Figure 1. A three-layer workflow connects cell identity, tissue location, and local metabolic phenotype.

Single-cell spatial transcriptomics metabolomics workflow design helps researchers connect cell identity, tissue location, and local metabolic phenotype in one evidence chain. This article explains when to combine single-cell transcriptomics, spatial transcriptomics, and spatial metabolomics, how each layer contributes to tissue mechanism research, and what analysis logic can support candidate pathway and regulator discovery.

Key takeaways

• Single-cell transcriptomics identifies cell types and cell states, but removes cells from their native tissue location.
• Spatial transcriptomics maps gene expression patterns back to tissue architecture.
• Spatial metabolomics adds local metabolite distribution and phenotype-level evidence.
• A strong workflow links cell identity, spatial domains, pathway changes, metabolite enrichment, and candidate regulators.
• This strategy is most useful when the research question involves tissue heterogeneity, developmental transitions, microenvironment remodeling, or metabolic reprogramming.

Why Combine Three Layers

Single-cell transcriptomics, spatial transcriptomics, and spatial metabolomics answer different parts of the same tissue biology question.

Single-cell transcriptomics helps define which cell types or cell states are present. It is useful for discovering major cell groups, rare populations, transitional states, marker genes, and condition-associated cell populations.

The limitation is location. Tissue dissociation removes cells from their original environment. After dissociation, researchers can still classify cells, but they lose native tissue architecture.

Spatial transcriptomics adds that context. It shows where gene expression programs occur across a tissue section. This supports spatial domain analysis, cell-type mapping, region-specific marker discovery, and tissue architecture-aware interpretation.

Spatial metabolomics adds a third layer. It helps show where metabolites accumulate or change inside the tissue system. This is important when transcriptional changes need phenotype-level support.

Together, the three layers create a practical research logic:

1. Cell identity: Which cells are involved?
2. Spatial location: Where are these cells or gene programs located?
3. Metabolic phenotype: What local metabolite changes support the mechanism?
4. Candidate mechanism: Which genes, pathways, or regulators connect these signals?

For teams planning single-cell sequencing, spatial transcriptomics, or spatial metabolomics, the question is not whether more data are better. The question is whether each data layer answers a different biological uncertainty.

Core Study Questions

A single-cell spatial transcriptomics metabolomics workflow is most useful when the project cannot be answered by one data layer alone.

Which cells drive the phenotype?

Single-cell transcriptomics is often the starting point when the tissue contains mixed cell populations.

It can help identify major cell types, rare cell groups, transitional states, and condition-associated cell populations.

Useful outputs include:

• Cell clusters
• Cell type labels
• Marker genes
• Cell proportion differences
• Trajectory or pseudotime patterns
• Candidate cell populations for downstream mapping

This layer is useful when the study needs to distinguish whether a phenotype comes from a known cell type, a rare cell group, or a changing cell state.

Where do molecular changes occur?

Spatial transcriptomics is useful when the location of molecular changes matters.

A gene program may appear important in single-cell data, but its biological meaning can depend on where it occurs. It may localize to a tissue boundary, vascular niche, tumor margin, developmental zone, or specialized tissue compartment.

Useful outputs include:

• Spatial expression maps
• Spatial domains
• Region-specific genes
• Cell-type spatial distribution
• Spatial deconvolution results
• Tissue architecture-associated expression patterns

For projects focused on tissue organization, spatial transcriptomics should be planned early instead of added after sample collection.

Which metabolites support the mechanism?

Spatial metabolomics is useful when the study needs local biochemical evidence.

For example, a transcriptomic pathway may suggest lipid metabolism, amino acid metabolism, oxidative stress, or secondary metabolite activity. Spatial metabolomics can help determine whether related metabolites show regional enrichment.

Useful outputs include:

• Metabolite feature tables
• Metabolite spatial maps
• Region-enriched metabolite signals
• Pathway-level metabolite interpretation
• Gene-metabolite association results

This layer helps move the project from "gene expression changed" toward "a spatially localized phenotype may support this mechanism."

Sample and Group Design

The workflow should begin with the biological contrast, not the platform list.

A strong study design defines what will be compared before deciding how each omics layer will be generated. Common contrasts include developmental stages, treatment groups, tissue regions, genotypes, disease models, stress conditions, or phenotype groups in research specimens.

Species, tissue, and condition selection

The tissue model should match the research question.

Plant studies may focus on seed development, nutrient accumulation, fruit ripening, stress response, or tissue-specific metabolism. Animal and biomedical research models may focus on tumor microenvironments, organ development, immune niches, tissue injury, or treatment-associated molecular changes in research settings.

Before sample collection, researchers should define:

• Species and tissue type
• Experimental conditions
• Biological replicate plan
• Tissue region of interest
• Key phenotype or endpoint
• Whether cell dissociation is feasible
• Whether adjacent sections can be prepared

This information helps decide whether the project should include single-cell transcriptomics, single-nucleus transcriptomics, spatial transcriptomics, spatial metabolomics, or a smaller pilot design.

Matched and adjacent sections

The three data layers do not always come from the exact same section.

Single-cell transcriptomics often requires a cell or nuclei suspension. Spatial transcriptomics and spatial metabolomics usually use tissue sections. Because each layer has different sample handling needs, matched tissue regions and adjacent sections are important for interpretation.

Good planning should document:

• Which tissue block or region is used for each assay
• Whether sections are adjacent or region-matched
• How morphology will be recorded
• How metadata will connect samples across platforms
• Whether tissue heterogeneity may affect interpretation

If tissue regions are not matched, cross-omics integration may become difficult. The analysis may still be possible, but the biological interpretation becomes less direct.

Time points and treatment groups

Time points should be chosen based on the expected biology.

For developmental or response studies, multiple stages may be useful when the goal is to reconstruct a trajectory. For simpler contrast studies, fewer groups may be enough if the phenotype is clear and well controlled.

The key is to avoid collecting many weakly defined groups. A smaller design with clear metadata and matched tissue regions is often more interpretable than a broad design with limited biological structure.

Recommended Workflow

A practical single-cell spatial transcriptomics metabolomics workflow usually follows five planning steps.

Step 1: Define the biological contrast

Start with the comparison that matters.

Examples include genotype A versus genotype B, early versus late development, treated versus untreated research models, resistant versus sensitive tissue regions, or high-metabolite versus low-metabolite phenotypes.

The contrast should be linked to a measurable endpoint, such as tissue morphology, developmental stage, metabolite class, pathway activity, or region-specific phenotype.

Step 2: Build the cell atlas

Single-cell transcriptomics can define the cellular composition of the tissue.

The analysis may include read processing, cell filtering, normalization, clustering, marker gene detection, cell type annotation, cell proportion comparison, and trajectory analysis when appropriate.

The output is not just a cluster plot. The goal is to produce a reference cell atlas that can guide spatial interpretation.

For tissue types where dissociation is difficult, fragile, frozen, or biased toward certain cell types, researchers may consider related approaches such as snRNA sequencing.

Step 3: Map cell types in tissue

Spatial transcriptomics helps place gene expression and cell-type signals back into the tissue.

Depending on the platform and resolution, the analysis may include spatial clustering, spatially variable gene detection, marker visualization, deconvolution, cell-type mapping, pathway enrichment, and tissue region annotation.

If the main challenge is data analysis rather than sample generation, spatial transcriptomics data analysis can support workflows such as cell-type mapping, spatial domain analysis, and visualization.

Step 4: Add metabolite distribution

Spatial metabolomics maps metabolite features across tissue regions.

This layer is useful when local biochemical changes are central to the mechanism. Examples include lipid distribution, amino acid accumulation, secondary metabolite localization, oxidative stress-related metabolites, or treatment-associated metabolic shifts in research models.

Spatial metabolomics should not be added only because it sounds advanced. It should answer a defined question that transcriptomics alone cannot resolve.

Step 5: Prioritize candidate mechanisms

The final step is integration.

A useful analysis does not simply place three datasets side by side. It connects cell clusters, spatial domains, differentially expressed genes, pathway enrichment, metabolite maps, and candidate regulators into a ranked evidence chain.

Possible integrated outputs include:

• Cell type to tissue region mapping
• Region-specific pathway modules
• Gene-metabolite correlation tables
• Candidate metabolic enzymes
• Transcription factor candidates
• Ligand-receptor or cell communication signals
• Ranked mechanism shortlist for follow-up validation

This is where multi-omics becomes decision-supporting instead of data-heavy.

Mid-article CTA:
If your tissue study already has a defined phenotype but the right omics combination is unclear, CD Genomics can help review the study question, sample format, and data layers before finalizing the workflow.

Figure 2. Integrated analysis logic for connecting cell identity, tissue location, and metabolite distribution in a single-cell spatial transcriptomics metabolomics workflow.

Integrated Analysis Logic

The workflow does not end after each omics layer is processed.

The main value comes from connecting cell atlas construction, spatial region mapping, metabolite distribution, and candidate mechanism prioritization. Figure 2 summarizes how these layers can be organized into a mechanism-oriented analysis framework.

From study design to omics input

The first layer of analysis is the experimental design.

A project may compare different species, genotypes, treatments, tissue regions, or response groups. For developmental or dynamic biological questions, multiple stages may be considered when the biology requires trajectory or temporal analysis.

The next layer is the omics input:

• Single-cell transcriptomics
• Spatial transcriptomics
• Spatial metabolomics

Each input should be connected to a defined question. Single-cell data should support cell atlas construction. Spatial transcriptomics should support tissue mapping. Spatial metabolomics should support local phenotype interpretation.

From basic analysis to key cells

After data processing, the analysis can move into three linked modules.

Cell atlas construction defines the cell populations and cell states. Outputs may include cell clustering, cell type annotation, marker genes, and trajectory-related results.

Differential cell or region analysis compares groups, regions, or stages. Outputs may include differentially expressed genes, differential metabolite features, GO / KEGG / GSEA enrichment, and region-specific pathway modules.

Key cell-type prioritization helps identify the cell populations most relevant to the research question. Evidence may include cell proportion shifts, group-specific cell changes, and pathway-enriched cell populations.

This step prevents the project from becoming a broad list of clusters. It focuses the analysis on cells and regions that explain the phenotype.

From multi-omics mining to mechanism hypotheses

The deeper analysis stage connects each data layer into a mechanism-oriented framework.

Based on single-cell data, researchers can refine core cell populations, infer trajectory or state transitions, and build transcription factor-target or pathway networks.

Integrated with spatial transcriptomics, the workflow can map cells to tissue regions, define spatial expression domains, and identify in situ gene programs.

Integrated with spatial metabolomics, the workflow can locate metabolite-enriched regions, link genes with metabolite patterns, and build region-gene-metabolite evidence chains.

The goal is not to claim causality from correlation. The goal is to prioritize mechanism hypotheses that can be tested with follow-up experiments.

Advanced enrichment and comparison

Advanced analysis may include cell-cell communication analysis, co-expression module analysis, spatial niche analysis, or spatial trajectory analysis.

For comparative studies, the workflow may also support cross-condition or cross-species comparison. This can help identify conserved cell-type patterns, shared pathway modules, or condition-specific gene-metabolite regulation.

These analyses are most useful when the sample design, metadata, and tissue matching are planned before the project begins.

Analysis Logic

The main goal of integrated analysis is to turn multi-layer data into interpretable biological evidence.

A useful framework is:

Cell clustering and annotation

Single-cell analysis usually starts with quality filtering, normalization, clustering, marker gene identification, and annotation.

Cell type labels should be supported by marker genes and biological knowledge. When possible, annotations can also be compared with published references or internal project knowledge.

The final goal is a usable cell reference, not only a visualization.

Spatial domain and deconvolution analysis

Spatial transcriptomics analysis can identify spatially variable genes, tissue domains, and region-specific expression programs.

When the spatial platform does not directly provide single-cell resolution, deconvolution can help estimate which cell types are likely present in each spot or region. This is where single-cell data and spatial transcriptomics become complementary.

A related resource, how to integrate scRNA-seq with spatial transcriptomics, can support readers who want a deeper explanation of this specific analysis step.

Gene-metabolite association

Gene-metabolite association analysis helps connect transcriptomic programs with local metabolic features.

For example, a spatial domain may show increased expression of metabolic enzymes and enriched metabolite signals from a related pathway. This does not prove causality by itself, but it helps prioritize stronger mechanism candidates.

Useful outputs may include:

• Gene-metabolite correlation pairs
• Region-level pathway enrichment
• Metabolic enzyme expression patterns
• Spatial overlap between pathway genes and metabolites
• Candidate regulators linked to both gene and metabolite layers

Candidate regulator selection

A strong integrated report should help researchers move toward follow-up experiments.

Candidate regulators may include transcription factors, metabolic enzymes, signaling molecules, ligand-receptor pairs, or pathway genes. These candidates should be ranked based on multiple evidence types, not only fold change.

Evidence may include cell-type specificity, spatial localization, pathway relevance, metabolite association, and consistency across biological groups.

When Spatial Metabolomics Matters

Spatial metabolomics is most valuable when the biological question involves local metabolic function.

It may not be necessary for every spatial transcriptomics project. If the main goal is only to classify cell types or map gene expression regions, single-cell and spatial transcriptomics may be enough.

Add it for metabolic reprogramming

Spatial metabolomics should be considered when the project focuses on metabolic reprogramming.

Examples include:

• Nutrient accumulation in plant seeds
• Fruit ripening or senescence
• Tumor metabolic niches
• Tissue stress response
• Local lipid or amino acid changes
• Treatment-associated molecular effects in research models

In these cases, transcriptomics can suggest pathway activity, but metabolomics can provide local phenotype evidence.

Add it for phenotype validation

Spatial metabolomics can also support phenotype interpretation.

If a spatial transcriptomics result suggests that a tissue region is involved in lipid synthesis, flavonoid accumulation, oxidative stress, or amino acid metabolism, spatial metabolomics can test whether relevant metabolite features are enriched in the same region.

This helps strengthen the evidence chain:

gene program - tissue location - metabolite pattern - candidate mechanism

Defer it when the question is only cell mapping

Spatial metabolomics may be deferred when the project mainly asks:

• What cell types are present?
• Where are the major cell populations located?
• Which genes mark each tissue region?
• How do cell proportions differ between groups?

In these cases, single-cell transcriptomics plus spatial transcriptomics may answer the core question.

A staged plan can be practical: first define the cell and spatial gene architecture, then add spatial metabolomics if the first results point to metabolic mechanisms.

Figure 3. A decision tree helps determine when spatial metabolomics adds value.

Application Scenarios

This workflow is useful across several tissue-based research areas.

Plant development and seed biology

Plant tissues often have strong spatial organization. Different regions may show different cell types, gene programs, and metabolite accumulation patterns.

A three-layer workflow can help connect cell differentiation, tissue zones, nutrient storage, and candidate regulatory genes. This is useful for studies of seed development, fruit tissue, root zones, stress response, and specialized metabolism.

Readers planning plant-focused projects may also explore spatial omics solutions for plant research.

Fruit ripening and stress response

Fruit ripening, post-harvest senescence, and stress response often involve both transcriptional changes and metabolite shifts.

Single-cell transcriptomics can identify changing cell populations. Spatial transcriptomics can locate these changes across tissue layers. Spatial metabolomics can show whether antioxidant compounds, amino acids, lipids, or other metabolite groups follow a region-specific pattern.

This is useful when the research goal is to explain how local tissue programs connect to visible phenotypes.

Tumor microenvironment research

In tumor microenvironment research, different tissue regions may contain different cell states, immune niches, stromal areas, and metabolic patterns.

Single-cell data can define cell states. Spatial transcriptomics can show where these states are organized. Spatial metabolomics can support questions about local metabolic differences across tumor regions in research models.

For this topic, spatial omics should be framed as research-use molecular characterization, not diagnosis or treatment decision-making.

Developmental biology

Developmental biology often requires spatial and temporal interpretation.

Single-cell transcriptomics can define developmental trajectories. Spatial transcriptomics can place those trajectories into tissue architecture. Spatial metabolomics can show whether local biochemical states align with lineage progression, differentiation zones, or organ patterning.

This makes the workflow useful for studies where cell fate, tissue region, and metabolic state are linked.

Common Design Pitfalls

Multi-omics studies can fail at the design stage even when each platform works technically.

Unmatched tissue regions

The most common issue is mismatched tissue input.

If single-cell data come from one region and spatial data come from another, integration becomes weaker. If metabolomics uses a different tissue layer with different morphology, gene-metabolite interpretation may become indirect.

The best practice is to plan tissue blocks, adjacent sections, morphology review, and metadata capture before sample processing.

Weak metadata design

Metadata are essential for integration.

Important metadata may include sample group, tissue region, developmental stage, treatment condition, section position, preservation method, batch, and morphology notes.

Without metadata, the analysis may still generate plots, but biological interpretation becomes difficult.

Correlation without prioritization

Multi-omics integration often produces many associations.

Not every gene-metabolite correlation is biologically meaningful. A useful workflow should rank candidates based on several evidence types, such as cell-type specificity, spatial overlap, pathway relevance, group difference, and consistency across biological groups.

The final output should support decision-making for follow-up validation.

Deliverables and Interpretation

Deliverables should be planned before the project starts.

This helps researchers understand what each data layer can provide and how the final report will support interpretation.

Single-cell deliverables

Possible single-cell deliverables include:

• Raw sequencing data
• Processed expression matrix
• Quality control summary
• Cell clustering results
• Marker gene tables
• Cell type annotation
• Cell proportion comparison
• Trajectory or cell-state analysis when appropriate

These results help define the cell populations that will be used for spatial interpretation.

Spatial transcriptomics deliverables

Possible spatial transcriptomics deliverables include:

• Spatial expression matrix
• Tissue image registration
• Spot or cell coordinate files
• Spatial expression maps
• Spatial domain results
• Spatially variable gene tables
• Cell-type mapping or deconvolution outputs
• Pathway enrichment results

These outputs help connect gene expression to tissue architecture.

Spatial metabolomics deliverables

Possible spatial metabolomics deliverables include:

• Metabolite feature table
• Spatial metabolite distribution maps
• Region-specific metabolite patterns
• Pathway-level metabolite interpretation
• Group comparison results when applicable
• Integration-ready metabolite profiles

These outputs help support local phenotype interpretation.

Integrated report

The integrated report should bring the three layers together.

A useful report may include:

• Study design summary
• Cross-platform QC overview
• Cell type and spatial domain interpretation
• Gene expression and metabolite association
• Candidate pathway modules
• Ranked regulator or mechanism shortlist
• Figures prepared for research discussion
• Notes on limitations and recommended follow-up

The report should make clear which conclusions are directly supported by the data and which are hypotheses for validation.

FAQ

What is a single-cell spatial transcriptomics metabolomics workflow?
It is a multi-omics study design that combines single-cell transcriptomics, spatial transcriptomics, and spatial metabolomics. Single-cell transcriptomics identifies cell types and cell states. Spatial transcriptomics places gene expression patterns back into tissue context. Spatial metabolomics maps local metabolite features. Together, they help researchers connect cell identity, tissue location, pathway activity, and metabolic phenotype.

When should spatial metabolomics be added to a spatial transcriptomics study?
Spatial metabolomics should be considered when the research question involves metabolic phenotype, nutrient accumulation, metabolic reprogramming, stress response, or pathway-level biochemical evidence. It is especially useful when transcriptomics suggests a metabolic pathway but the study still needs local metabolite evidence. If the goal is only cell-type mapping or gene expression localization, spatial metabolomics may be deferred.

Can single-cell transcriptomics and spatial transcriptomics use the same tissue sample?
They usually require coordinated but different sample preparation. Single-cell transcriptomics often uses dissociated cells or isolated nuclei, while spatial transcriptomics uses tissue sections. The best design uses matched tissue regions, adjacent sections, and complete metadata. This helps connect cell type information with spatial gene expression patterns during analysis.

What sample information is needed before designing a spatial multi-omics study?
Useful information includes species, tissue type, preservation method, tissue region, experimental groups, biological replicate plan, phenotype description, sectioning plan, and whether cell or nuclei isolation is feasible. Morphology images, pilot data, or previous RNA quality information can also help. For difficult or archived tissues, feasibility review is recommended before finalizing the workflow.

What QC metrics are important for single-cell and spatial multi-omics projects?
QC should cover sample quality, tissue morphology, cell or nuclei quality, library quality, sequencing data quality, mapping performance, spot or cell detection, background signal, clustering quality, and consistency across groups. For spatial metabolomics, QC may also include feature detection, spatial signal stability, and region-level interpretability. Exact metrics depend on platform and sample type.

How are cell types mapped from single-cell RNA-seq to spatial transcriptomics data?
Single-cell RNA-seq can provide a reference cell atlas. Spatial transcriptomics data can then be compared with this reference using cell-type marker genes, deconvolution methods, or mapping algorithms. The goal is to estimate where cell types or cell states are located in the tissue. Results should be interpreted with morphology, platform resolution, and tissue context.

How can gene expression and metabolite distribution be interpreted together?
Researchers can compare spatial gene expression patterns with local metabolite maps. For example, a tissue region may show enriched metabolic enzyme expression and related metabolite features. This can support a pathway hypothesis. The strongest interpretations usually combine spatial overlap, pathway knowledge, group differences, and candidate regulator analysis. Correlation alone should not be treated as proof of causality.

Are these workflows intended for diagnosis or treatment decisions?
No. The workflows described here are for research use only. They are intended to support molecular characterization, tissue biology studies, mechanism research, and hypothesis generation. They are not intended for diagnosis, treatment decisions, disease monitoring, therapeutic decision-making, or individual health assessment.

Plan an Integrated Spatial Multi-Omics Study

A single-cell spatial transcriptomics metabolomics workflow is most useful when each omics layer answers a different research question.

If your project involves tissue heterogeneity, spatially localized gene programs, or metabolic phenotype interpretation, CD Genomics can help review:

• Tissue type and sample format
• Biological contrast and study groups
• Single-cell or single-nucleus feasibility
• Spatial transcriptomics platform fit
• Spatial metabolomics value
• QC and deliverable expectations
• Integrated bioinformatics needs

To discuss whether this workflow fits your tissue study, contact CD Genomics for a research-use project consultation.

Compliance and Trust Statement

This content is intended for research use only. The services and workflows described are not intended for diagnosis, treatment decisions, disease monitoring, therapeutic decision-making, or individual health assessment.

Project feasibility depends on tissue type, sample preservation, morphology, input condition, platform requirements, and study design. CD Genomics recommends sample and study design review before project initiation, especially for archived, degraded, difficult, or limited tissue specimens.

Customer project information and sample details should be handled according to applicable confidentiality and data protection requirements.

References

1. Single-cell spatial (scs) omics: Recent developments in data analysis
   Supports the broader context of single-cell and spatial omics integration across transcriptomics, genomics, epigenomics, proteomics, and metabolomics.
2. Deep Learning in Single-Cell and Spatial Transcriptomics Data Analysis: Advances and Challenges from a Data Science Perspective
   Supports discussion of high-dimensional, sparse, noisy, and multimodal data challenges in single-cell and spatial transcriptomics analysis.
3. Multimodal Spatial Omics: From Data Acquisition to Computational Integration
   Supports the article's framing that multimodal spatial omics can profile molecular layers in tissue context and requires computational integration across data modalities.
4. The technological landscape and applications of single-cell multi-omics
   Supports the article's discussion of single-cell multi-omics as a broader technology landscape for connecting multiple molecular layers.
5. Evaluating Integrative Strategies for Incorporating Phenotypic Features in Spatial Transcriptomics
   Supports the concept that spatial transcriptomics can be integrated with other phenotypic and imaging-derived features for tissue-context interpretation.