HyperCom: a hypergraph-based method to infer cell-resolved cell-cell communication

Abstract

Cell-cell communication underpins intercellular responses allowing cells to dynamically change states across conditions; however current methods typically aggregate signals by cell type annotations which may introduce biases. We developed HyperCom, a hypergraph-based method to infer cell-resolved cell-cell communication from single cell transcriptomics data. HyperCom constructs a hypergraph from known ligand-receptor interactions and the expression of those ligands and receptors in individual cells and performs graph diffusion on the hypergraph to score cells on a continuous spectrum from sender to receiver for each LR pair. HyperCom then generates a cell-type agnostic ranking of LR pairs by communication either over the whole dataset or in a condition-aware manner. Validation of HyperCom on physical cell-to-cell contact, spatial, and multicondition datasets highlight its flexibility to capture subtle per cell communication information. HyperCom is available as a species-agnostic, scalable, and open source R package: https://github.com/pawelab/HyperCom.