Understanding Spatial Genomics: Principles and Techniques

Inquiry

Spatial genomics is rapidly emerging as a core technology driving the frontiers of life science research. By precisely integrating gene expression profiles with the native spatial organization of tissues, it offers scientists an unprecedented view into cell-cell interactions, tissue heterogeneity, and the molecular mechanisms underlying health and disease. In this guide, we present a comprehensive overview of spatial genomics, covering its fundamental concepts, leading technology platforms — including Visium, Slide-seq, DBiT-seq, and CosMx — and a comparative analysis of resolution, throughput, and sample compatibility. We also provide practical recommendations for platform selection and key quality-control considerations, followed by an outlook on future trends. Whether your work focuses on basic research, mechanistic studies of complex diseases, or the integration of spatial multi-omics data, this resource is designed to deliver structured, data-driven, and globally relevant insights to help your research stand out in the competitive international scientific landscape.

What is Spatial Genomics? Definition and Research Context

Definition

Spatial genomics refers to a class of high-resolution molecular profiling technologies that preserve the native spatial organization of cells within intact tissues while capturing genome-, transcriptome-, or multi-omics-level information. Unlike conventional bulk or single-cell sequencing, which dissociates cells and loses positional context, spatial genomics directly maps molecular readouts back to their precise coordinates within the tissue architecture. This integration of morphology and molecular data enables a far more accurate interpretation of biological processes.

Research Context

The concept of spatially resolved molecular analysis has matured over the past decade, accelerated in recent years by advances in high-throughput sequencing, in situ hybridization (ISH), and imaging-based assays. These developments have led to the rapid growth of spatial transcriptomics, spatial proteomics, and other spatial-omics approaches.

Key drivers behind this rise include:

• Increasing demand for detailed characterization of tissue heterogeneity and microenvironments
• The need for higher-precision mapping in oncology, developmental biology, and neuroscience
• Improved compatibility of workflows with challenging sample types such as FFPE sections
• Maturation of in situ amplification, high-plex imaging, and barcoding technologies

In recognition of its transformative potential, Nature Methods named "Spatially Resolved Transcriptomics" as the 2020 Method of the Year, underscoring its growing influence across multiple research domains.

Applications in Research

• Developmental biology – mapping cell fate decisions and lineage trajectories in situ
• Tumor microenvironment profiling – revealing spatial relationships among immune cells, stromal cells, and cancer cells
• Neuroscience – characterizing neuron subtypes, synaptic organization, and spatial signaling patterns
• Multi-omics integration – combining spatial transcript data with proteomics, metabolomics, or epigenomics to capture a more complete systems-level view

From a research perspective, the power of spatial genomics lies not only in its data richness but in its ability to preserve histological context. This allows molecular findings to be interpreted in the light of tissue architecture, bridging the gap between classical histopathology and modern omics science.

Main Technology Platforms – Visium vs Slide-seq vs DBiT-seq vs CosMx

Spatial genomics has evolved into a diverse field with multiple commercial and academic platforms, each optimized for different research needs. While all share the goal of preserving tissue architecture and capturing spatially resolved molecular information, their underlying chemistries, resolutions, and data outputs differ substantially. Four widely adopted platforms are described below.

10x Genomics Visium

Visium uses spatially barcoded capture spots printed on glass slides to bind poly-A RNA molecules directly from tissue sections. Each spot captures transcripts from one or more cells, enabling downstream cDNA synthesis and sequencing. Visium supports both fresh-frozen and formalin-fixed paraffin-embedded (FFPE) samples, with workflows tailored for unbiased transcriptome-wide capture (fresh-frozen) or targeted gene panels (FFPE probe-based).

Gene expression and clustering of stitched data. (Eagles, N. J., et al., 2024, BMC genomics)

Slide-seq V2

Slide-seq employs a surface densely coated with DNA-barcoded microbeads, each with a known spatial coordinate. Tissue sections are transferred onto the bead array, and mRNA hybridizes to bead-bound oligos for reverse transcription and sequencing. The small bead size (~10 µm) allows near- cell-level resolution and higher spatial granularity than Visium, though with increased sensitivity to tissue handling quality.

SlideCNA identifies spatial CNA patterns in Slide-seq MBC samples. (Zhang, D., et al., 2025, Genome Biology)

DBiT-seq (Deterministic Barcoding in Tissue Sequencing)

DBiT-seq integrates microfluidics to deliver barcoded oligonucleotides across tissue sections in two orthogonal directions, creating a grid of spatial coordinates. This approach supports multi-omics profiling — for example, combining spatial transcriptomics with protein detection via oligo-conjugated antibodies. DBiT-seq is compatible with both fresh-frozen and FFPE tissue and is particularly suited for integrative spatial biology studies.

Deconvolution of ST data of varying resolution from multiple technologies by STdeconvolve. (Miller, B. F., et al., 2022, Nature communications)

NanoString CosMx Spatial Molecular Imager

CosMx uses high-plex, single-molecule fluorescence imaging to detect RNA or protein targets in situ at subcellular resolution. It can analyze hundreds to thousands of targets per experiment, with workflows optimized for FFPE preservation. Unlike sequencing-based methods, CosMx directly images transcripts or proteins, producing highly quantitative spatial maps that retain morphological details.

Comparative Overview

From a researcher's perspective, platform selection depends on the scientific question:

• For broad transcriptome profiling in diverse sample types, Visium offers a robust entry point.
• For higher spatial resolution and fine-grained cellular mapping, Slide-seq V2 is advantageous.
• For integrating RNA and protein in the same spatial framework, DBiT-seq is the method of choice.
• For ultra-high-plex, morphology-rich imaging, CosMx provides unmatched subcellular detail.

Comparative Analysis — Resolution, Throughput, and Sample Compatibility

Benchmarking studies have highlighted that no single spatial genomics technology excels in all performance dimensions. Resolution, molecular capture efficiency, throughput, and sample compatibility vary depending on platform design and chemistry. Choosing the right tool therefore requires weighing these factors against the biological question and available resources.

Resolution

Resolution defines the smallest physical unit in which transcripts or proteins can be spatially assigned.

• Highest resolution (<10 µm) — Slide-seq V2 and DBiT-seq can achieve near- cell-level mapping, enabling detailed reconstruction of cell-cell interfaces and microenvironment boundaries.
• Subcellular resolution — CosMx offers imaging-based detection at the level of individual transcripts or proteins within compartments.
• Lower resolution (~55 µm) — Visium spots capture multiple cells, making it more suitable for mapping tissue-level patterns rather than cell-level heterogeneity.

Throughput and Coverage

Throughput describes the number of spatial features (spots, beads, pixels) profiled per unit area.

• High-density capture — Stereo-seq and Slide-seq V2 have high spot density, enabling finer mapping of large tissue areas.
• Moderate density with broad coverage — Visium captures fewer, larger spots but can process multiple sections in parallel, making it suitable for comparative studies across tissues or conditions.
• Targeted high-plex imaging — CosMx offers high molecular multiplexing but covers smaller areas per run compared to sequencing-based platforms.

Sample Compatibility

The type and preservation status of the tissue critically impact platform performance.

• Fresh-frozen — All four platforms perform optimally with fresh-frozen tissue, preserving RNA integrity.
• FFPE — Visium (probe-based), DBiT-seq, and CosMx have workflows compatible with FFPE, which is advantageous for archived or clinical research specimens.
• Multi-omics readiness — DBiT-seq and some advanced CosMx protocols allow integration of RNA and protein profiling in the same tissue section.

Summary Table of Key Metrics

Research Take-Home Message

If the aim is high-resolution mapping of cell-type niches or fine anatomical structures, Slide-seq V2 or CosMx is recommended. For integrating multiple molecular layers in the same tissue, DBiT-seq stands out. For broader surveys with a balance of cost and data depth, Visium remains a well-established choice. Importantly, the performance of each platform can vary across tissue types, making pilot runs or reviewing platform-specific benchmarks essential before large-scale studies.

Practical Recommendations — Platform Selection and Key QC Considerations

Selecting a spatial genomics platform is not merely a matter of technical specifications; it requires aligning the platform's strengths with the specific biological questions, tissue types, and downstream analysis goals. Based on both published benchmarking data and hands-on laboratory experience, the following recommendations can help researchers maximize the quality and interpretability of their spatial datasets.

Platform Selection Strategy

• Broad transcriptome surveys with robust pipelines — Visium is well-suited for projects aiming to map large-scale gene expression patterns across tissues, especially when reproducibility and commercial support are priorities.
• Fine-scale mapping of microenvironments — Slide-seq V2 is ideal for resolving small cellular niches, developmental gradients, and tissue borders at near- cell-level resolution.
• Integrated multi-omics studies — DBiT-seq allows concurrent measurement of RNA and protein markers, making it powerful for dissecting cell-cell signaling pathways and multimodal spatial networks.
• High-plex targeted profiling — CosMx excels when the research requires subcellular localization of a predefined set of transcripts or proteins, particularly in FFPE samples where sequencing methods may be limited.

Key QC Considerations

1. Tissue Handling and Sectioning

• Use fresh-frozen or optimally fixed FFPE sections to preserve molecular integrity.
• Maintain consistent section thickness (typically 5–10 µm) to balance RNA yield and spatial fidelity.
• Avoid mechanical damage and ensure the section adheres firmly to the capture surface or imaging substrate.

2. Permeabilization and Hybridization Conditions

• Over-permeabilization can cause RNA diffusion, blurring spatial boundaries; under-permeabilization reduces capture efficiency.
• Optimize conditions for each tissue type — manufacturer defaults are a starting point but not always optimal.

3. RNA Integrity and Quality Control

• Assess RNA integrity number (RIN) or DV200 metrics before committing to library preparation.
• For FFPE samples, confirm probe performance on small pilot areas to avoid wasting full slides.

4. Sequencing Depth and Read Structure

• Spatial datasets often require deeper sequencing than cell-resolved RNA-seq to capture low-abundance transcripts.
• Monitor per-spot read counts during pilot runs to determine the optimal depth for full experiments.

5. Data Processing and Annotation

• Integrate spatial data with cell-resolved RNA-seq references to improve cell-type annotation accuracy.
• Use spatially aware analysis tools (e.g., Seurat with spatial extension, SEDR, Giotto) to preserve tissue context in downstream analysis.

Practical Tip for Research Teams

A small-scale pilot experiment on the chosen platform — using the same tissue type, preservation method, and staining strategy planned for the main study — can save significant resources by identifying protocol adjustments early. This is particularly important when working with rare or irreplaceable samples.

Conclusion and Future Trends

Spatial genomics has rapidly transformed from a niche technology into a central pillar of modern molecular biology. By uniting spatial context with high-content molecular profiling, it enables researchers to uncover patterns and interactions that were previously hidden in dissociated bulk or cell-resolved datasets. Over the past few years, platforms such as Visium, Slide-seq, DBiT-seq, and CosMx have each contributed unique strengths — from robust whole-transcriptome coverage to subcellular resolution and multi-omics integration.

Current State

At present, no single platform universally outperforms the others across all metrics. The optimal choice depends on the interplay of resolution requirements, tissue type, molecular targets, and experimental scale. Benchmark studies consistently show that careful matching of platform capabilities to the biological question is more important than simply chasing the highest resolution or throughput.

Emerging Trends

1. Subcellular Resolution Mapping
   • Next-generation platforms are pushing below the 1 µm scale, enabling visualization of transcript and protein localization within specific cellular compartments such as nuclei, dendrites, or synaptic terminals.
   • This fine-grained view will be crucial for neuroscience, developmental biology, and intracellular signaling studies.
2. Integrative Spatial Multi-Omics
   • Combining RNA, protein, chromatin accessibility, and metabolite mapping in the same tissue section is becoming feasible.
   • Technologies like DBiT-seq and advanced imaging-sequencing hybrids will expand the dimensionality of spatial data, supporting systems-level modeling of tissue function.
3. AI-Driven Analysis and Interpretation
   • Machine learning is increasingly used to detect subtle spatial patterns, infer cell-cell communication networks, and integrate spatial data with clinical or phenotypic metadata.
   • Spatial transcriptomics datasets will benefit from automated feature extraction, particularly for large-scale projects spanning multiple tissues or conditions.
4. Greater Accessibility and Standardization
   • As protocols become more streamlined and costs decrease, spatial genomics is expected to move from specialized core facilities into routine use in many research laboratories.
   • Efforts to establish data standards and interoperable formats will improve reproducibility and cross-study comparisons.

Final Perspective

For research teams, the strategic adoption of spatial genomics will not only enhance the resolution and depth of biological insights but also open the door to novel hypotheses that can only be addressed in a spatially resolved framework. As instrumentation, chemistry, and computational methods continue to advance, spatial genomics will likely become as foundational to tissue-based research as next-generation sequencing is to bulk molecular profiling today.

Researcher-Focused Collaboration Invitation

At CD Genomics, we recognize that every spatial genomics project presents unique scientific questions and technical challenges. Our team combines deep expertise in experimental design, tissue processing, and advanced spatial data analysis to deliver results that meet the highest standards of reproducibility and interpretability. Whether you require high-resolution mapping of cellular microenvironments, integrative RNA–protein profiling, or robust analysis pipelines for complex datasets, we provide tailored solutions that align with your research objectives.

If you are planning a spatial genomics study and want to ensure optimal platform selection, precise execution, and insightful interpretation, our scientists are ready to collaborate with you. Contact us to discuss your project needs and explore how our spatial genomics services can accelerate your research.

References

1. Eagles, N. J., Bach, S. V., Tippani, M., Ravichandran, P., Du, Y., Miller, R. A., ... & Collado-Torres, L. (2024). Integrating gene expression and imaging data across Visium capture areas with visiumStitched. BMC genomics, 25(1), 1077.
2. Nguyen, D. H., Duque, V., Phillips, N., Mecawi, A. S., & Cunningham, J. T. (2023). Spatial transcriptomics reveal basal sex differences in supraoptic nucleus gene expression of adult rats related to cell signaling and ribosomal pathways. Biology of sex Differences, 14(1), 71.
3. Zhang, D., Segerstolpe, Å., Slyper, M., Waldman, J., Murray, E., Strasser, R., ... & Klughammer, J. (2025). SlideCNA: spatial copy number alteration detection from Slide-seq-like spatial transcriptomics data. Genome Biology, 26(1), 112.
4. Liu, Y., Yang, M., Deng, Y., Su, G., Enninful, A., Guo, C. C., ... & Fan, R. (2020). High-spatial-resolution multi-omics sequencing via deterministic barcoding in tissue. Cell, 183(6), 1665-1681.
5. Miller, B. F., Huang, F., Atta, L., Sahoo, A., & Fan, J. (2022). Reference-free cell type deconvolution of multi-cellular pixel-resolution spatially resolved transcriptomics data. Nature communications, 13(1), 2339.