Skip to content
DraconiticResearch
Search Sources API
197,768 studies31 sourcesUpdated 43 min ago
Peer-reviewedCC BYPLOS Genetics

Enhancing cell type deconvolution in spatial transcriptomics through incorporating multiple reference panels

  • Peiyao Zhao1,2
  • Justin A Colacino3,4
  • Ying Ma5,6
  • Xiang Zhou7
7 affiliations
  1. 1Department of Biostatistics, School of Public Health, University of Michigan, Ann Arbor, Michigan, United States of America.
  2. 2Center for Statistical Genetics, University of Michigan, Ann Arbor, Michigan, United States of America.
  3. 3Department of Environmental Health Sciences, University of Michigan School of Public Health, Ann Arbor, Michigan, United States of America.
  4. 4Department of Nutritional Sciences, University of Michigan School of Public Health, Ann Arbor, Michigan, United States of America.
  5. 5Department of Biostatistics, School of Public Health, Brown University, Providence, Rhode Island, United States of America.
  6. 6Center for Computational Molecular Biology, Brown University, Providence, Rhode Island, United States of America.
  7. 7Department of Statistics and Data Science, Yale University, New Haven, Connecticut, United States of America.
Read full textPDFDOICite

Abstract

Spatial transcriptomics (ST) captures gene expression with spatial context but often at resolutions where each spatial location contains multiple cells. Deconvolution of cell-type proportions is therefore essential, yet existing methods typically rely on a single scRNA-seq reference, limiting accuracy when references are incomplete or mismatched especially in complex tissues. These limitations highlight the need for a framework that can integrate multiple references while leveraging spatial information. We introduce SWIMR (Spatially Weighted Integration for Multi-Reference deconvolution), a statistical framework that integrates multiple scRNA-seq references while accounting for spatial correlations. SWIMR models each spatial location as a weighted combination of references, with both cell-type proportions and reference weights varying across tissue locations. Across comprehensive simulations and diverse ST datasets, SWIMR consistently outperformed single-reference approaches. In developing human heart tissue, it recovered ground-truth cell-type compositions with the highest accuracy. In human lung and breast cancer, SWIMR delineated tumor, stromal, and immune compartments, uncovered tumor-adjacent populations, and detected cell type–specific spatially variable genes missed by other methods. In mouse embryo and liver cancer, it leveraged context-specific references to resolve organ-specific cell types and increase sensitivity for biologically relevant pathways. Together, these results establish SWIMR as a robust framework for multi-reference deconvolution in spatial transcriptomics. Author summary: Spatial transcriptomics measures gene expression at many locations across tissues, revolutionizing genomics. However, many widely used spatial technologies have limited resolution: each measured location may contain multiple cells, often from different cell types. Researchers therefore commonly use computational deconvolution methods to estimate cell-type composition at each location, typically using single-cell RNA-seq reference data. However, it is often difficult to find a single-cell reference that closely matches a spatial experiment in tissue type, developmental stage, disease state, or other characteristics. The growing availability of public single-cell datasets provides an opportunity to combine information from multiple related references. This, however, creates a new challenge: how to decide which reference, or combination of references, to use, as this choice can substantially affect deconvolution results.We developed SWIMR, a computational method that integrates multiple single-cell reference datasets for spatial transcriptomics deconvolution. SWIMR automatically learns the contribution of each reference and allows these contributions to vary across tissue locations, while leveraging spatial organization to stabilize cell-type estimates. Across simulations and datasets, SWIMR effectively integrates complementary reference information and produces biologically meaningful maps of cell-type distributions. SWIMR provides a flexible approach for leveraging the growing diversity of single-cell reference data for spatial deconvolution.

Related

  1. WDGTA: a wavelet-denoised graph attention framework for spatial transcriptomics deconvolution.
    Hao Zhu, Chenggong Ma, Zhenchun Ran +5
    Briefings in bioinformaticsPeer-reviewedCC BYFull textMolecular biology & gene regulationEurope PMC
  2. Systematic evaluation of robustness of deconvolution methods for spatial transcriptomics data in case of cell type mismatch.
    Utkarsh M Mahamune, Aldo Jongejan, Antoine H C van Kampen +2
    Briefings in bioinformaticsPeer-reviewedCC BYFull textMolecular biology & gene regulationEurope PMCPMC
  3. Enhancing the accuracy of qPCR normalization in Bactrocera zonata: A systematic study for the selection of condition-optimized reference genes for gene expression studies
    T. C Suman, Selva Babu Selvamani, N. Veeramanikanta Reddy +5
    Scientific ReportsPeer-reviewedCC BY-NC-ND 4.0Full textMolecular biology & gene regulationCrossrefDOAJ
  4. Enhancing double-duplex invasion by using multiple parallel-stranded peptide nucleic acids
    Masanari Shibata, Kota Ito, Osami Shoji +1
    ChemRxivPreprintCC BYFull textMolecular biology & gene regulationChemRxiv
  5. Single-cell Transcriptomics in Autoimmune Diseases: Mechanistic Insights, Cross-disease Convergence, and Translational Challenges in Rheumatoid Arthritis, Systemic Lupus Erythematosus, Type 1 Diabetes, and Multiple Sclerosis.
    Thanh-Hong Nguyen, Tuan-Dinh Nguyen, Oanh-Thi Hoang +5
    Clinical reviews in allergy & immunologyPeer-reviewedCC BYFull textMolecular biology & gene regulationEurope PMC

Cite

BibTeXRIS
BibTeX
@article{zhao2026enhancing,
  title = {{Enhancing cell type deconvolution in spatial transcriptomics through incorporating multiple reference panels}},
  author = {Peiyao Zhao and Justin A Colacino and Ying Ma and Xiang Zhou},
  journal = {PLOS Genetics},
  year = {2026},
  month = nov,
  doi = {10.1371/journal.pgen.1012345},
  url = {https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1012345},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - Enhancing cell type deconvolution in spatial transcriptomics through incorporating multiple reference panels
AU  - Peiyao Zhao
AU  - Justin A Colacino
AU  - Ying Ma
AU  - Xiang Zhou
PY  - 2026
DA  - 2026/11/01
JO  - PLOS Genetics
DO  - 10.1371/journal.pgen.1012345
UR  - https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1012345
AB  - Spatial transcriptomics (ST) captures gene expression with spatial context but often at resolutions where each spatial location contains multiple cells. Deconvolution of cell-type proportions is therefore essential, yet existing methods typically rely on a single scRNA-seq reference, limiting accuracy when references are incomplete or mismatched especially in complex tissues. These limitations highlight the need for a framework that can integrate multiple references while leveraging spatial information. We introduce SWIMR (Spatially Weighted Integration for Multi-Reference deconvolution), a statistical framework that integrates multiple scRNA-seq references while accounting for spatial correlations. SWIMR models each spatial location as a weighted combination of references, with both cell-type proportions and reference weights varying across tissue locations. Across comprehensive simulations and diverse ST datasets, SWIMR consistently outperformed single-reference approaches. In developing human heart tissue, it recovered ground-truth cell-type compositions with the highest accuracy. In human lung and breast cancer, SWIMR delineated tumor, stromal, and immune compartments, uncovered tumor-adjacent populations, and detected cell type–specific spatially variable genes missed by other methods. In mouse embryo and liver cancer, it leveraged context-specific references to resolve organ-specific cell types and increase sensitivity for biologically relevant pathways. Together, these results establish SWIMR as a robust framework for multi-reference deconvolution in spatial transcriptomics. Author summary: Spatial transcriptomics measures gene expression at many locations across tissues, revolutionizing genomics. However, many widely used spatial technologies have limited resolution: each measured location may contain multiple cells, often from different cell types. Researchers therefore commonly use computational deconvolution methods to estimate cell-type composition at each location, typically using single-cell RNA-seq reference data. However, it is often difficult to find a single-cell reference that closely matches a spatial experiment in tissue type, developmental stage, disease state, or other characteristics. The growing availability of public single-cell datasets provides an opportunity to combine information from multiple related references. This, however, creates a new challenge: how to decide which reference, or combination of references, to use, as this choice can substantially affect deconvolution results.We developed SWIMR, a computational method that integrates multiple single-cell reference datasets for spatial transcriptomics deconvolution. SWIMR automatically learns the contribution of each reference and allows these contributions to vary across tissue locations, while leveraging spatial organization to stabilize cell-type estimates. Across simulations and datasets, SWIMR effectively integrates complementary reference information and produces biologically meaningful maps of cell-type distributions. SWIMR provides a flexible approach for leveraging the growing diversity of single-cell reference data for spatial deconvolution.
SN  - 1553-7404
SN  - 1553-7390
ER  -