0
0This paper introduce SSpMosaic, a novel computational framework designed to unify single-cell and spatial omics data through the use of interpretable metaprograms. By extracting and aligning high-order gene programs, the tool facilitates the integration of diverse datasets across different batches, species, and molecular modalities like transcriptomics and proteomics. This methodology enhances cell-type annotation and allows for high-resolution spatial deconvolution, mapping cellular distributions from broad tissue areas down to subcellular scales. A key innovation of the framework is its ability to perform reference-free characterization, identifying conserved tissue structures without the need for matched single-cell references. Comparative benchmarks demonstrate that SSpMosaic excels in maintaining biological fidelity while effectively correcting for technical noise and batch effects. Ultimately, this technology provides a more transparent and scalable approach to understanding the functional architecture of complex tissues.
References:
Zhang Y, Ming W, Yu B, et al. Robust integration and annotation of single-cell and spatial omics data using interpretable gene programs[J]. Cell Genomics, 2026, 6(4).
