Single-cell spatial resolution is necessary when your conclusion depends on which cells express a molecular signal, cell state, or the relative position of cells. It is not necessary for every spatial transcriptomics study. If your question can be answered by comparing tumour with stroma, diseased with unaffected tissue, or one tissue region with another, a regional or spot-based approach may suffice. Select the lowest spatial resolution that preserves the biology needed for the next development decision.
What is single-cell spatial resolution?
In a cell-resolved workflow, detected transcripts are mapped to spatial coordinates and assigned to individual cells. This can produce a cell-by-gene matrix while retaining each cell’s tissue position.
However, cell boundaries are usually computational estimates. Segmentation may use nuclear staining, boundary markers, interior markers, or combined image features. Platform documentation describes segmentation as an approximation for assigning transcripts to cells. Dense tissue, irregular cell shapes and weak staining can affect the result.
You should treat single-cell resolution as an analytical capability, not simply an instrument specification. Reliable interpretation requires quality control of the tissue image, segmentation, transcript assignment, and cell-type annotation.
Single-cell resolution is essential when averaging would obscure the relevant biological signal
The clearest justification for single-cell spatial analysis is that a mixed measurement would obscure the biology you need to see.
1. Identifying the cellular source of signal
A region may contain tumour cells, fibroblasts, endothelial cells, and multiple immune populations. A regional increase in a transcript does not reveal which population produced it. If the development question depends on attributing expression to a particular cell type or state, individual-cell assignment becomes important.
2. Rare cell populations may influence the conclusion
A small but biologically relevant population can disappear within an averaged signal. Cell-resolved analysis can locate rare populations and determine whether they occupy a consistent niche. This may matter in biomarker discovery, treatment-response studies, and investigations of resistance.
The study still needs enough tissue sections, samples, and biological replicates to establish that the pattern is reproducible. Detecting many cells in one specimen does not compensate for inadequate patient-level replication.
3. Spatial neighbourhoods are part of the hypothesis
Single-cell resolution is usually best when the analysis asks whether defined cell types co-localise, remain separated or form recurrent neighbourhoods. Examples include assessing immune exclusion at a tumour boundary or testing whether a cell state is enriched near a particular stromal population.
Spatial proximity can support a hypothesis about potential cell-cell interaction, but it does not by itself demonstrate functional communication. Ligand-receptor analysis, protein-level evidence or functional validation may still be needed.
4. Treatment response differs between nearby cells
Two cells in the same tissue compartment may respond differently to treatment. If you need to determine which cellular states change after exposure, or whether a response occurs only in a defined subpopulation, regional profiling can merge distinct effects. Single-cell spatial resolution allows you to analyse those differences without removing the cells from their tissue context.
When is regional or spot-based resolution sufficient?
Single-cell analysis may add complexity without changing the answer when the main question concerns larger tissue structures or broad molecular programmes.
