EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

ASAOT: an accentuated structure-differentiated asymmetric optimal transport model for cell type and single-cell spatial mapping

Ziyan Sun, Zhixin Shi, Yuan Zhang, Mengdi Nan, Jie Gao iD

Briefings in Bioinformatics · 02.11.2026 · 27 (6)

DOI: 10.1093/bib/bbag615 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Briefings in Bioinformatics, online erschienen am 13.11.2026. Verfasst von Ziyan Sun und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Mathematics and Data Science, Jiangnan University, 1800 Lihu Avenue, Binhu District, Wuxi, Jiangsu 214122, China. Hauptthemen laut MEDLINE: Single-Cell Analysis. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Spatial transcriptomics (STs) characterizes spatiotemporal gene expression and facilitates analysis of cellular architecture and cell–cell crosstalk within tissue microenvironments. However, existing spatial deconvolution methods employ globally fixed smoothing without local sharpening, which blurs cellular boundaries and impedes reconstruction of layered, heterogeneous cellular patterns in brain tissue. Here, we present Accentuated Structure-differentiated Asymmetric Optimal Transport (ASAOT), a deconvolution framework built on structure-enhanced asymmetric optimal transport. Coupled with spatial neighborhood constraints, ASAOT introduces a differentiated accentuated sharpening term to maintain global spatial continuity, enhance boundary detection, and preserve spot-level expression heterogeneity. Using asymmetric optimal transport, the framework jointly estimates spot-wise cell-type proportions and performs spatial assignment of reference single-cell coordinates. Benchmarks on diverse simulated datasets demonstrate that ASAOT achieves competitive overall accuracy and robustness across moderate spatial resolutions, with the best average comprehensive rank among tested methods. Applications to mouse olfactory bulb and cerebral cortex datasets demonstrate that ASAOT reduces signal diffusion and boundary distortion, effectively recovering layered brain anatomy and stratified cell distributions to build interpretable spatial maps of reference single-cell profiles. Collectively, ASAOT addresses critical drawbacks of conventional methods via a sharpening regularization term, serving as a reliable tool for STs analysis and mechanism exploration of tissue spatial organization.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Single-Cell AnalysisAnimalsMiceSpatial TranscriptomicsAlgorithmsBrainOlfactory BulbSingle-Cell Gene Expression Analysis

Stichwörter der Autor:innen

cell-type proportionoptimal transportsingle-cell spatial mappingspatial deconvolutionspatial transcriptomics

Autor:innen und Institutionen

  • School of Mathematics and Data Science, Jiangnan University, 1800 Lihu Avenue, Binhu District, Wuxi, Jiangsu 214122, China.

ORCID-Profile: Jie Gao

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Sun, Z., Shi, Z., Zhang, Y., Nan, M., & Gao, J. (2026). ASAOT: an accentuated structure-differentiated asymmetric optimal transport model for cell type and single-cell spatial mapping. Briefings in Bioinformatics, 27(6). https://doi.org/10.1093/bib/bbag615

Vancouver

Sun Z, Shi Z, Zhang Y, Nan M, Gao J. ASAOT: an accentuated structure-differentiated asymmetric optimal transport model for cell type and single-cell spatial mapping. Briefings in Bioinformatics. 2026;27(6). doi:10.1093/bib/bbag615

BibTeX
@article{Sun2026,
  author = {Sun, Ziyan and Shi, Zhixin and Zhang, Yuan and Nan, Mengdi and Gao, Jie},
  title = {ASAOT: an accentuated structure-differentiated asymmetric optimal transport model for cell type and single-cell spatial mapping},
  journal = {Briefings in Bioinformatics},
  year = {2026},
  volume = {27},
  number = {6},
  doi = {10.1093/bib/bbag615}
}
Kostenlos starten →