概要
This skill enhances the MaxFuse framework by introducing spatial context into the cell-matching process, specifically designed for complex tissue environments like the spleen. By accounting for distinct tissue regions—such as B cell follicles, T cell zones, and red pulp—it prevents biologically implausible matches where cells from scRNAseq are incorrectly paired with incompatible CODEX regions. It utilizes a three-pronged approach involving prior-weighted distance interpolation, neighborhood-augmented features, and DBSCAN-based region detection to ensure high-fidelity multi-omic integration while maintaining the spatial architectural integrity of the sample.
