home / skills / gptomics / bioskills / cell-communication

cell-communication skill

safe

This skill helps infer and compare cell-cell communication networks from scRNA-seq data using LIANA, CellChat, and NicheNet to identify ligand-receptor

npx playbooks add skill gptomics/bioskills --skill cell-communication

Review the files below or copy the command above to add this skill to your agents.

Files (4)

SKILL.md

8.4 KB

---
name: bio-single-cell-cell-communication
description: Infer cell-cell communication networks from scRNA-seq data using CellChat, NicheNet, and LIANA for ligand-receptor interaction analysis. Use when inferring ligand-receptor interactions between cell types.
tool_type: mixed
primary_tool: CellChat
---

## Version Compatibility

Reference examples tested with: ggplot2 3.5+, scanpy 1.10+

Before using code patterns, verify installed versions match. If versions differ:
- Python: `pip show <package>` then `help(module.function)` to check signatures
- R: `packageVersion('<pkg>')` then `?function_name` to verify parameters

If code throws ImportError, AttributeError, or TypeError, introspect the installed
package and adapt the example to match the actual API rather than retrying.

# Cell-Cell Communication Analysis

**"Infer cell-cell communication from my scRNA-seq data"** → Predict ligand-receptor interactions between cell types and visualize intercellular signaling networks.
- R: `CellChat::createCellChat()` → `computeCommunProb()` → `netAnalysis()`
- Python: `liana.method.cellchat()` (LIANA framework)

## CellChat (R)

**Goal:** Infer and quantify intercellular communication networks from scRNA-seq data using curated ligand-receptor databases.

**Approach:** Create a CellChat object from a Seurat object with cell type labels, select a signaling database subset, identify overexpressed ligands/receptors, compute communication probabilities using the trimean method, then aggregate into pathway-level networks.

```r
library(CellChat)
library(Seurat)

# Create CellChat object from Seurat
cellchat <- createCellChat(object = seurat_obj, group.by = 'cell_type')

# Set ligand-receptor database
CellChatDB <- CellChatDB.human  # or CellChatDB.mouse
cellchat@DB <- CellChatDB

# Subset to secreted signaling (optional)
CellChatDB.use <- subsetDB(CellChatDB, search = 'Secreted Signaling')
cellchat@DB <- CellChatDB.use

# Preprocessing
cellchat <- subsetData(cellchat)
cellchat <- identifyOverExpressedGenes(cellchat)
cellchat <- identifyOverExpressedInteractions(cellchat)

# Compute communication probability
cellchat <- computeCommunProb(cellchat, type = 'triMean')
cellchat <- filterCommunication(cellchat, min.cells = 10)

# Infer signaling pathways
cellchat <- computeCommunProbPathway(cellchat)
cellchat <- aggregateNet(cellchat)
```

## CellChat Visualization

```r
# Network plots
netVisual_circle(cellchat@net$count, vertex.weight = groupSize, weight.scale = TRUE,
                 label.edge = FALSE, title.name = 'Number of interactions')

netVisual_circle(cellchat@net$weight, vertex.weight = groupSize, weight.scale = TRUE,
                 label.edge = FALSE, title.name = 'Interaction strength')

# Heatmap of interactions
netVisual_heatmap(cellchat, color.heatmap = 'Reds')

# Specific pathway visualization
netVisual_aggregate(cellchat, signaling = 'WNT', layout = 'circle')
netVisual_aggregate(cellchat, signaling = 'WNT', layout = 'chord')

# Bubble plot
netVisual_bubble(cellchat, sources.use = c(1, 2), targets.use = c(3, 4),
                 remove.isolate = FALSE)

# Chord diagram for ligand-receptor pairs
netVisual_chord_gene(cellchat, sources.use = 1, targets.use = c(2, 3, 4),
                     lab.cex = 0.5, legend.pos.x = 10)
```

## CellChat Pathway Analysis

```r
# Identify signaling roles
cellchat <- netAnalysis_computeCentrality(cellchat, slot.name = 'netP')

# Signaling role heatmap
netAnalysis_signalingRole_heatmap(cellchat, signaling = c('WNT', 'TGFb', 'BMP'))

# Dominant senders/receivers
netAnalysis_signalingRole_scatter(cellchat)

# Compare pathways
rankNet(cellchat, mode = 'comparison', stacked = TRUE, do.stat = TRUE)
```

## Compare Conditions (CellChat)

```r
# Create separate CellChat objects
cellchat_ctrl <- createCellChat(subset(seurat_obj, condition == 'control'), group.by = 'cell_type')
cellchat_treat <- createCellChat(subset(seurat_obj, condition == 'treatment'), group.by = 'cell_type')

# Process both (same steps as above)
# ...

# Merge for comparison
cellchat_list <- list(Control = cellchat_ctrl, Treatment = cellchat_treat)
cellchat_merged <- mergeCellChat(cellchat_list, add.names = names(cellchat_list))

# Compare interactions
compareInteractions(cellchat_merged, show.legend = FALSE)

# Differential interactions
netVisual_diffInteraction(cellchat_merged, weight.scale = TRUE)
netVisual_heatmap(cellchat_merged)

# Pathway comparison
rankNet(cellchat_merged, mode = 'comparison', stacked = TRUE)
```

## NicheNet (R)

```r
library(nichenetr)
library(Seurat)
library(tidyverse)

# Load NicheNet databases
ligand_target_matrix <- readRDS('ligand_target_matrix.rds')
lr_network <- readRDS('lr_network.rds')
weighted_networks <- readRDS('weighted_networks.rds')

# Define sender and receiver cells
sender_celltypes <- c('Macrophage', 'Dendritic')
receiver <- 'T_cell'

# Get expressed genes
expressed_genes_sender <- get_expressed_genes(sender_celltypes, seurat_obj, pct = 0.10)
expressed_genes_receiver <- get_expressed_genes(receiver, seurat_obj, pct = 0.10)

# Define gene set of interest (e.g., DE genes in receiver)
geneset_oi <- FindMarkers(seurat_obj, ident.1 = 'activated_T', ident.2 = 'naive_T') %>%
    filter(p_val_adj < 0.05, avg_log2FC > 0.5) %>% rownames()

# Background genes
background_genes <- expressed_genes_receiver

# Define potential ligands
ligands <- lr_network %>% pull(from) %>% unique()
expressed_ligands <- intersect(ligands, expressed_genes_sender)

receptors <- lr_network %>% pull(to) %>% unique()
expressed_receptors <- intersect(receptors, expressed_genes_receiver)

potential_ligands <- lr_network %>%
    filter(from %in% expressed_ligands & to %in% expressed_receptors) %>%
    pull(from) %>% unique()

# NicheNet ligand activity analysis
ligand_activities <- predict_ligand_activities(
    geneset = geneset_oi,
    background_expressed_genes = background_genes,
    ligand_target_matrix = ligand_target_matrix,
    potential_ligands = potential_ligands
)

# Top ligands
best_ligands <- ligand_activities %>% top_n(20, pearson) %>% arrange(-pearson) %>% pull(test_ligand)
```

## NicheNet Visualization

```r
# Ligand-target heatmap
active_ligand_target_links <- best_ligands %>%
    lapply(get_weighted_ligand_target_links, geneset_oi, ligand_target_matrix, n = 200) %>%
    bind_rows() %>% drop_na()

vis_ligand_target <- prepare_ligand_target_visualization(
    ligand_target_df = active_ligand_target_links,
    ligand_target_matrix = ligand_target_matrix,
    cutoff = 0.33
)

p_ligand_target <- vis_ligand_target %>%
    make_heatmap_ggplot('Prioritized ligands', 'Target genes',
                        color = 'purple', legend_position = 'top')

# Ligand-receptor heatmap
lr_network_top <- lr_network %>%
    filter(from %in% best_ligands & to %in% expressed_receptors) %>%
    distinct(from, to)

vis_ligand_receptor <- get_exprs_avg(seurat_obj, 'cell_type') %>%
    inner_join(lr_network_top, by = c('gene' = 'to'))

p_ligand_receptor <- vis_ligand_receptor %>%
    make_heatmap_ggplot('Ligands', 'Receptors', color = 'mediumvioletred')

# Ligand expression by cell type
p_ligand_expression <- DotPlot(seurat_obj, features = best_ligands, cols = 'RdYlBu') +
    RotatedAxis()
```

## LIANA (Python)

```python
import liana as li
import scanpy as sc

adata = sc.read_h5ad('adata.h5ad')

# Run LIANA with multiple methods
li.mt.rank_aggregate(adata, groupby='cell_type', resource_name='consensus',
                     expr_prop=0.1, verbose=True)

# Get results
liana_results = adata.uns['liana_res']

# Filter significant interactions
sig_interactions = liana_results[liana_results['liana_rank'] < 0.01]

# Visualize
li.pl.dotplot(adata, colour='magnitude_rank', size='specificity_rank',
              source_groups=['Macrophage'], target_groups=['T_cell'])
```

## LIANA with Tensor Decomposition

```python
# Multi-sample/condition analysis
li.mt.rank_aggregate(adata, groupby='cell_type', resource_name='consensus',
                     use_raw=False, verbose=True)

# Build tensor for decomposition
li.multi.build_tensor(adata, sample_key='sample', groupby='cell_type',
                      ligand_key='ligand_complex', receptor_key='receptor_complex')

# Run tensor decomposition
li.multi.decompose_tensor(adata, n_components=5)

# Visualize factor loadings
li.pl.factor_loadings(adata, factor_idx=0)
```

## Related Skills

- single-cell/clustering - Define cell types first
- single-cell/trajectory-inference - Communication along trajectory
- spatial-transcriptomics/spatial-communication - Spatial context
- pathway-analysis/go-enrichment - Pathway enrichment of targets

Overview

This skill infers cell–cell communication networks from single-cell RNA-seq data using CellChat, NicheNet, and LIANA. It provides end-to-end patterns for building ligand–receptor interaction maps, scoring ligand activities, and comparing conditions or samples. Use it to prioritize signaling pathways, visualize sender–receiver relationships, and identify candidate ligands driving target gene programs.

How this skill works

The skill shows canonical workflows for three tools: CellChat (R) to create CellChat objects, compute communication probabilities, and aggregate pathway networks; NicheNet (R) to predict ligand activities and link ligands to target genes; and LIANA (Python) to run multiple LR methods, aggregate ranks, and perform multi-sample tensor decomposition. It covers preprocessing steps, expression filters, selection of expressed ligands/receptors, statistical filtering, and common visualization types (circle, heatmap, chord, bubble, dotplot).

When to use it

You have annotated cell types or clusters and want intercellular signaling maps.
You need to prioritize ligands that explain differential expression in a receiving cell population.
You want to compare communication networks between conditions, timepoints, or samples.
You plan to run multiple LR methods and aggregate consensus predictions.
You want pathway-level summaries and role analysis of senders/receivers.

Best practices

Verify package versions and adapt example calls to the installed API signatures before running.
Filter genes by expression proportion and minimum cell counts to reduce false positives.
Use the same preprocessing and grouping strategy across conditions when comparing networks.
Combine complementary tools: use NicheNet for ligand→target prioritization and CellChat/LIANA for network topology.
Visualize both edge-level interactions and aggregated pathway-level summaries to interpret results.

Example use cases

Infer macrophage→T cell signaling and identify ligands driving T cell activation gene programs with NicheNet.
Build CellChat networks from a Seurat object, visualize dominant senders/receivers, and compare control vs treatment.
Run LIANA consensus aggregation across multiple datasets, build a communication tensor, and decompose to find condition-specific factors.
Generate ligand–receptor heatmaps and dotplots to prioritize candidate receptor targets for functional validation.
Use CellChat pathway aggregation to inspect changes in WNT or TGFb signaling across developmental stages.

FAQ

Do I need cell type labels to run these workflows?

Yes. Accurate cell type or cluster labels are required for grouping, sender/receiver definitions, and interpretable network summaries.

Which tool should I pick for ligand prioritization?

Use NicheNet to link ligands to target gene programs. Use CellChat or LIANA to map network topology and aggregate predictions from multiple methods.