Table 2.

Existing tools for measuring cell–cell communication

Tool	Method overview	Output	Visualization	Available in	URL	Refs
Differential combinations
CellTalker	Uses differentially expressed ligands and receptors in each cluster to identify unique interactions between clusters	Upregulated and downregulated interactions between all clusters	Circos plot of differential interactions between clusters	R	https://github.com/arc85/celltalker	61
iTALK	Enumerates differentially expressed ligand and receptor values to identify LRIs between different clusters	Upregulated and downregulated interactions between all clusters	CCI networks, Circos plots and boxplots	R	https://github.com/Coolgenome/iTALK	106
PyMINEr	Uses differentially expressed ligand and receptor pairs to identify altered signalling pathways. Detects both activation and inhibition	Upregulated and downregulated interactions between all clusters	Network visualization and Circos plots	Python and standalone application	https://www.sciencescott.com/pyminer	107
Expression permutation
CellChat	Modelling of LRI is generalized from the Hill equation, including expression of both agonists and antagonists. Significance is computed by permutation	Likelihood of CCC between all clusters for all interactions	Alluvial and Circos plots of communication pathways, dot plots of interactions between clusters	R and Web interface	https://github.com/sqjin/CellChat	79
CellPhoneDB	Randomly permutes cluster labels to generate a null distribution of LRI scores using protein complex subunit architecture to identify significant interactions	Upregulated and downregulated interactions between all clusters	Heatmap of significant interaction counts, dot plot of LRIs and cluster combinations	Python and Web interface	https://github.com/Teichlab/cellphonedb	30,83
Giotto	Randomly permutes cluster labels to generate a null distribution of LRI scores using spatial information to identify significant interactions	Upregulated and downregulated interactions between all clusters	Heatmap of significant interaction counts, dot plot of LRIs and cluster combinations	R	https://github.com/RubD/Giotto	111
ICELLNET	Sums the product of all LRI scores between two clusters to compute an overall CCI score	Intergroup communication scores, P values for these scores and LRI scores	Stacked bar plot of LRIs, network visualization of interacting groups and pathway-level analysis	R	https://github.com/soumelis-lab/ICELLNET	84
SingleCellSignalR	Uses a regularized ligand–receptor expression product to measure extent of CCC	Interaction scores for each LRI between all clusters in the dataset	Circos plot, tables and graph visualizations of interactions between clusters	R	https://github.com/SCA-IRCM	112
Graph or network
CCCExplorer	A graph of all signalling pathways is built, then a Fisher-like statistic is computed using ligand, receptor and downstream TF expression to identify significant interactions	Graph visualizations of all interactions	Interactive directed graphs	Standalone application	https://github.com/methodistsmab/CCCExplorer	67
NicheNet	A network of ligand–receptor pathway interactions is used to measure the predictive power of the ligand for its downstream pathway targets as an interaction score, based on a personalized PageRank algorithm	Ligand interaction scores and expressing cell types for provided target pathway	Circos plot of interactions between cells or clusters	R	https://github.com/saeyslab/nichenetr	54
SoptSC	Integrates downstream signalling measurements into an LRI scoring function	Upregulated and downregulated interactions between all clusters	Circos plot of interactions between cells	MATLAB and R	https://github.com/WangShuxiong/SoptSC https://github.com/mkarikom/RSoptSC	25
SpaOTsc	An optimal transport model is used to infer CCC from ligand, receptor and downstream component expression	Likelihood of CCC between all clusters for all LRIs	Not implemented	Python	https://github.com/zcang/SpaOTsc	108
Tensor based
scTensor	Tucker decomposition on a tensor of order three to identify key LRIs present in certain cell types	HTML file with summaries of clustering, decomposition and interaction components	Many options for interaction, expression and pattern visualization	R	https://github.com/rikenbit/scTensor	113

CCC, cell–cell communication; CCI, cell–cell interaction; LRI, ligand–receptor interaction; TF, transcription factor.