CellsigN

Single-cell RNA sequencing provides unprecedented opportunities to investigate cellular communication. Numerous computational methods have been developed to infer ligand–receptor interactions and downstream transcription factor (TF)–target gene（TG） regulation. However, existing approaches remain limited in reconstructing the detailed signaling cascades that bridge receptors and TFs. Here, we present CellSigN, a computational framework that reconstructs complete transcellular signaling pathways from single-cell RNA sequencing data. CellSigN integrates ligand–receptor interactions, TF–TG regulation, and six types of intercellular signaling events—chemical catalysis, transport regulation, state change control, physical interactions, protein complex formation, and phosphorylation. Applied to eight cancer datasets, CellSigN achieved an average of 16.13% improvement in recall for seven categories of cancer-related interactions compared with state-of-the-art methods. A breast cancer case study further demonstrated CellSigN’s ability to uncover disease mechanisms, highlighting its potential for biomarker discovery and therapeutic target development.