Chapter 11: Cell-Cell Communication & Ligand-Receptor Networks

Chapter 11: Cell-Cell Communication & Ligand-Receptor Networks

No cell operates in biological isolation. Tissues, organs, and tumors are coordinated by intercellular signaling networks where sender cells secrete ligand proteins that bind to cognate cell-surface receptors on target cells, activating downstream signaling pathways.

1. Intercellular Communication with CellChat in R

CellChat models communication probability using the law of mass action, incorporating multimeric ligand-receptor complexes (e.g. TGF-\(eta\) heterotetramers) along with soluble cofactors and inhibitors.

library(CellChat)
library(patchwork)

# Create CellChat object from Seurat
cellchat <- createCellChat(object = pbmc, group.by = "cell_type", assay = "RNA")

# Load CellChatDB (Human or Mouse)
CellChatDB <- CellChatDB.human
# Filter to specific interaction categories (Secreted Signaling, ECM-Receptor, Cell-Cell Contact)
cellchat@DB <- CellChatDB

# Preprocess expression data and identify overexpressed signaling genes
cellchat <- subsetData(cellchat)
cellchat <- identifyOverExpressedGenes(cellchat)
cellchat <- identifyOverExpressedInteractions(cellchat)

# Calculate communication probabilities using mass action modeling
cellchat <- computeCommunProb(cellchat, raw.use = TRUE)
cellchat <- filterCommunication(cellchat, min.cells = 10)
cellchat <- computeCommunProbPathway(cellchat)
cellchat <- aggregateNet(cellchat)

2. Visualizing Signaling Cross-Talk

# 1. Circle plot: Aggregate number of interactions between cell types
par(mfrow = c(1,2))
netVisual_circle(cellchat@net$count, vertex.weight = groupSize, weight.scale = T, label.edge= F, title.name = "Number of interactions")
netVisual_circle(cellchat@net$weight, vertex.weight = groupSize, weight.scale = T, label.edge= F, title.name = "Interaction weights/strength")

# 2. Hierarchy and Circle plot for specific pathways (e.g. VEGF or CXCL)
pathways.show <- c("CXCL")
netVisual_aggregate(cellchat, signaling = pathways.show, layout = "circle")

# 3. Bubble plot showing specific ligand-receptor pairs
netVisual_bubble(cellchat, sources.use = c("Monocytes"), targets.use = c("CD4 T cells", "CD8 T cells"), remove.isolate = FALSE)

3. CellPhoneDB Permutation Testing in Python

CellPhoneDB calculates the mean expression of ligands and receptors across interacting cell pairs and assesses statistical significance using 1,000 empirical label permutations.

# Run CellPhoneDB statistical analysis from terminal
cellphonedb method statistical_analysis \
    meta_table.tsv \
    counts_normalized.tsv \
    --counts-data=gene_name \
    --threads=16 \
    --output-path=./cellphonedb_results/

# Outputs:
#   ├── pvalues.txt               -> Significance of each ligand-receptor pair
#   ├── means.txt                 -> Mean expression level of interactions
#   └── significant_means.txt     -> Filtered significant interactions