Prioritize cohort-scale expression
For treatment effects across many independent samples, bulk RNA-seq may offer the most defensible balance of depth, replication, covariate modeling, and cost.
Default: Bulk RNA-seq
Technical Bulletin
Bulk RNA-Seq, Single-Cell RNA-Seq or Spatial Transcriptomics?
How to Choose the Right Approach
Discuss Your StudyDecision first
Choose the Biological Resolution You Need
Bulk RNA-seq, single-cell RNA-seq, and spatial transcriptomics are not successive upgrades of the same assay. They preserve different information, impose different sample constraints, and support different levels of inference.
The shortest useful answer
Choose bulk RNA-seq when the sample-level transcriptome is the intended unit of analysis and cohort scale matters. Choose single-cell or single-nucleus RNA-seq when cellular composition or cell-state heterogeneity is central. Choose spatial transcriptomics when the location of a signal within tissue architecture is necessary to answer the question.
When more than one resolution is essential, a staged or integrated design may be more defensible than forcing one assay to serve every objective.
Bulk RNA-seqSample-level average with broad transcriptome coverage and efficient cohort scaling.
Single-cell RNA-seqCell-resolved identities, states, trajectories, and compositional changes without native coordinates.
Spatial transcriptomicsGene expression mapped to tissue locations at spot-, cell-, or subcellular resolution depending on platform.
Core comparison
Bulk RNA-Seq vs. Single-Cell RNA-Seq vs. Spatial Transcriptomics
The table separates biological resolution from practical design. “Higher resolution” is not automatically “better” if it reduces replication, conflicts with specimen quality, or measures a level that is unnecessary for the endpoint.
| Decision dimension | Bulk RNA-seq | Single-cell / single-nucleus RNA-seq | Spatial transcriptomics |
|---|---|---|---|
| Primary analytical unit | Whole sample, tissue piece, sorted population, or experimental replicate | Individual cell or nucleus, nested within a biological sample | Spatial spot, bin, segmented cell, region of interest, or molecular coordinate |
| Biological view | Average expression across the sampled cell mixture | Cell types, rare populations, cell states, composition, and transitions | Expression associated with tissue domains, cellular neighborhoods, or histological structures |
| Native spatial context | No | No; dissociation or nuclei isolation removes original coordinates | Yes, although resolution and assignment accuracy are platform-dependent |
| Transcriptome breadth | Broad and generally deep at the sample level | Broad for whole-transcriptome assays, but sparse per cell and sensitive to library depth | Whole-transcriptome or targeted, depending on chemistry; coverage and resolution trade off |
| Sample compatibility | Flexible for fresh, frozen, low-input, and many fixed-tissue workflows | Fresh viable cells or isolated nuclei are common; compatibility depends on tissue and protocol | Requires sections and platform-compatible preservation, morphology, dimensions, and analyte quality |
| Replication and cohort scale | Often strongest Easier to prioritize more biological replicates | More expensive and operationally demanding per sample; multiplexing may help | Slide area, tissue geometry, imaging, and processing can limit cohort scale |
| Rare population detection | Indirect; signals may be diluted or require deconvolution | Direct if the population survives preparation and enough cells are sampled | Possible when resolution, target coverage, segmentation, and tissue sampling support it |
| Differential analysis | Mature sample-level models for replicated designs | Must account for cells nested within samples; pseudobulk or mixed models are often appropriate | Must account for spatial dependence, repeated sections, regions, and sample-level replication |
| Main quality risks | RNA degradation, library composition, cell-mixture confounding, batch effects | Dissociation bias, viability, ambient RNA, doublets, sparsity, annotation and pseudoreplication | Section quality, morphology, capture efficiency, imaging, segmentation, spot mixing and spatial batch effects |
| Typical best fit | Large-cohort expression studies, perturbation screens, validation, homogeneous samples | Cell atlas construction, immune heterogeneity, rare states, cell-type-specific response | Tissue niches, invasive boundaries, spatial pathology, local interactions, anatomical gradients |
| Interpretation boundary | Cannot assign a mixed-tissue signal to a specific population without additional evidence | Cell-state association does not establish causality; spatial relationships are not measured directly | Proximity or co-localization does not demonstrate communication or causal mechanism |
Research-goal framework
Is the endpoint sample-level?
Is heterogeneity the endpoint?
Does location change meaning?
Are discovery and validation different?
Start with the Decision the Data Must Support
A useful selection process moves from research endpoint to resolution, specimen feasibility, replication, analysis, and validation—not from a preferred platform backward to a biological story.
Resolve cell identities and states
When a mixed specimen contains biologically important populations or rare states, cell- or nucleus-resolved profiling can separate composition from within-cell-state change.
Default: Single-cell / nucleiRetain tissue architecture
If invasive margin, immune exclusion, fibrotic niche, vascular compartment, or anatomical layer is central, dissociated data alone cannot answer the primary question.
Default: SpatialUse a staged design
High-resolution discovery in a focused subset can define cell states or spatial signatures, followed by bulk, targeted, imaging, or orthogonal validation in a larger cohort.
Default: Combined strategyFeasibility gate
Let the Specimen and Study Design Constrain the Choice
A theoretically ideal modality can fail if tissue handling, sample count, section geometry, or metadata do not support it.
Specimen reality
- Fresh viability and time to processing
- Availability of frozen nuclei
- FFPE age and RNA quality
- Tissue size, morphology, and sectioning
- Expected cell fragility or dissociation bias
Design reality
- Independent donors or experimental units
- Balanced groups, batches, and time points
- Paired or longitudinal structure
- Clinical covariates and tissue regions
- Discovery versus validation cohort
Operational reality
- Sequencing and imaging capacity
- Slide area and specimen throughput
- Storage and computational resources
- Annotation and pathology input
- Turnaround and validation budget
Do not trade away replication for nominal resolution without examining the consequence. Thousands of cells or spatial locations within one specimen are observational units nested inside that specimen; they are not substitutes for independent donors, animals, cultures, or other biological replicates.
Combination strategies
When One Modality Is Not Enough
Combination does not mean running every assay on every sample. A staged allocation can preserve both high-resolution discovery and cohort-level inference.
Bulk first, resolution second
Use a scalable bulk cohort to identify robust sample-level programs, subtypes, or response groups. Select representative specimens for single-cell or spatial profiling to identify the cellular source or tissue location of those programs.
Best when cohort structure is the starting asset.Single-cell reference plus spatial mapping
Use single-cell or single-nucleus data to define cell states and reference profiles, then map or deconvolve those states in tissue sections. The reliability of mapping depends on compatible biology, feature overlap, platform resolution, and reference quality.
Best when both identity and location matter.High-resolution discovery, practical validation
Discover a cell-state or spatial signature in a focused set, then translate it into a targeted expression panel, pathology assay, flow-cytometry strategy, or bulk signature for validation across more independent samples.
Best when translation is the endpoint.Open-access evidence
Spatial Transcriptomics Is Itself a Family of Choices
The label “spatial transcriptomics” covers materially different measurement systems. Imaging-based and sequencing-based approaches vary in gene coverage, resolution, sensitivity, field of view, tissue compatibility, processing time, and dependence on segmentation or deconvolution.
The open-access comparison shown below illustrates how targeted imaging platforms can use different probe designs and optical decoding strategies. This matters because a decision to retain spatial context is only the first step. The platform must then be matched to whether the study requires whole-transcriptome discovery or a targeted panel, large tissue area or subcellular localization, fresh-frozen or fixed specimens, and discovery or validation.
Published guidance also emphasizes that platform capabilities evolve quickly. Specifications, compatible sample types, gene panels, capture area, and analytical software should therefore be verified at project start rather than treated as permanent properties.
Project-start checklist
Information to Lock Before Samples Enter the Workflow
The following items are more useful for selecting a modality than a generic request for the “highest resolution” assay.
Primary endpointDefine the comparison, decision, or mechanistic question in one sentence.
Biological unitIdentify donors, animals, cultures, organoids, sections, and repeated measurements.
Required resolutionSpecify whether sample, cell type, cell state, tissue region, cell, or subcellular location is necessary.
Specimen inventoryRecord preservation, age, viability, dimensions, pathology, available material, and processing history.
Group and batch mapConfirm that condition, site, operator, collection date, and assay batch are not inseparable.
Expected biologyEstimate cell diversity, rare-population frequency, anatomical structure, and effect size where possible.
Analysis unit and modelPredefine sample-aware differential analysis, spatial structure, covariates, and multiplicity control.
Quality thresholdsDefine assay-specific review points without relying on universal cutoffs across tissues.
Validation routeChoose an independent cohort, targeted assay, imaging method, protein measurement, or functional experiment.
Decision deliverableSpecify whether the output is a pathway, cell state, target list, spatial niche, or biomarker hypothesis.
Related reading and support
Continue from Technology Choice to Analysis Strategy
Choose the Assay Around the Decision
Share the research question, sample inventory, replicate structure, available metadata, and intended validation route. ComputaBio can help translate these constraints into a fit-for-purpose transcriptomics analysis plan.
Discuss Your ProjectReferences
Selected Scientific Sources
- Lim, H. J.; Wang, Y.; Buzdin, A.; et al. A practical guide for choosing an optimal spatial transcriptomics technology from seven major commercially available options. BMC Genomics 2025, 26, 47. https://doi.org/10.1186/s12864-025-11235-3. Distributed under Open Access license CC BY 4.0.
- Heath, J. R.; Ribas, A.; Mischel, P. S. Single-cell analysis tools for drug discovery and development. Nature Reviews Drug Discovery 2016, 15, 204–216. https://doi.org/10.1038/nrd.2015.16.
- Stark, R.; Grzelak, M.; Hadfield, J. RNA sequencing: the teenage years. Nature Reviews Genetics 2019, 20, 631–656. https://doi.org/10.1038/s41576-019-0150-2.
- Zimmerman, K. D.; Espeland, M. A.; Langefeld, C. D. A practical solution to pseudoreplication bias in single-cell studies. Nature Communications 2021, 12, 738. https://doi.org/10.1038/s41467-021-21038-1.
- Heumos, L.; Schaar, A. C.; Lance, C.; et al. Best practices for single-cell analysis across modalities. Nature Reviews Genetics 2023, 24, 550–572. https://doi.org/10.1038/s41576-023-00586-w.
- You, Y.; Fu, Y.; Li, L.; et al. Systematic comparison of sequencing-based spatial transcriptomic methods. Nature Methods 2024, 21, 1743–1754. https://doi.org/10.1038/s41592-024-02325-3.
- Chen, T.-Y.; You, L.; Hardillo, J. A. U.; Chien, M.-P. Spatial transcriptomic technologies. Cells 2023, 12, 2042. https://doi.org/10.3390/cells12162042.
- Conesa, A.; Madrigal, P.; Tarazona, S.; et al. A survey of best practices for RNA-seq data analysis. Genome Biology 2016, 17, 13. https://doi.org/10.1186/s13059-016-0881-8.
Online Inquiry
Submit your project details below, and our team will respond within 24 hours.
x
Need help getting the data you need?
Talk to our technical team about your project!
I Want To Talk