SpaCCI key features
- Integration of Multiple Ligand-Receptor Databases: SpaCCI leverages widely used ligand-receptor interaction databases, including CellChat, CellPhoneDB, Cellinker, ICELLNET, and ConnectomeDB, for both human and mouse species. This ensures comprehensive coverage and accuracy in identifying potential cell-cell interactions.
- Spatial Awareness: Unlike traditional methods, SpaCCI incorporates spatial information from ST platforms like 10x Visium. This allows for more accurate detection of cell-cell interactions by considering the physical proximity of cells and the spatial distribution of cell types across tissue samples.
- Cell Type Mixture Modeling: SpaCCI acknowledges the mixture of cell types within each spot of the tissue sample. It uses non-negative least square regression combined with cell type proportion data to estimate cell type-specific gene expression, offering a more accurate quantification of interactions.
- Global, Regional, and Local Interaction Quantification: The framework is designed to analyze cell-cell interactions at local, regional, and global scales. By borrowing information from neighboring spots, SpaCCI provides a detailed view of interaction patterns within and across different regions of the tissue.