This weeks top spatial transcriptomics paper
🧬 Disease Atlases & Precision Medicine
Spatial atlas of diabetic kidney disease reveals a B cell-rich subgroup.
Using high-resolution spatial transcriptomics, the authors construct a detailed atlas of diabetic kidney disease (DKD) and uncover a molecularly distinct subgroup characterized by dense B-cell infiltration. This B cell–rich endotype links spatial immune cell organization with clinical heterogeneity, offering new angles for patient stratification and targeted therapy design.
Impact: Provides a spatially resolved DKD atlas that reveals an immunologically distinct B cell–rich disease subtype with therapeutic implications.
Dumoulin B et al., https://doi.org/10.1038/s41586-026-10363-4
Spatial transcriptomics atlas of inflammatory bowel disease to guide implementation in research consortiums and clinical trials.
By profiling over three million cells from >100 intestinal sections, this study builds a single-cell–resolved spatial atlas for ulcerative colitis and Crohn’s disease across multiple consortia. Comparative benchmarking of CosMx and Xenium platforms highlights performance differences and validates robust detection of regulatory T cell biology, establishing standards for multi-center clinical IBD studies.
Impact: Delivers a reference spatial IBD atlas and technology benchmark to standardize and de-risk spatial omics use in large translational consortia and clinical trials.
Li Y et al., https://doi.org/10.1038/s41467-026-72482-w
🧠Neurobiology & Sensory Systems
Spatial organization and detection of social odors in mouse primary olfactory system.
Using MERFISH-based spatial transcriptomics, this work maps the full mouse olfactory receptor repertoire across the main olfactory epithelium and bulb, revealing ordered gradients of sensory neuron distribution and projection patterns. Co-imaging receptor and activity markers identifies distinct spatial domains tuned to ethologically relevant social odors, uncovering a topographic logic for odor processing.
Impact: Establishes a molecularly and spatially resolved atlas of the primary olfactory system that links receptor expression, circuit topology, and social odor coding.
Bintu B et al., https://doi.org/10.1016/j.cell.2026.03.053
Retinoic acid drives cell fate specification, maturation and retinal regionality in human retinal organoids.
By temporally and dose-wise modulating retinoic acid during retinal organoid differentiation, the authors dissect how RA levels shape photoreceptor versus interneuron abundance, maturation, and spatial organization. Integrating single-cell and spatial transcriptomics, they show that low RA promotes macula-like patterning while high RA biases toward peripheral retina, clarifying mechanisms of human retinal regionalization.
Impact: Demonstrates RA dosage control as a tunable lever to engineer macula- versus periphery-like retinal organoids for disease modelling and cell therapy.
Lim BY et al., https://doi.org/10.1038/s41467-026-72130-3
🧪 Technology & Methods Development
Interpretable, flexible and spatially aware integration of multiple spatial transcriptomics datasets from diverse sources.
INSPIRE is a deep-learning framework that uses adversarial learning and graph neural networks to integrate heterogeneous spatial transcriptomics datasets while preserving spatial context. Coupled with non-negative matrix factorization, it yields interpretable spatial factors and gene programs, enabling cross-platform integration, 3D reconstruction, and refined analysis of tumor microenvironments and developmental processes at scale.
Impact: Introduces a scalable, spatially aware integration method that unifies multi-platform ST data into interpretable maps of tissue architecture and biology.
Zhao J et al., https://doi.org/10.1038/s41588-026-02579-x
Spatial-aware detection of copy number alterations from spatial transcriptomics using SpaCNA.
SpaCNA integrates neighborhood gene expression, image-derived morphology, and a hidden Markov random field to infer copy number alterations directly from spatial transcriptomics data. It achieves high accuracy in calling CNAs, delineates tumor boundaries, and reconstructs 3D CNA landscapes, revealing spatially distinct subclones and evolutionary trajectories in multiple solid tumors.
Impact: Provides a dedicated tool to extract high-resolution, 3D genomic architecture and clonal structure from spatial transcriptomics cancer datasets.
Zhang Z et al., https://doi.org/10.1038/s41467-026-72284-0
GR2ST: Spatial Transcriptomics Prediction based on Graph-Enhanced Multimodal Contrastive Learning.
GR2ST is a deep learning model that predicts spatial gene expression patterns from histology images by jointly modeling image and spatial transcriptomic data through graph-enhanced contrastive learning. By capturing fine-grained relationships between tissue morphology and transcriptional states, it aims to reduce cost and turnaround time for spatial omics while preserving biological signal.
Impact: Advances morphology-to-transcriptome prediction, enabling more accessible and scalable use of spatial information when full ST sequencing is impractical.
Zhou J et al., https://doi.org/10.1093/bioinformatics/btag209
🧫 Cancer Research & Tumor Microenvironment
Spatial-aware detection of copy number alterations from spatial transcriptomics using SpaCNA.
SpaCNA aggregates expression from morphologically and spatially similar spots and models spatial continuity to robustly infer CNAs from tumor spatial transcriptomics. In breast, colorectal, and head and neck cancers, it sharpens tumor–normal boundaries, identifies spatially segregated subclones, and reconstructs 3D evolutionary trajectories of tumor progression.
Impact: Enhances dissection of intratumoral heterogeneity and clonal evolution directly in situ, strengthening the utility of ST in cancer biology and pathology.
Zhang Z et al., https://doi.org/10.1038/s41467-026-72284-0
Activation of the Integrin αV-YAP-CTGF Axis in Liver Sinusoidal Endothelial Cells Promotes Liver Fibrogenesis, Leading to Portal Hypertension and Liver Carcinogenesis in Congestive Hepatopathy.
Focusing on congestive hepatopathy, this study dissects how liver sinusoidal endothelial cells activate the integrin αV–YAP–CTGF signaling axis to drive fibrosis, portal hypertension, and eventual carcinogenesis. Spatial and molecular profiling of LSECs under chronic congestion uncovers endothelial-centric mechanisms that couple vascular stress to fibrotic remodeling and tumor initiation.
Impact: Identifies an endothelial integrin–YAP signaling axis as a spatially localized driver of fibrogenesis and liver cancer risk in congestive hepatopathy.
Kato S et al., https://doi.org/10.1053/j.gastro.2025.11.014
Acinar Metaplastic Cells Generate Semi-Homogeneous Niches and Interact With Immune Cells.
Building on previous single-cell work, this paper maps the spatial distribution of seven acinar metaplastic cell subtypes in premalignant pancreatic lesions and their crosstalk with stromal and immune populations. The authors show that semi-homogeneous niches of metaplastic cells orchestrate local immune interactions and microenvironmental remodeling that may prime progression to pancreatic ductal adenocarcinoma.
Impact: Provides a spatial blueprint of metaplastic niches and immune interactions that underlie early pancreatic tumorigenesis.
Arcila-Barrera S et al., https://doi.org/10.1053/j.gastro.2025.12.014
🧠🧪 Neuro-Immune & Inflammatory Disease
Spatial transcriptomics atlas of inflammatory bowel disease to guide implementation in research consortiums and clinical trials.
Leveraging high-plex imaging-based ST across multiple centers, this work charts immune, epithelial, and stromal cell states across inflamed and non-inflamed regions in ulcerative colitis and Crohn’s disease. The atlas highlights stable detection of regulatory T cell–associated programs and demonstrates that CosMx maintains data quality across varied pre-analytic conditions important for real-world trials.
Impact: Sets a new reference for spatially resolved immune and epithelial organization in IBD and clarifies platform performance for clinical deployment.
Li Y et al., https://doi.org/10.1038/s41467-026-72482-w
đź’» Future Computing & Single-Cell / Spatial Omics
Advancing single-cell omics and cell-based therapeutics with quantum computing.
This Roadmap article outlines current challenges in spatiotemporal single-cell and spatial multi-omics analysis, including model complexity, data integration, and simulation of cellular dynamics. It proposes how hybrid quantum–classical algorithms could accelerate key computational bottlenecks and presents a case study on integrating quantum computing into the design and optimization of cell-based therapies.
Impact: Frames a forward-looking agenda for combining quantum computing with high-resolution single-cell and spatial omics to enable next-generation precision therapeutics.
Bose A et al., https://doi.org/10.1038/s41580-025-00918-0