Unraveling Tissue Complexity Through Single-Cell and Spatial Transcriptomics

Abstract

We are pleased to present this special section of Small Methods, which highlights the rapidly advancing fields of single-cell and spatial transcriptomics.

Single-cell transcriptomics and spatial transcriptomics have emerged as transformative tools for high-resolution profiling of gene expression. Single-cell approaches reveal cellular diversity at unprecedented resolution, while spatial transcriptomics preserves the spatial context of gene activity, enabling precise mapping of tissue architecture. Together, these technologies provide complementary insights into biological systems, uncovering cellular heterogeneity, dynamic interactions, and spatially driven molecular processes across diverse fields including developmental biology, cancer biology, immunology, and neuroscience. This collection includes 3 reviews and 10 research articles from prominent scientists with their valuable insights.

In this special section, we explore the transformative potential of spatial transcriptomics data analysis. Gao et al. (smtd.2401451) introduce PASSAGE, a deep learning framework for identifying phenotype-associated signatures across heterogeneous spatial slices. Yang et al. (smtd.2400975) propose a novel cell segmentation method UCS, optimized for large-scale subcellular spatial transcriptomics data. Ishaque et al. (smtd.2401123) present Sainsc, a cell-segmentation-free approach for transcriptome-wide, nanoscale-resolution spatial data. Yuan et al. (smtd.2401199) present a kernel-based strategy to model spatially continuous variations of the tissue microenvironment using a new kernel-based strategyFei et al. (smtd.2401056) introduce a membrane-based method that greatly increases the number of genes captured in cells compared to the number captured using nucleus-based methods. These innovations highlight the power of spatially resolved transcriptomics in decoding tissue complexity.

Transitioning to integrative analysis of single-cell transcriptomics, spatial transcriptomics data, and other data, this field is reshaping our understanding of tissue architecture and gene expression. Liu et al. (smtd.2401145) developed QR-SIDE, a computational framework that maps spatial heterogeneity and optimizes marker gene contributions for robust deconvolution. Chen et al. (smtd.2401163) developed SpaDA, a spatially aware domain adaptation method integrating transcriptomics, histology, and spatial data to resolve cell-type distributions. These tools exemplify the synergy of multi-modal data in advancing oncology, neuroscience, and precision medicine. Last, Su et al. (smtd.2400991) introduce scPDS, a transformer-based deep learning method that predicts drug sensitivities from scRNA-seq data via pathway activation mapping, bridging transcriptomics, and therapeutic development.

The true power of these technologies lies in their applications. Hicks et al. (smtd.2401194) first review the challenges and opportunities of integrating spatially-resolved transcriptomics data across tissues and individuals. Yu et al. (smtd.2401107) review cutting-edge spatial transcriptomics (ST) technologies, computational tools, and their neuroscientific applications. Zeng et al. (smtd.2401171) review the advancements and strategies of spatial omics technologies, summarize their applications in biomedical research, and highlight the power of spatial omics technologies in advancing the understanding of life sciences related to development and disease. Then, Peng et al. (smtd.2401272) developed STExplore, a comprehensive spatial transcriptomics data analysis platform for spatial data integrative analysis. Finally, Ye et al. (smtd.2401192) combine single-cell and spatial transcriptomics to reveal interactions between POSTN+ cancer-associated fibroblasts (CAFs) and CDK16+ tumor cells, driving chemotherapy resistance. These studies exemplify how the integration of these technologies can drive breakthroughs in understanding complex biological systems.

Overall, this special section highlights innovative frameworks that integrate single-cell and spatial transcriptomics to map tissue architecture, decode cellular interactions, and uncover disease mechanisms. Advanced machine learning methods, scalable platforms, and multi-omics integration strategies are pushing the boundaries of biomedical research, offering new insights into tumor microenvironments, neural circuits, and immune dynamics.

Finally, we extend our heartfelt thanks to the authors, reviewers, and editorial team for their invaluable contributions to this special section. It is our hope that this collection of work will inspire further innovation and collaboration, driving the field forward and unlocking new frontiers in biological research.

The authors declare no conflict of interest.

S.S., L.S., and C.Y. wrote the editorial. All authors provided a critical review of the editorial and approved the final editorial before submission.