Spatial Transcriptomics in Cancer Research: Applications and Case Studies

Spatial transcriptomics is a set of molecular profiling technologies that measure gene expression directly within intact tissue sections, preserving the physical coordinates of each measurement. For cancer researchers, this means tumor heterogeneity, microenvironment organization, and cell-cell communication can be studied in their native architectural context instead of in dissociated cell suspensions. This article provides a practical review of how spatial transcriptomics is applied across cancer types, with emphasis on study design decisions, data integration workflows, quality controls, and interpretation limits. The content is directed at students, researchers, analysts, and life-science professionals who need concrete guidance on when and how to use these technologies.

What Spatial Transcriptomics Measures and Why It Matters in Cancer

Traditional bulk RNA sequencing averages gene expression across millions of cells, obscuring the fact that a tumor is a mosaic of malignant, immune, stromal, and vascular populations. Single-cell RNA sequencing (scRNA-seq) resolves this cellular heterogeneity but requires tissue dissociation, which destroys the spatial relationships between cells. Spatial transcriptomics fills this gap by capturing transcript abundance at defined spatial coordinates within an intact tissue section, allowing researchers to ask where specific cell states and signaling programs are located relative to tumor margins, blood vessels, necrotic regions, and immune infiltrates.

The core value proposition in cancer research is architectural. Tumors are not uniform masses, they contain niches where hypoxia, inflammation, immune suppression, and drug resistance develop in spatially restricted patterns. Spatial transcriptomics reveals these patterns directly. For example, in glioblastoma, spatial analysis has shown that tumor cell states are not randomly distributed. Neural progenitor-like and oligodendrocyte progenitor-like tumor cells segregate in distinct regions, and perinecrotic areas are markedly more immunosuppressive than perivascular zones, which are more pro-inflammatory [22]. This kind of positional information cannot be recovered from dissociated cells.

A second major value is the ability to map cell-cell interactions in situ. By integrating spatial transcriptomic data with ligand-receptor analysis, researchers can infer which cells are communicating with which neighbors and whether those interactions are spatially restricted. In colorectal cancer, for instance, spatial analysis has identified immune-excluded zones characterized by cancer-associated fibroblast (CAF)-mediated extracellular matrix barriers and SPP1-positive tumor-associated macrophages, alongside immune-active niches featuring T cell-myeloid cooperation [11]. These spatially defined interaction networks are candidate therapeutic targets.

A third value is translational. Spatial transcriptomic data can be used to derive biomarkers that stratify patients by survival or treatment response. In small cell lung cancer, a spatial analysis of primary tumors and lymph node metastases identified a pan-immune hotspot whose abundance independently predicted survival [9]. In breast cancer, spatial transcriptomic data have been used to define tumor microenvironment subtypes with distinct survival outcomes and to predict response to chemotherapy and trastuzumab [5]. These findings illustrate the trajectory from descriptive mapping to clinically actionable readouts.

Core Principles of Spatial Transcriptomic Study Design

Choosing a Platform Based on Resolution and Gene Coverage

Spatial transcriptomic platforms differ substantially in spatial resolution, gene throughput, and tissue compatibility. The two most widely used categories are array-based approaches, such as 10x Visium, and imaging-based approaches, such as the CosMx Spatial Molecular Imager. Array-based methods capture whole transcriptomes at spots that contain multiple cells, typically 5 to 10 cells per spot. Imaging-based methods use targeted or whole-transcriptome probes to measure individual transcripts at subcellular resolution, enabling single-cell-level analysis within the tissue context.

The choice of platform depends on the research question. If the goal is to discover novel gene programs across the entire transcriptome at moderate resolution, array-based whole-transcriptome methods are appropriate. If the goal is to resolve cell types and states at single-cell resolution within defined regions of interest, imaging-based methods offer higher granularity. In small cell lung cancer research, the CosMx platform was used to profile over 600,000 cells across 105 primary and metastatic lymph node specimens, generating a single-cell spatial atlas that identified metastasis-enriched malignant subclusters with distinct metabolic and angiogenic programs [9]. This scale of single-cell spatial profiling is feasible with imaging-based platforms but not with array-based approaches.

Integrating Single-Cell and Spatial Data

A common and powerful workflow integrates scRNA-seq with spatial transcriptomics. The scRNA-seq data provide high-resolution cell type definitions and transcriptional states, while the spatial data provide the tissue coordinates for those states. Integration typically involves deconvolution of spatial spots using single-cell reference profiles, or mapping of single-cell clusters onto spatial coordinates using computational tools.

This integration has been applied extensively in prostate cancer research. A comprehensive atlas integrating 127 scRNA-seq samples and 9 spatial transcriptomic profiles spanning the disease continuum from healthy prostate to neuroendocrine carcinoma defined four evolutionarily connected malignant epithelial subtypes and revealed that the tumor microenvironment undergoes coordinated reprogramming during progression [12]. The spatial component was essential for showing that specific macrophage subsets and CAF populations maintain spatial co-localization and facilitate immune cell recruitment after neoadjuvant hormone therapy.

In colorectal cancer, integration of single-cell and spatial data has revealed spatially organized tumor-stroma-immune networks, including CAF-mediated extracellular matrix barriers, T cell-myeloid cooperation niches, and tertiary lymphoid structure-associated domains with both activating and suppressive microenvironments [11]. These findings depend on the complementary strengths of both technologies.

Defining Regions of Interest and Spatial Niches

Spatial transcriptomic analysis often begins with histology. Hematoxylin and eosin (H&E)-stained sections are used to identify regions of interest such as tumor core, invasive margin, necrotic areas, and lymphoid aggregates. Spatial transcriptomic data are then overlaid on these regions to characterize gene expression within each niche.

In glioblastoma, this approach has revealed that perinecrotic regions are more immunosuppressive than the endogenous tumor microenvironment, while perivascular regions are more pro-inflammatory [22]. Gradient analysis further showed that oligodendrocyte progenitor-like tumor cells tend to reside closer to tumor vasculature than to necrosis, possibly reflecting increased oxygen requirements. These spatial gradients are invisible in dissociated cell preparations.

Cellular neighborhood analysis is a related approach that defines recurrent multicellular configurations across the tissue. In small cell lung cancer, neighborhood analysis delineated distinct multicellular niches and identified a pan-immune hotspot whose abundance was an independent predictor of survival [9]. This type of analysis moves beyond single-cell identities to characterize the tissue-level organization of the tumor ecosystem.

At a Glance: Spatial Transcriptomics Applications Across Cancer Types

Practical Workflow for Applying Spatial Transcriptomics to Cancer Research

Step 1: Define the Biological Question and Required Resolution

Before selecting a platform, specify whether the question requires single-cell resolution, whole-transcriptome coverage, or both. Questions about cell type co-localization and neighborhood organization can be addressed with array-based platforms if cell types are well defined. Questions about rare cell states, subcellular localization, or precise cell-cell contacts require imaging-based platforms. Questions about large cohort biomarker discovery may not require spatial transcriptomics at all if histopathology-based prediction models can be trained on existing spatial data [5].

Step 2: Assemble the Cohort and Tissue Samples

Spatial transcriptomics requires fresh frozen or optimally preserved formalin-fixed paraffin-embedded tissue sections. Sample quality is critical. Necrotic tissue, surgical cautery artifacts, and prolonged ischemia time degrade RNA and compromise spatial data quality. For each sample, document tissue source, preservation method, section thickness, and quality metrics. In studies of treatment response, define response criteria before analysis. In muscle-invasive bladder cancer research, patients were classified as responders or non-responders according to RECIST 1.1-based radiological evaluation, with responders defined as those achieving complete or partial response after neoadjuvant immunochemotherapy [18].

Step 3: Generate Single-Cell Reference Data

If the spatial platform does not provide single-cell resolution, generate matched scRNA-seq data from the same or representative samples. The single-cell data serve as the reference for deconvolution and cell state annotation. In prostate cancer research, the integration of 127 scRNA-seq samples with 9 spatial profiles was central to defining malignant epithelial subtypes and their spatial distribution [12]. In colorectal cancer, an atlas of 4.27 million single cells from 1,670 patient samples was complemented with spatial transcriptomic data from 3.7 million cells [8].

Step 4: Perform Spatial Transcriptomic Profiling

Select the spatial platform based on the resolution and throughput requirements defined in Step 1. For whole-transcriptome discovery, array-based methods are appropriate. For single-cell spatial resolution, imaging-based methods such as CosMx are appropriate. Document all technical parameters, including probe sets, imaging depth, and quality control metrics. In small cell lung cancer research, the CosMx platform generated a comprehensive atlas of over 600,000 cells from 105 primary and metastatic lymph node specimens [9].

Step 5: Integrate and Analyze the Data

Integration involves mapping single-cell identities onto spatial coordinates, defining spatial niches, and analyzing cell-cell communication. Key analytical steps include:

• Deconvolution of spatial spots using single-cell reference profiles
• Clustering of spatial spots or cells to define recurrent spatial states
• Differential expression analysis between spatial niches
• Ligand-receptor analysis to infer cell-cell communication
• Cellular neighborhood analysis to define recurrent multicellular configurations
• Gradient analysis to characterize positional relationships between cell types and tissue structures

In glioblastoma, integration of snRNA-seq and spatial transcriptomics revealed patterns of segregation of tumor cell states and identified significant pathways in functionally relevant niches [22]. In colorectal cancer, communication analysis demonstrated that the MIF signaling pathway plays a prominent role within the tumor microenvironment communication network, with MIF-CD74-CD44 and MIF-CD74-CXCR4 identified as dominant receptor complexes [14].

Step 6: Validate Findings with Independent Methods

Spatial transcriptomic findings should be validated with orthogonal methods. Immunohistochemistry can confirm protein expression and localization. Functional experiments using patient-derived organoids or cell lines can test causal relationships. In colorectal cancer research, functional experiments using patient-derived organoids showed KRAS-dependent pro-tumorigenic polarization of neutrophils [8]. In glioblastoma research, knockdown of S100A6 reduced tumor cell proliferation, migration, and invasion, confirming the functional relevance of a lactylation-associated gene identified through spatial analysis [23].

Step 7: Derive and Test Biomarkers

If the goal is biomarker discovery, translate spatial findings into clinically applicable readouts. This may involve training machine learning models on histopathology images to predict spatial gene expression or tumor microenvironment states. In breast cancer research, a deep learning model called Path2Space was trained to predict spatial gene expression directly from histopathology slides, enabling the charting of the tumor microenvironment of 976 TCGA tumors and the identification of three spatially defined breast cancer subgroups with distinct survival outcomes [5]. In a pan-cancer tertiary lymphoid structure atlas, an artificial intelligence framework was trained to predict TLS maturation states directly from H&E-stained images and evaluated across TCGA and independent therapy cohorts [6].

Options and Tradeoffs in Spatial Transcriptomic Analysis

Whole-Transcriptome Versus Targeted Approaches

Whole-transcriptome spatial methods capture all expressed genes but at lower resolution or sensitivity. Targeted methods measure a predefined panel of genes at higher resolution and sensitivity. The tradeoff is between discovery power and precision. Whole-transcriptome methods are appropriate for hypothesis-generating studies. Targeted methods are appropriate for validating specific cell types, pathways, or biomarkers across large cohorts.

Fresh Frozen Versus Formalin-Fixed Paraffin-Embedded Tissue

Fresh frozen tissue generally yields higher RNA quality and more complete transcript coverage. Formalin-fixed paraffin-embedded tissue is more widely available in clinical archives and enables integration with histopathology. Many spatial platforms now support both tissue types, but the choice affects RNA integrity, probe penetration, and data quality. Document the tissue preservation method for each sample and include it as a covariate in downstream analyses.

Single-Cell Versus Spot-Level Analysis

Array-based spatial methods measure gene expression at spots containing multiple cells. Analysis can be performed at the spot level or deconvolved to estimate cell type proportions. Imaging-based methods measure individual transcripts and enable true single-cell analysis. The choice affects the granularity of cell type identification, the ability to detect rare cell states, and the complexity of the computational analysis.

Cost and Throughput Considerations

Spatial transcriptomic assays are expensive relative to bulk RNA sequencing. The high cost limits their application in large-scale biomarker discovery [5]. Researchers should consider whether the research question genuinely requires spatial information or whether bulk or single-cell approaches would suffice. For large cohort studies, histopathology-based prediction models trained on spatial data may offer a scalable and cost-effective alternative [5, 6].

Observations and Measurements in Spatial Transcriptomic Studies

Cell Type Composition and Spatial Distribution

Spatial transcriptomic data provide quantitative measurements of cell type abundance and distribution across tissue sections. In colorectal cancer, analysis of an atlas with 4.27 million single cells from 1,670 patient samples allowed tumor classification into immune desert, B cell enriched, T cell enriched, and myeloid cell enriched immune phenotypes [8]. These classifications have implications for therapy response and prognosis.

Tumor Cell State Segregation

Spatial analysis can reveal whether distinct tumor cell states occupy different regions of the tumor. In glioblastoma, oligodendrocyte progenitor-like and neural progenitor-like tumor cells significantly segregated in two of three samples analyzed [22]. This segregation has implications for understanding tumor evolution and resistance mechanisms.

Ligand-Receptor Interaction Networks

By integrating spatial expression data with ligand-receptor databases, researchers can infer which cells are communicating and whether those interactions are spatially restricted. In colorectal cancer, the MIF signaling pathway was identified as a prominent component of the tumor microenvironment communication network, with MIF-CD74-CD44 and MIF-CD74-CXCR4 as dominant receptor complexes [14]. Spatial analysis revealed distinct spatial functional partitioning of these signaling axes.

Immune Cell Infiltration Patterns

Spatial transcriptomics can characterize the distribution of immune cells relative to tumor cells and stromal structures. In bladder cancer, high SUV39H1 expression defined an immunosuppressive tumor microenvironment characterized by increased infiltration of regulatory T cells and M2 macrophages, along with reduced cytotoxic CD8-positive T cell activity, indicative of a cold tumor phenotype [13]. In small cell lung cancer, spatial analysis revealed vascular-immune crosstalk wherein endothelial cells orchestrate immune activation through avoidance of malignant cells while forming functional perivascular niches with cytotoxic T cells during lymph node metastasis [9].

Tertiary Lymphoid Structure Maturation

Spatial transcriptomics can characterize the maturation states of tertiary lymphoid structures, which are critical regulators of antitumor immunity. A pan-cancer spatial atlas spanning 12 cancer types characterized TLS spatial architecture and maturation states, revealing that TLS maturation is accompanied by coordinated remodeling of distinct niche cell populations and distance-dependent gradients in tumor programs [6].

Records and Documentation for Reproducible Spatial Analysis

Metadata Standards

Document the following for every sample:

• Tissue source and anatomical site
• Preservation method and fixation duration
• Section thickness and orientation
• Platform and probe set used
• Sequencing or imaging depth
• Quality control metrics including RNA integrity, spot or cell counts, and gene detection rates
• Clinical annotations including diagnosis, stage, treatment history, and response

Data Management and Sharing

Spatial transcriptomic datasets are large and complex. Establish a data management plan that addresses storage, versioning, and access. Public repositories such as the NCBI Data Resources provide infrastructure for depositing and accessing genomic data [2]. The EMBL-EBI Training program offers educational resources on data management and analysis [1]. Follow the FAIR Guiding Principles to ensure that data are findable, accessible, interoperable, and reusable [4].

Compliance with Data Sharing Policies

When working with human tissue samples, comply with applicable data sharing policies. The NIH Genomic Data Sharing Policy sets expectations for the sharing of genomic data generated from NIH-funded research [3]. Ensure that informed consent documents and institutional review board approvals are consistent with the intended data sharing and secondary use.

Common Failure Patterns and How to Avoid Them

Poor RNA Quality

Degraded RNA produces noisy spatial data with low gene detection rates. Avoid this by using fresh frozen tissue whenever possible, minimizing ischemia time, and verifying RNA integrity before library preparation. For formalin-fixed paraffin-embedded tissue, optimize fixation and retrieval protocols for the specific platform.

Inadequate Single-Cell Reference Data

Deconvolution of spatial spots depends on high-quality single-cell reference data. If the reference data do not capture the relevant cell types or states, deconvolution will produce inaccurate estimates. Generate matched single-cell data from the same or representative samples, and validate cell type annotations with orthogonal markers.

Batch Effects

Spatial transcriptomic data are sensitive to batch effects arising from sample processing, platform differences, and sequencing runs. Include technical replicates, randomize sample processing, and use computational batch correction methods. Document all technical parameters to enable downstream correction.

Overinterpretation of Spatial Correlations

Spatial co-localization does not prove functional interaction. Ligand-receptor analysis identifies potential communication axes, but these require functional validation. In colorectal cancer research, functionally validated mechanisms include IL1R1-positive inflammatory CAFs inducing M2 polarization and MDK-SDC4 interactions promoting regulatory T cell migration [11]. Without functional validation, spatial correlations remain hypotheses.

Ignoring Tissue Architecture

Spatial transcriptomic analysis should be guided by histology. Regions of interest should be defined based on tissue architecture, not solely on molecular clustering. In glioblastoma, the identification of perinecrotic and perivascular niches depended on integrating molecular data with histologic features [22].

Limitations of Spatial Transcriptomic Technologies

Resolution Limits

Array-based spatial methods measure gene expression at spots containing multiple cells, limiting the ability to resolve individual cell identities. Deconvolution methods can estimate cell type proportions but cannot fully recover single-cell expression profiles. Imaging-based methods offer higher resolution but are often limited to targeted gene panels.

Gene Coverage and Sensitivity

Whole-transcriptome spatial methods may miss low-abundance transcripts. Targeted methods measure only predefined genes, limiting discovery. The choice between coverage and sensitivity should be guided by the research question.

Data Integration Challenges

Integrating single-cell and spatial data is computationally demanding and requires careful validation. Differences in platform, tissue processing, and data distribution can introduce artifacts. The review literature on spatial transcriptomics in breast cancer notes bottlenecks in data integration, spatial resolution, standardization, and the need for functional validation [16].

Clinical Translation Barriers

Spatial transcriptomic assays are expensive and technically demanding, limiting their use in routine clinical practice. The translation of spatial findings into clinically applicable biomarkers requires validation in large cohorts and the development of scalable assays. Histopathology-based prediction models trained on spatial data offer a potential path forward [5, 6].

Safety and Regulatory Context

Human Tissue Research Oversight

Research involving human tissue samples is subject to institutional review board oversight and applicable regulations. Ensure that tissue acquisition, processing, and data sharing comply with institutional policies and informed consent requirements. The NIH Genomic Data Sharing Policy provides guidance on the sharing of genomic data from human subjects [3].

Data Privacy and Security

Spatial transcriptomic data derived from human tissues may contain identifiable information. Implement data de-identification procedures and restrict access to authorized researchers. Follow institutional data security policies and applicable privacy regulations.

Reproducibility Standards

Adopt reproducibility standards that enable other researchers to verify and build on your findings. This includes depositing raw and processed data in public repositories, documenting analysis workflows, and sharing code. The FAIR Guiding Principles provide a framework for data management and stewardship [4].

Professional Escalation Criteria

When to Seek Specialized Expertise

Spatial transcriptomic experiments require specialized expertise in experimental design, platform selection, and computational analysis. Consult with core facility staff, bioinformaticians, and statisticians before initiating experiments. If your institution lacks these resources, consider collaborating with institutions that have established spatial transcriptomic programs.

When to Reassess the Experimental Approach

Reassess the experimental approach if:

• RNA quality metrics fall below platform-specific thresholds
• Spatial data show poor alignment with histologic features
• Deconvolution results are inconsistent with known biology
• Batch effects dominate biological variation
• Findings cannot be validated with orthogonal methods

When to Escalate to Clinical or Regulatory Consultation

Escalate to clinical or regulatory consultation if:

• Findings have potential implications for patient care or treatment decisions
• The research involves investigational therapies or biomarkers
• Data sharing or secondary use raises ethical or regulatory questions
• Intellectual property considerations arise from biomarker discovery

Frequently Asked Questions

What is the difference between single-cell RNA sequencing and spatial transcriptomics?

Single-cell RNA sequencing measures gene expression in individual cells after tissue dissociation, providing high-resolution cell type identification but losing spatial context. Spatial transcriptomics measures gene expression within intact tissue sections, preserving the physical coordinates of each measurement. The two technologies are complementary and are often integrated to combine cell type resolution with spatial architecture.

How do I choose between array-based and imaging-based spatial transcriptomic platforms?

Array-based platforms such as 10x Visium capture whole-transcriptome data at spots containing multiple cells, making them suitable for discovery studies. Imaging-based platforms such as CosMx measure individual transcripts at subcellular resolution, making them suitable for single-cell spatial analysis and targeted validation. The choice depends on whether the research question requires whole-transcriptome coverage or single-cell resolution.

Can spatial transcriptomics be performed on formalin-fixed paraffin-embedded tissue?

Yes, many spatial transcriptomic platforms support formalin-fixed paraffin-embedded tissue, which is widely available in clinical archives. However, RNA quality may be lower than in fresh frozen tissue, and probe penetration may be reduced. Optimize fixation and retrieval protocols for the specific platform and verify RNA quality before library preparation.

How is spatial transcriptomic data integrated with single-cell RNA sequencing data?

Integration typically involves using single-cell data as a reference to deconvolve spatial spots or map single-cell clusters onto spatial coordinates. Computational tools align cell type identities with spatial positions, enabling the characterization of tissue architecture and cell-cell interactions. In prostate cancer research, integration of 127 scRNA-seq samples with 9 spatial profiles defined malignant epithelial subtypes and their spatial distribution [12].

What are tertiary lymphoid structures and why are they important in cancer?

Tertiary lymphoid structures are organized aggregates of immune cells that form in non-lymphoid tissues, including tumors. They are critical regulators of antitumor immunity. Spatial transcriptomics has enabled the construction of a pan-cancer TLS atlas spanning 12 cancer types, revealing that TLS maturation is accompanied by coordinated remodeling of distinct niche cell populations and distance-dependent gradients in tumor programs [6].

How can spatial transcriptomics predict treatment response?

Spatial transcriptomic data can identify tumor microenvironment features associated with treatment response. In biliary tract cancer, integration of urinary proteomics with single-cell and spatial transcriptomics revealed that patients achieving durable clinical benefit from immune checkpoint inhibitors exhibited enrichment of immune activation pathways, and a machine learning-derived four-protein panel robustly predicted durable benefit [7]. In breast cancer, spatial tumor microenvironment landscapes enabled more accurate predictions of patient response to chemotherapy and trastuzumab compared with conventional bulk-sequencing-based biomarkers [5].

What are the main limitations of spatial transcriptomic technologies?

Main limitations include resolution limits in array-based methods, gene coverage and sensitivity tradeoffs, data integration challenges, high cost, and barriers to clinical translation. The review literature on spatial transcriptomics in breast cancer notes bottlenecks in data integration, spatial resolution, standardization, and the need for functional validation [16].

What quality controls are essential for spatial transcriptomic experiments?

Essential quality controls include verification of RNA integrity, assessment of spot or cell counts, evaluation of gene detection rates, alignment of molecular data with histologic features, and validation of findings with orthogonal methods such as immunohistochemistry or functional experiments. Document all technical parameters to enable batch correction and reproducibility.

Related Bioinformatics Guides

• Single-Cell Transcriptomics of Host-Pathogen Interactions During Viral Infection
• Modern Transcriptomics: From Bulk RNA-Seq to Single-Cell and Spatial Resolution
• Spatial Transcriptomics: Mapping the Cellular Atlas
• Spatial Transcriptomics Alignment and Cellular Neighborhood Analysis
• Computational Modeling of Avian Influenza Virus Hemagglutinin, Receptor Interactions: Implications for Host Tropism

References and Further Reading

• EMBL-EBI Training. European Bioinformatics Institute.
• NCBI Data Resources. National Center for Biotechnology Information.
• Genomic Data Sharing Policy. National Institutes of Health.
• The FAIR Guiding Principles. Scientific Data.
• AI-predicted spatial transcriptomics unlocks breast cancer biomarkers from pathology.. Cell, 2026.
• Pan-cancer spatial atlas of tertiary lymphoid structures.. Science (New York, N.Y.), 2026.
• Dynamic urinary proteomics integrates single-cell and spatial transcriptomics to reveal tumour microenvironment and predict immunotherapy response in biliary tract cancer.. Gut, 2026.
• Single-cell integration and multi-modal profiling reveals phenotypes and spatial organization of neutrophils in colorectal cancer.. Cancer cell, 2026.
• Single-cell spatial transcriptomics reveals tumor microenvironment heterogeneity in primary and lymph node-metastatic small cell lung cancer.. Cell reports. Medicine, 2026.
• Integrated spatial morpho-transcriptomics predicts functional traits in pancreatic cancer.. Science advances, 2025.
• Integrated single-cell and spatial transcriptomics reveal the colorectal cancer microenvironment.. Critical reviews in oncology/hematology, 2026.
• Integrated single-cell and spatial transcriptomic profiling decodes lineage plasticity and immune microenvironment remodeling in prostate cancer progression.. Molecular cancer, 2026.
• Single-cell and spatial transcriptomics reveal SUV39H1 as a master epigenetic driver of immunosuppressive niche and stemness in bladder cancer.. 2026.
• Integrated Single-Cell and Spatial Transcriptomic Analysis Reveals the Immunoregulatory Role of MIF Signaling in Colorectal Cancer.. 2026.
• Perspective on the integration of radiomics and spatial omics in the analysis of the tumor microenvironment of bladder cancer and prospects for precision diagnosis and treatment.. 2026.
• Spatial Transcriptomics in Breast Cancer: Advances and Applications.. 2026.
• Single-Cell and Spatial Transcriptomics Reframe the Immunosuppressive Microenvironment of Neuroendocrine Neoplasms.. 2026.
• Remodeling of the tumor microenvironment in muscle-invasive bladder cancer: insights from single-cell and spatial transcriptomics after neoadjuvant immunochemotherapy.. 2026.
• Decoding early lung adenocarcinoma progression by single-cell and spatial transcriptomics reveals a CMA-related prognostic signature.. 2026.
• Integrated analysis of single-cell RNA-seq and spatial transcriptomics to identify the lactylation-related protein TUBB2A as a potential biomarker for glioblastoma in cancer cells by machine learning. Frontiers in Immunology, 2025.
• Spatial transcriptomics and multi-omics reveal relapse and resistance mechanisms of EndMT-derived CAFs mediated by TNC and FLNC in glioblastoma. Journal of Translational Medicine, 2025.
• Spatial transcriptomics reveals segregation of tumor cell states in glioblastoma and marked immunosuppression within the perinecrotic niche. Acta Neuropathologica Communications, 2024.
• Single-cell and spatial transcriptomics reveal lactylation-associated tumor cell clusters and define a prognostic risk model in glioblastoma. BMC Cancer, 2025.
• Machine learning-enhanced discovery of a basement membrane-related gene signature in glioblastoma via single-cell and Spatial transcriptomics. Journal of Translational Medicine, 2025.
• Spatial transcriptomics in breast cancer: providing insight into tumor heterogeneity and promoting individualized therapy. Frontiers in Immunology, 2024.
• Current Role and Future Frontiers of Spatial Transcriptomics in Genitourinary Cancers. Cancers, 2025.
• The promising application of cell-cell interaction analysis in cancer from single-cell and spatial transcriptomics. Seminars in Cancer Biology, 2023.
• Advances in spatial transcriptomics and its applications in cancer research. Molecular Cancer, 2024.

This article is educational and does not replace validated analysis plans, institutional policy, clinical interpretation, or specialist review.