This paper introduces Scriabin, a novel computational framework designed for the comparative analysis of cell-cell communication (CCC) at single-cell resolution using single-cell RNA sequencing (scRNA-seq) data. The authors explain that traditional methods often aggregate cell data, which can obscure important biological details, and detail how Scriabin overcomes this by analyzing interactions at the individual cell level without aggregation or downsampling. The framework offers three main workflows—the cell-cell interaction matrix (CCIM), summarized interaction graph, and interaction program discovery—to handle various dataset sizes and analytical goals. Through multiple published datasets and experimental validations, the source demonstrates that Scriabin accurately identifies CCC networks that are often missed by agglomerative techniques, showcasing its utility in contexts like T cell exhaustion, developmental biology, and longitudinal infection studies. Furthermore, the discussion addresses the challenges of data sparsity and scalability in single-cell CCC analysis, offering algorithmic solutions and future directions for improving the method.

References:

• Wilk A J, Shalek A K, Holmes S, et al. Comparative analysis of cell–cell communication at single-cell resolution[J]. Nature Biotechnology, 2024, 42(3): 470-483.