Figure 1. Plant ligand-receptor analysis considers peptides, hormones, receptor complexes, and co-receptors.

Summary

Plant ligand-receptor analysis for single-cell omics helps researchers infer how plant cell types may exchange molecular signals through peptide ligands, hormones, receptor-like kinases, receptor complexes, co-receptors, and other signaling systems. In plant single-cell RNA-seq, single-nucleus RNA-seq, and spatial transcriptomics, this analysis can move a study beyond "which cell types are present" toward "which cell types may send or receive specific developmental, immune, nutrient, or stress-related signals."

This guide explains why ligand-receptor analysis is different in plants, what input data are needed, how plant-specific databases and tools such as PlantCellChatDB can support cell-cell communication analysis, and how to interpret communication scores, source-target networks, pathway-level outputs, and spatial context. It also outlines QC checks, deliverables, project planning questions, and research-use-only limitations.

Key Takeaways

• Plant ligand-receptor analysis connects cell-type expression profiles with possible signaling relationships.
• Plant communication analysis should account for hormones, peptides, receptor-like kinases, receptor-like proteins, co-receptors, receptor complexes, agonists, and antagonists.
• Plant-specific resources are important because many ligand-receptor databases were built around human or mouse biology.
• Reliable analysis depends on cell type annotation, gene ID mapping, expression depth, replicate design, and database coverage.
• Spatial transcriptomics can add tissue-position context to communication inference.
• Results should be treated as hypothesis-generating unless supported by independent experimental validation.

Why Plant Ligand-Receptor Analysis Matters

Single-cell and spatial omics have changed how plant researchers study tissue organization. Instead of averaging gene expression across a whole root tip, leaf, embryo, flower, or stress-treated tissue, researchers can now examine gene expression at the level of cell clusters, cell states, nuclei, or spatial regions.

That resolution creates a new question. Once cell types are identified, how do they coordinate with each other?

Plant ligand-receptor analysis addresses this question by combining expression data with known or predicted molecular interactions. A source cell type is evaluated for ligand or signaling molecule expression. A target cell type is evaluated for receptor, co-receptor, or receptor-complex expression. The result is an inferred communication relationship that can be ranked, filtered, visualized, and compared across tissues or conditions.

This is especially useful in plants because development and stress adaptation depend on coordinated signals across tissues. Root meristem maintenance, vascular differentiation, stomatal regulation, immune response, regeneration, organ initiation, and hormone response all require information exchange between neighboring or distant cells.

From Cell Identity to Cell Interaction

A typical single-cell RNA-seq study can identify epidermal cells, cortex cells, endodermal cells, vascular cells, mesophyll cells, guard cells, pollen-related cell types, or developmental subpopulations. Marker genes and clustering can describe who is present.

Ligand-receptor analysis asks what these cell types may be saying to each other.

For example, a study may identify a stress-responsive cell population with elevated hormone response genes. A communication analysis can then ask whether the same cell population appears to send ABA-related, JA-related, SA-related, peptide-mediated, or immune-related signals to another cell type.

The output does not prove physical binding. It provides a ranked model of possible sender-receiver relationships supported by expression evidence and database annotations.

Signals That Shape Plant Development and Stress Response

Plant communication differs from many animal-focused ligand-receptor models. Plants use protein ligands and receptors, but they also rely heavily on phytohormones, mobile peptides, nutrient signals, ions, defense-related signals, and mobile RNA-related regulation.

Important signal categories may include:

• Auxin, cytokinin, gibberellin, ABA, ethylene, brassinosteroid, jasmonate, and salicylic acid pathways.
• Small peptide signals involved in meristem maintenance, organ development, vascular patterning, and defense.
• Pattern-triggered immunity signals such as flg22, elf18, and chitin-related recognition.
• Nutrient and ion-related signals such as nitrate, potassium, calcium, and magnesium.
• Receptor-like kinases and receptor-like proteins that act alone or with co-receptors.

Because plant signaling systems are diverse, a plant-specific ligand-receptor workflow should not treat every interaction as a simple "one ligand, one receptor" event.

Why Bulk Data Cannot Resolve Sender-Receiver Context

Bulk RNA-seq can show that a hormone response pathway is active in a tissue sample. It cannot usually identify which cell type is sending the signal, which cell type is receiving it, or whether the signal is localized to a tissue domain.

Single-cell omics adds cell identity. Spatial transcriptomics adds tissue position.

Together, they help researchers ask more specific questions:

• Which cell clusters express the ligand or signaling enzyme?
• Which clusters express the receptor, co-receptor, or downstream response genes?
• Are inferred source and target cells adjacent, regionally separated, or linked across a developmental axis?
• Does a stress condition change the direction, strength, or pathway composition of inferred communication?
• Are the same interactions observed across biological replicates?

These questions are central for plant developmental biology, stress biology, crop research, tissue mapping, and spatial transcriptomics data interpretation.

What Counts as a Plant Ligand-Receptor Pair

A plant ligand-receptor pair is not always a simple extracellular protein binding a single membrane receptor. In many plant systems, a biologically meaningful signaling event may involve a small molecule, a peptide, a receptor-like kinase, a receptor-like protein, a co-receptor, receptor-complex subunits, and pathway modulators.

For single-cell omics, the practical question is: which components can be detected or inferred from expression data, and which interactions can be mapped to a curated or computationally predicted database?

Protein Ligands and Receptor-Like Kinases

Many plant signaling pathways use secreted peptides or extracellular proteins as ligands. These ligands may bind receptor-like kinases, receptor-like proteins, or receptor complexes on target cells.

Examples of relevant plant receptor families include leucine-rich repeat receptor-like kinases and other membrane-associated receptors. These proteins are often involved in development, cell identity maintenance, immunity, and environmental response.

In a single-cell analysis, a protein ligand-receptor event may be represented as:

A strong analysis should retain these fields in the output table. Without them, interpretation becomes difficult because users cannot tell whether a ranked interaction is based on literature support, predicted pairing, a receptor complex, or a single-gene expression pattern.

Hormones and Small Signaling Molecules

Plant ligand-receptor analysis also needs to represent small signaling molecules. These include phytohormones such as ABA, auxin, cytokinin, ethylene, gibberellin, jasmonate, salicylic acid, and brassinosteroid-related signals.

Small-molecule signaling creates a modeling challenge. The molecule itself is not always measured directly in scRNA-seq or spatial transcriptomics. Instead, the analysis may use receptor genes, biosynthesis genes, transporters, downstream response genes, or curated signal-receptor relationships.

This is why hormone-related communication results should be interpreted carefully. A pathway-level score may suggest that a target cell type is transcriptionally equipped to receive a signal, but it does not directly quantify hormone concentration.

For deeper interpretation, researchers may combine ligand-receptor analysis with:

• Marker gene analysis.
• Hormone biosynthesis or response gene modules.
• Spatial domain annotation.
• Mutant or treatment comparisons.
• Targeted validation assays.
• Metabolite or hormone measurements when available.

Co-Receptors, Complexes, Agonists, and Antagonists

Plant signaling often depends on receptor complexes and cofactors. A receptor may not be sufficient on its own. A co-receptor may be required for recognition, or an antagonist may reduce pathway activity.

For this reason, plant cell-cell communication analysis should track more than a ligand and receptor column.

Useful output fields include:

• Interaction name.
• Signal or pathway name.
• Ligand or signaling molecule.
• Receptor.
• Complex subunits.
• Co-receptor or auxiliary receptor.
• Agonist.
• Antagonist.
• Evidence source.
• Interaction type.
• Source cell type.
• Target cell type.
• Communication score or probability.
• Statistical filtering result.

These fields help researchers distinguish a plausible biological signal from a loose gene-pair match.

Plant-Specific Databases and PlantCellChatDB

Many widely used cell-cell communication tools were first developed for human or mouse single-cell data. Those resources are valuable for animal systems, but they cannot be directly transferred to plants without careful database adaptation.

Plant signaling has different ligands, receptors, pathways, gene families, and interaction evidence. A plant analysis must account for plant-specific receptor-like kinases, phytohormones, peptide families, nutrient signals, pathogen-associated molecular pattern recognition, and species-specific gene identifiers.

Species and Signal Coverage

PlantCellChatDB is a useful example of a plant-focused communication resource. It is presented as an integrated knowledgebase and computational framework for exploring plant cell-cell communication, with ligand-receptor information across multiple plant species. The current public database page reports 12,712 ligand-receptor pairs across five species, including Arabidopsis thaliana and Oryza sativa.

The value of this type of resource is not only the number of interactions. It is the plant-specific structure of the database.

A plant-focused database can include:

• Interacting protein-based ligand-receptor pairs.
• Signaling molecule-receptor relationships.
• Hormone-related signaling categories.
• Receptor-complex information.
• Co-receptors.
• Agonists and antagonists.
• Evidence fields such as literature identifiers.
• Species-specific gene IDs.

For plant single-cell omics, this makes the database more biologically relevant than an animal-only ligand-receptor library.

Interaction Fields Researchers Should Understand

Before interpreting results, researchers should understand what each interaction field means.

A typical interaction table may include:

These fields should be kept in the final deliverables. They allow the researcher to trace each visual result back to its molecular basis.

When Manual Curation May Be Needed

Plant databases are improving, but coverage is not complete for every species, cultivar, tissue, or gene model version. Manual curation may be needed when working with non-model plants, polyploid crops, orphan genes, custom annotations, or species with incomplete receptor annotation.

Manual curation can include:

• Mapping orthologs from a model species.
• Converting gene IDs between genome versions.
• Removing low-confidence interactions.
• Adding project-specific ligand or receptor candidates.
• Separating experimentally supported interactions from predicted interactions.
• Reviewing pathway relevance with plant biology experts.

This step is especially important for crop species and specialized tissues. A database match is only as useful as the gene model, annotation, and biological context supporting it.

Input Data and QC Checks

Plant ligand-receptor analysis for single-cell omics begins with processed expression data, but the quality of the inference depends on several upstream choices. The analysis is sensitive to cell type annotation, detected gene depth, dropout, sample preparation, tissue dissociation effects, species mapping, and metadata quality.

A communication workflow should never start from an expression matrix alone without checking these factors.

Required Inputs for scRNA-seq

For single-cell RNA-seq or single-nucleus RNA-seq, the core inputs usually include:

• Gene-by-cell expression matrix.
• Cell barcode metadata.
• Cluster labels or cell type annotations.
• Marker gene tables.
• Sample and condition metadata.
• Species and reference genome version.
• Gene ID format.
• Batch or replicate information.
• Optional treatment, tissue zone, time point, or genotype metadata.

Cell type annotation is one of the most important inputs. If a cluster is mislabeled, the communication result will also be misleading. For example, a receptor expressed by a vascular-related cluster cannot be interpreted correctly if that cluster was incorrectly annotated as epidermal tissue.

For plant data, researchers should also consider whether the data came from protoplasts, nuclei, or spatially captured sections. Protoplasting may enrich or reduce some stress-response signals. Nuclei-based data may capture different RNA populations than whole-cell data. Spatial data may have mixed-cell spots or variable resolution depending on the platform.

Additional Inputs for Spatial Data

For spatial transcriptomics, the analysis can include tissue position.

Useful spatial inputs include:

• Spatial coordinate files.
• Histology or tissue image alignment.
• Spot, bin, or cell segmentation metadata.
• Tissue region annotation.
• Spatial domain labels.
• Matched single-cell or single-nucleus reference data.
• Region-of-interest information.
• Distance or neighborhood information when supported by the method.

Spatial coordinates help answer whether inferred communication is local, regional, or associated with tissue architecture. However, spatial data may also have lower gene detection per cell or spot than some single-cell RNA-seq datasets. This makes QC and interpretation especially important.

QC Metrics That Affect Communication Inference

QC should be reported as metric types rather than fixed universal thresholds. Plant tissues differ widely in cell wall structure, organelle content, RNA quality, cell size, and dissociation sensitivity.

Important QC metric types include:

• Detected genes per cell, nucleus, spot, bin, or segmented cell.
• Molecular count or read-depth distribution.
• Fraction of reads assigned to cells or tissue regions.
• Mitochondrial and chloroplast read fractions.
• Ambient RNA assessment when available.
• Doublet or multiplet estimates for single-cell data.
• Cluster marker consistency.
• Batch or replicate separation.
• Annotation confidence.
• Spatial tissue coverage and image alignment quality.
• Gene ID mapping success rate.
• Expression coverage of ligand and receptor genes.

For communication analysis, it is especially important to examine whether low-abundance ligands or receptors are consistently detected. A pathway may be missed because the relevant gene is not captured, not because the signal is biologically absent.

Analysis Workflow

Figure 2. A practical workflow for plant ligand-receptor analysis from annotated single-cell data to communication networks.

A plant communication workflow should be structured, traceable, and easy to audit. Each inferred interaction should be linked back to the input expression matrix, cell annotation, database entry, score, and visualization.

Step 1: Confirm Cell Types and Gene IDs

The workflow begins with annotated single-cell, single-nucleus, or spatial data.

The analysis team should confirm:

• Species.
• Genome build.
• Gene ID style.
• Cluster or cell type labels.
• Sample group labels.
• Replicates or biological conditions.
• Target tissue and developmental stage.
• Whether the study uses scRNA-seq, snRNA-seq, spatial transcriptomics, or integrated data.

Gene ID conversion is often needed. This is not a minor step. A mismatch between the expression matrix and the ligand-receptor database can lead to missing interactions.

For example, a dataset may use locus IDs, transcript IDs, gene symbols, or a custom annotation. The database may use another naming format. Mapping should be documented in the final report.

Step 2: Map Ligands, Receptors, and Cofactors

The next step is to map expressed genes to the plant ligand-receptor database.

The analysis should identify:

• Expressed ligands.
• Expressed receptors.
• Expressed co-receptors.
• Complex components.
• Signaling molecule-related receptors.
• Pathway categories.
• Evidence labels.

Lowly expressed genes may be filtered depending on method settings. This step should be documented because filtering can influence which interactions appear in the final network.

For plant hormone signaling, the mapping may be more complex than direct ligand mRNA detection. Some signals are small molecules rather than gene products. In those cases, the interpretation should focus on receptor expression, pathway assignment, downstream response context, or curated signaling relationships.

Step 3: Infer Source-Target Communication

After mapping, the method estimates possible communication between source and target cell groups. Many frameworks use expression-based scoring, probability models, permutation tests, or mass-action-inspired assumptions.

A source-target pair may be ranked higher when:

• The source group expresses the ligand or signal-related component.
• The target group expresses the receptor and required cofactors.
• The interaction is supported by a database entry.
• The expression pattern is specific to a cell type or condition.
• Statistical filtering supports the interaction.
• In spatial workflows, source and target cells or regions have relevant positional relationships.

The score is a model output. It is not a direct measurement of ligand concentration, receptor binding, or downstream phosphorylation.

Step 4: Visualize Networks and Pathways

Communication results are usually easier to interpret when shown at multiple levels.

Common visual outputs include:

• Source-target network diagrams.
• Pathway-level network plots.
• Ligand-receptor bubble plots.
• Heatmaps of communication strength.
• Bar plots of significant interactions by signal category.
• Tables of ranked ligand-receptor pairs.
• Condition comparison plots.
• Spatial overlays when spatial coordinates are available.
• UMAP or cluster maps linked to source and target cell types.

A useful visualization should not be decorative. It should answer a biological question: which cell type is sending, which is receiving, which pathway is involved, and whether the pattern changes across condition, tissue region, genotype, or developmental stage.

Planning a Communication Analysis Project

If you already have plant scRNA-seq, snRNA-seq, or spatial transcriptomics data, CD Genomics can review your species, gene ID format, annotation table, sample groups, and research goals to evaluate whether ligand-receptor analysis is appropriate for your dataset. This review can be paired with single-cell RNA-seq data analysis, spatial transcriptomics data analysis, or intercellular communication analysis services for research-use-only projects.

How to Interpret Results

Figure 3. Communication outputs should be interpreted as model-based hypotheses supported by expression and metadata.

Ligand-receptor analysis is powerful because it converts large expression matrices into interpretable communication hypotheses. It is also easy to overinterpret. A strong report should explain what the scores mean, what they do not mean, and how results should be prioritized.

Source, Target, and Directionality

The source cell type is the inferred sender. The target cell type is the inferred receiver.

Directionality is based on the model and database structure. If a ligand is expressed in cell type A and a receptor is expressed in cell type B, the interaction may be represented as A to B. If both cell types express both components, the network may show reciprocal communication.

In plant tissues, directionality should be interpreted with biology in mind. Some signals move through specific tissues or routes. Others may be local. Some hormone signals depend on synthesis, transport, perception, and response gradients.

A source-target arrow is therefore an inference, not a live image of molecular movement.

Communication Probability and Statistical Filters

Communication probability or interaction score is usually a model-based measure of inferred signal strength. A higher score suggests that, under the model assumptions, the source-target pair has stronger support from expression and interaction data.

P values or adjusted statistical filters may help remove weak or unstable signals. However, statistical significance does not automatically make an interaction biologically important. A low-frequency but tissue-critical interaction may be missed if expression is sparse. A broad housekeeping-like receptor expression pattern may generate a high number of interactions that need biological filtering.

Useful interpretation combines:

• Score magnitude.
• Statistical support.
• Cell type specificity.
• Pathway relevance.
• Replicate consistency.
• Known plant biology.
• Spatial position when available.
• Independent validation evidence.

Pathway-Level Networks and Biological Hypotheses

Pathway-level aggregation can make large tables easier to interpret. Instead of focusing on thousands of gene pairs, the analysis may summarize ABA-related, auxin-related, jasmonate-related, brassinosteroid-related, peptide-related, or immune-related communication.

This can help identify broad biological patterns.

For example:

• ABA-related communication may increase in stress-responsive tissue regions.
• Peptide-receptor interactions may concentrate around meristem or vascular cell types.
• Immune-related signals may become enriched after pathogen exposure.
• Auxin-related interactions may align with developmental gradients.
• Cytokinin-related patterns may differ between dividing and differentiating cell populations.

These results should be framed as hypotheses. Follow-up experiments may include reporter lines, targeted gene expression assays, mutant analysis, in situ hybridization, immunostaining when antibodies are available, or perturbation experiments.

Single-Cell vs Spatial Omics for Communication Analysis

Figure 4. Single-cell and spatial omics answer different but complementary questions in plant communication analysis.

Single-cell and spatial omics answer related but different questions. The best choice depends on whether the research question is mainly about cell identity, tissue position, or both.

When scRNA-seq Is Enough

Single-cell RNA-seq is often suitable when the main goal is to identify cell types, compare conditions, and infer communication between annotated clusters. It can provide strong cell-type resolution, especially when dissociation is feasible and cell populations are well captured.

scRNA-seq is useful for:

• Discovering cell populations.
• Building marker gene sets.
• Comparing treated and control conditions.
• Finding cell-type-specific ligand and receptor expression.
• Ranking source-target communication between clusters.
• Generating hypotheses for follow-up spatial or functional studies.

For plant tissues that are difficult to dissociate, single-nucleus RNA-seq may be considered. It can reduce some dissociation-related constraints, but it may capture a different transcript profile from whole-cell RNA-seq.

When Spatial Coordinates Matter

Spatial transcriptomics becomes important when tissue architecture is part of the question.

Spatial data can help evaluate whether source and target populations are located in the same tissue region, adjacent regions, or distinct domains. This is useful in roots, leaves, meristems, embryos, vascular tissues, reproductive tissues, and stress-treated regions where position is biologically meaningful.

Spatial communication analysis can consider:

• Tissue section position.
• Neighborhood relationships.
• Region-specific signaling.
• Distance-aware models.
• Spatially localized pathway activity.
• Communication patterns across tissue boundaries.

Not every spatial dataset has single-cell resolution. Some platforms capture mixed-cell spots or bins. In those cases, deconvolution or reference mapping may be needed before communication analysis.

Why Integrated Analysis Can Be Stronger

Integrated single-cell and spatial analysis can combine the strengths of both data types. Single-cell data can provide high-resolution cell identity. Spatial data can place those identities back into the tissue.

This is useful when a study needs to connect:

• Cell types with tissue zones.
• Communication pathways with spatial domains.
• Ligand-receptor scores with histology or anatomy.
• Stress-responsive cell states with local tissue context.
• Developmental gradients with source-target relationships.

For projects that need both cell identity and tissue architecture, integrated analysis of single-cell and spatial transcriptome data can help connect expression-based communication inference with spatial position.

Deliverables Researchers Should Expect

A plant ligand-receptor analysis project should produce more than a set of attractive network plots. The deliverables should allow researchers to audit, reuse, and explain the results.

Core Output Tables

Core tables may include:

• Filtered expression matrix or processed object summary.
• Cell or cluster annotation table.
• Gene ID mapping table.
• Ligand-receptor interaction table.
• Source-target communication table.
• Pathway-level communication table.
• Condition comparison table when multiple groups are analyzed.
• Evidence field for each interaction when available.
• Statistical filtering fields.
• Notes on removed or unmapped genes.

These tables are important for reproducibility. A figure may show the main pattern, but the table lets the researcher trace which ligand, receptor, co-receptor, or pathway produced the result.

Visualization Files

Visual deliverables may include:

• UMAP or clustering plots.
• Cell type marker plots.
• Source-target network diagrams.
• Ligand-receptor bubble plots.
• Pathway heatmaps.
• Bar charts of significant interactions.
• Chord or circle plots when appropriate.
• Spatial overlays for spatial transcriptomics data.
• Condition comparison plots.
• High-resolution figure files for reporting.

Plots should be designed to answer specific questions. Overly dense network diagrams can be hard to interpret if too many cell types or interactions are shown at once. Filtering and pathway-level summaries are often needed.

Interpretation and Reporting Notes

A useful report should explain:

• Which database or resource was used.
• How genes were mapped.
• Which interactions were filtered.
• How scores were calculated or ranked.
• Which pathways were highlighted.
• Which results were robust across replicates or groups.
• Which results need cautious interpretation.
• Which hypotheses may be suitable for validation.

This reporting layer is critical for CRO teams, core facilities, and project managers who need to pass results to plant biologists, breeders, or downstream experimental teams.

Limitations and Validation

Plant ligand-receptor analysis is an inference method. It is useful for prioritizing hypotheses, but it should not be treated as direct proof of molecular binding, signal movement, or functional pathway activation.

Why Inference Is Not Direct Proof

Most ligand-receptor workflows use transcript abundance as the starting point. Transcript presence does not always mean protein abundance, secretion, ligand availability, receptor localization, or pathway activation.

For small signaling molecules, the limitation is even stronger. Hormone concentration, transport, and metabolism are not measured directly in a standard scRNA-seq matrix.

Other common limitations include:

• Dropout of low-abundance genes.
• Missed ligands or receptors due to limited capture depth.
• Incorrect cluster annotation.
• Gene ID mapping failure.
• Incomplete database coverage.
• Batch effects.
• Species-specific gene model differences.
• Spatial resolution limits.
• Mixed-cell spots in some spatial platforms.
• Lack of protein-level or metabolite-level confirmation.

These limitations do not make the analysis invalid. They define how results should be used: as ranked, traceable, expression-supported communication hypotheses.

Species and Annotation Constraints

Model plants such as Arabidopsis and rice often have stronger annotation support than less-studied species. Tomato, maize, soybean, cotton, wheat, banana, and other crop species may require more careful gene mapping, orthology review, and manual curation.

Polyploid crops add another challenge. Homeologous genes may complicate read mapping, receptor annotation, and ligand-receptor matching. In these cases, the analysis should clearly document which gene IDs were retained, merged, mapped, or excluded.

For non-model species, researchers should ask:

• Is the reference genome suitable?
• Are gene models current?
• Are ligand and receptor homologs annotated?
• Can gene IDs be mapped to a plant communication database?
• Are target pathways conserved enough for interpretation?
• Are known marker genes available for cell type annotation?

If the answer is uncertain, a feasibility review should happen before full analysis.

Practical Validation Options

Validation depends on the study system and available tools.

Possible validation strategies include:

• Reporter lines for pathway activity.
• Mutant or overexpression studies.
• qPCR or targeted expression validation.
• RNA in situ hybridization.
• Spatial validation with targeted probes.
• Immunostaining when reliable antibodies are available.
• Hormone or metabolite assays.
• Perturbation experiments.
• Comparison across biological replicates.
• Cross-checking with published plant signaling literature.

A practical validation plan should focus on a small number of high-priority interactions. The best candidates are usually biologically relevant, cell-type-specific, supported by multiple output views, and consistent with known tissue organization or treatment response.

Project Planning Checklist

Before starting plant ligand-receptor analysis, researchers should prepare a clear data and metadata package. This reduces mapping errors and improves interpretability.

Information to Prepare Before Analysis

Useful project information includes:

• Plant species.
• Tissue or organ.
• Developmental stage.
• Treatment, genotype, condition, or time point.
• Data type: scRNA-seq, snRNA-seq, spatial transcriptomics, or integrated data.
• Raw or processed data status.
• Reference genome version.
• Gene ID format.
• Cell type annotation table.
• Marker gene table.
• Replicate and batch metadata.
• Spatial coordinates if available.
• Tissue image or region annotation if available.
• Target pathways or biological questions.
• Preferred output figure types.

For spatial omics solutions for plant research, this planning step is also useful for choosing whether the project should emphasize single-cell resolution, tissue position, integrated analysis, or downstream communication inference.

Questions to Ask Before Starting

A well-planned project should answer:

• Is the goal discovery, comparison, or validation support?
• Are cell type labels reliable enough for communication inference?
• Are biological replicates available?
• Are the target pathways represented in the database?
• Are gene IDs compatible with the selected resource?
• Is spatial position important to the research question?
• Are there known plant biology constraints for the tissue?
• What deliverables are needed for publication, internal review, or follow-up experiments?

These questions help align the analysis with the biological problem instead of producing a generic network report.

FAQ

What is plant ligand-receptor analysis for single-cell omics?

Plant ligand-receptor analysis for single-cell omics is a bioinformatics approach that infers possible communication between plant cell types. It combines expression data with a database of ligand-receptor, signal-receptor, co-receptor, and pathway relationships. The analysis identifies which cell groups may send signals and which cell groups may receive them. It is used to generate hypotheses about development, stress response, immunity, tissue organization, and spatial signaling.

What input data are needed for plant cell-cell communication analysis?

The core inputs include a gene expression matrix, cell or nucleus metadata, cell type annotation, marker gene information, sample group labels, species information, reference genome version, and gene ID format. For spatial transcriptomics, spatial coordinates, tissue images, segmentation or spot metadata, and tissue region labels may also be needed. Better metadata improves interpretation because communication scores depend on both expression and cell identity.

Can ligand-receptor analysis be performed on plant scRNA-seq data without spatial information?

Yes. Plant scRNA-seq data can support cell-cell communication inference when cell type annotation and gene mapping are reliable. The analysis can identify likely source-target relationships between cell clusters. However, without spatial coordinates, the result does not show whether the inferred sender and receiver cells are physically close in the tissue. Spatial transcriptomics is useful when location, tissue boundaries, or neighborhood relationships are central to the question.

When should spatial transcriptomics be added to plant communication analysis?

Spatial transcriptomics should be considered when tissue architecture matters. Examples include root zones, vascular regions, leaf layers, meristems, embryos, reproductive tissues, pathogen-infected regions, and localized stress responses. Spatial data can help determine whether inferred signaling is associated with specific tissue regions or neighboring cell populations. It can also support integrated analysis when paired with scRNA-seq or snRNA-seq reference data.

What QC metrics affect ligand-receptor inference in plant single-cell data?

Important QC metric types include detected genes per cell or nucleus, molecular count or read-depth distribution, organellar read fraction, ambient RNA assessment, doublet or multiplet estimates, batch effects, marker gene consistency, annotation confidence, and gene ID mapping success. For spatial data, tissue coverage, image alignment, segmentation quality, and spatial coordinate accuracy also matter. These metrics affect whether ligands and receptors can be detected and interpreted reliably.

How are communication probability scores interpreted?

A communication probability or score is a model-based estimate of inferred signaling strength between a source and target cell group. A higher score usually means stronger support from expression and database information under the method's assumptions. It does not directly prove ligand secretion, receptor binding, protein activity, or downstream pathway activation. Scores should be interpreted with statistical filters, biological context, replicate support, and validation options.

How is PlantCellChatDB different from animal ligand-receptor databases?

PlantCellChatDB is designed for plant cell-cell communication analysis. It includes plant-relevant interaction types such as signaling molecules, phytohormone-related pathways, receptor-like kinases, receptor complexes, co-receptors, agonists, and antagonists across multiple plant species. Animal databases are often centered on human or mouse ligand-receptor biology and may miss plant-specific signaling systems. For plant projects, database selection directly affects which interactions can be detected.

References

1. CellChat for systematic analysis of cell-cell communication from single-cell and spatially resolved transcriptomics. Nature Protocols, 2024.
2. Advancing spatial cellular communication inference with ligand diffusion and transport models. Communications Biology, 2025.
3. Recent progress in single-cell transcriptomic studies in plants. Cellular and Molecular Biology Letters, 2025.
4. Harnessing Single-Cell and Spatial Transcriptomics for Crop Improvement. Plants, 2024.
5. Single-cell and spatial transcriptomics in plants: from cell states to coordinated tissue function. Journal of Experimental Botany, 2026.