription content, access via your institution
Access options
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$32.99 / 30 days
cancel any time
Subscribe to this journal
Receive 12 print issues and online access
$259.00 per year
only $21.58 per issue
Buy this article
- Purchase on SpringerLink
- Instant access to the full article PDF.
USD 39.95
Prices may be subject to local taxes which are calculated during checkout
References
Lee, D. D. & Seung, H. S. Learning the parts of objects by non-negative matrix factorization. Nature 401, 788–791 (1999). Original nonnegative matrix factorization paper that introduces the concept of an interpretable parts-based representation.
Velten, B. et al. Identifying temporal and spatial patterns of variation from multimodal data using MEFISTO. Nat. Methods 19, 179–186 (2022). This paper showed how to combine real-valued dimension reduction with spatial correlation using Gaussian processes.
Wu, L. et al. Variational nearest neighbor Gaussian process. Preprint at arXiv https://doi.org/10.48550/arXiv.2202.01694 (2022). This study combines variational inference for nonconjugate likelihoods (also used by us) with nearest-neighbor approximations to enable greater scalability to large numbers of observations.
Cable, D. M. et al. Robust decomposition of cell type mixtures in spatial transcriptomics. Nat Biotechnol. 40, 517–526 (2022). This study uses single-cell reference data to assign cell types to spatial transcriptomics data and then examines spatial variation within each cell type.
Li, D. et al. Multi-group Gaussian processes. Preprint at arXiv https://doi.org/10.48550/arXiv.2110.08411 (2021). This study extends Gaussian processes to include both spatial locations and categorical labels such as cell or tissue type.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This is a summary of: Townes, F. W. & Engelhardt, B. E. Nonnegative spatial factorization applied to spatial genomics. Nat. Methods https://doi.org/10.1038/s41592-022-01687-w (2022)
Rights and permissions
About this article
Cite this article
Parts-based decomposition of spatial genomics data finds distinct tissue regions. Nat Methods 20, 187–188 (2023). https://doi.org/10.1038/s41592-022-01725-7
Published:
Version of record:
Issue date:
DOI: https://doi.org/10.1038/s41592-022-01725-7
