Spatial Transcriptomics vs Single-Cell RNA-Seq: Complementary Approaches

Spatial transcriptomics and single-cell RNA sequencing (scRNA-seq) are often framed as rivals, but they answer different questions about the same tissue. Single-cell RNA-seq resolves cell identity at high depth after dissociating a sample, while spatial transcriptomics keeps each transcript tied to its original location. Understanding what each method captures, and where they diverge, is the starting point for deciding when to use one, the other, or both together.

What spatial transcriptomics and scRNA-seq each capture

Single-cell RNA-seq isolates individual cells from a dissociated tissue sample and sequences the RNA from each cell, producing a transcriptome-wide expression profile for every cell captured. Because dissociation removes tissue architecture entirely, scRNA-seq answers questions about cell identity and state without reference to a cell's original location, and without needing to decide in advance which genes matter. That flexibility is part of what has made scRNA-seq the workhorse of modern cell biology.

Droplet-based scRNA-seq platforms can profile tens of thousands of individual cells from a single sample within minutes, a throughput enabled by massively parallel single-cell profiling methods that generate transcriptome data across large cell populations. This is one of several recent single-cell genomics advances that have expanded what scRNA-seq can detect. That combination of throughput and depth makes scRNA-seq the default tool for cataloging every cell type present in a tissue, including rare populations a targeted assay might otherwise miss. It underlies much of the cell-type discovery work behind reference atlases of human organs.

Spatial transcriptomics, sometimes called spatial RNA-seq, measures gene expression while preserving the coordinates of each transcript, either by capturing RNA on a spatially barcoded surface or by imaging individual molecules directly within the tissue section. Rather than a list of cell types, it produces a molecular map of gene activity across intact tissue, a capability central to ongoing work exploring tissue architecture using spatially resolved data. This theme is discussed at length across the broader spatial biology landscape. Sequencing- and imaging-based spatial transcriptomics platforms each capture that coordinate information differently, but both share the same goal of tying a gene expression measurement to a specific point in the tissue rather than discarding that context during preparation.

Single-cell RNA-seq and spatial transcriptomics provide complementary information. The choice between using one method, the other, or a combination depends on the biological question being asked (Figure 1).

Figure 1: A side-by-side comparison of the scRNA-seq and spatial transcriptomics workflows, from tissue sample to final data output. Credit: AI-generated image created using Google Gemini (2026).

Resolution vs location: The core trade-off between spatial transcriptomics and scRNA-seq

Single-cell RNA-seq and spatial transcriptomics trade off two properties: transcript depth per cell versus physical location within the tissue (Table 1). Single-cell RNA-seq measures more of the transcriptome per cell but discards spatial context entirely, while spatial transcriptomics preserves location but often measures less per cell or per position, depending on the platform. Neither limitation is a flaw so much as a direct consequence of how each method captures its data.

Table 1: A comparison of the core characteristics of scRNA-seq and spatial transcriptomics.

Dissociating tissue for scRNA-seq is not a neutral step. Digesting tissue into a single-cell suspension can itself alter gene expression in a way that mimics an injury response. A widely used dissociation protocol, for example, was shown to trigger dissociation-induced expression changes in a subpopulation of cells that would otherwise have been indistinguishable from unaffected cells in the same sample. That finding is a reminder that scRNA-seq data reflects both underlying biology and artifacts introduced by sample preparation, which matters when interpreting subtle expression differences.

Spatial transcriptomics carries its own resolution costs. Sequencing-based platforms, including barcoded capture platforms such as Visium by 10x Genomics, often capture RNA from several cells within a single spot, requiring computational deconvolution to separate cell types. Imaging-based platforms, by contrast, resolve individual molecules but only within a preselected gene panel rather than the whole transcriptome, trading breadth for precision. Choosing between those two spatial strategies mirrors, on a smaller scale, the broader resolution-versus-coverage decision that separates spatial transcriptomics from scRNA-seq in the first place.

When to use spatial transcriptomics vs scRNA-seq

Choosing between spatial transcriptomics and scRNA-seq starts with the question being asked, not with a platform's marketing claims. A project cataloging every cell type present in a tissue, with no strong hypothesis about which genes matter, is usually better served by scRNA-seq, since it can profile the whole transcriptome without excluding any cell type in advance.

A spatial transcriptomics technology overview cataloging the field's methods and trends since its earliest spatially resolved gene expression techniques found that new platform development has continued to accelerate, with no single method yet dominating the field. That continued diversification is why platform selection, including the choice between scRNA-seq and spatial transcriptomics, still depends on the specific biological question rather than a default option.

A project asking how a defined set of genes is organized within a tissue, such as mapping receptor-ligand pairs across a tumor margin or an immune infiltrate, is usually better served by spatial transcriptomics, where the physical arrangement of cells carries information that a dissociated cell list cannot provide. Evaluating the full range of spatial transcriptomics platform choices side by side remains the most reliable way to match resolution and coverage to a specific research question. Many research groups treat this as an iterative decision rather than a one-time choice, revisiting the trade-off as new instrument generations and lower-cost options reach the market.

A short framework helps narrow the decision before committing budget, time, and tissue to either method:

1. Define whether the project needs unbiased, whole-transcriptome discovery or a targeted, spatially resolved readout.
2. Confirm whether tissue architecture, such as cell neighborhoods or tumor margins, is central to the research question.
3. Assess whether a matched single-cell reference dataset already exists or would need to be generated.
4. Weigh the resolution and coverage trade-offs of available spatial platforms against the biology being studied.
5. Decide whether an integrated workflow, using both methods together, would serve the project better than either method alone.

How researchers integrate scRNA-seq and spatial transcriptomics data

Because neither method alone provides both transcriptome depth and spatial context, a growing share of spatial biology projects integrate scRNA-seq and spatial transcriptomics datasets computationally rather than choosing one method exclusively. Integration typically uses a scRNA-seq dataset as a reference to interpret ambiguous or lower-resolution spatial data, filling in cell-type detail that the spatial platform alone cannot resolve.

DMPK Playbook: Complete Workflows Not Workarounds

This eBook explores key challenges and solutions across metabolite identification, bioanalysis, and tissue distribution, highlighting complete workflow approaches.
View eBook

Advertisement

A review of integrating single-cell spatial data described the two methods as complementary, noting that scRNA-seq provides deep transcriptional information at true single-cell resolution while spatial transcriptomics maps the coordinates. That complementarity is the basis for most integration strategies developed over the past several years. It also explains why no single method has fully replaced the other despite rapid platform development on both sides.

The open-source tool Seurat, developed at Satija Lab, uses an anchor-based strategy for comprehensive single-cell data integration, identifying pairs of mutually similar cells across a scRNA-seq reference and spatial dataset to guide label transfer. Other open-source toolkits, including Scanpy developed by Theis Lab, offer comparable integration and label-transfer functions built around different underlying algorithms. This general approach enables researchers to transfer detailed cell-type labels from a well-characterized scRNA-seq atlas to spatial data that may lack the resolution to call those same cell types directly. Batch effects and differences in gene panel coverage between the two datasets still require careful handling during alignment.

A combined workflow for spatial transcriptomics and scRNA-seq

A typical combined workflow begins with a scRNA-seq experiment on a matched or similar tissue sample, generating a reference set of gene expression profiles for each expected cell type. That reference is then used to interpret the spatial dataset, whether the platform captures multiple cells per spot or resolves individual molecules. The quality of that reference has a direct effect on how accurately the final spatial map reflects true cell-type composition.

For spot-based spatial platforms, deconvolution methods estimate the proportion of each reference cell type present at every spot. For example, the open-source tool cell2location, developed at the Bayraktar Lab, was designed for fine-grained cell type mapping in spatial transcriptomics data by combining a single-cell reference with spot-level spatial measurements. It has since been applied to build detailed cellular maps of tissues, including the human gut and lymph node. Other deconvolution methods take a similar reference-based approach but differ in how they model technical noise, so comparing multiple methods against known tissue biology is common practice before settling on a final pipeline.

The resulting combined dataset retains both the transcriptional depth of scRNA-seq and the spatial context of the original tissue section, allowing researchers to ask not just which cell types are present but also where they sit relative to one another. Validating the combined result against known tissue anatomy or independent markers remains an important step before drawing biological conclusions, one piece of the broader spatial data analysis workflow that spans preprocessing, cell typing, and spatial statistics.

Spatial transcriptomics and scRNA-seq as complementary tools

Spatial transcriptomics and scRNA-seq answer different questions about the same underlying biology, and the choice between them depends on whether cell identity or tissue location matters more for a given project. scRNA-seq remains the more efficient route to an unbiased, whole-transcriptome catalog of cell types, while spatial transcriptomics remains the approach that preserves where each measurement originated within the tissue.

For many projects, the most productive path is not to choose one method over another but to combine them, using a single-cell reference to interpret spatial data and recover both depth and location in a single analysis. As integration tools mature and become more accessible, treating spatial transcriptomics and scRNA-seq as complementary, rather than competing, methods is likely to become the default across genomics research.

This content includes text that has been created with the assistance of generative AI and has undergone editorial review before publishing. Technology Networks' AI policy can be found here.