RankMap: rank-based reference mapping for fast and robust cell type annotation in spatial and single-cell transcriptomics

Cell type annotation is a key step in single-cell and spatial transcriptomics analysis, but many reference-based methods are computationally intensive, limiting their use on large spatial datasets. Here, we present RankMap ( https://github.com/jinming-cheng/RankMap ), a fast and flexible R package for reference-based cell type annotation. RankMap represents each cell using rank-based features derived from its top-expressed genes, reducing sensitivity to platform-specific differences in expression scale. It then applies elastic net-regularized multinomial regression to predict cell types with associated confidence scores. We benchmarked RankMap on five spatial datasets (Xenium, MERFISH, Stereo-seq) and two single-cell datasets, comparing it with SingleR , Azimuth , and RCTD . Across these settings, RankMap achieved competitive accuracy while substantially reducing runtime, particularly for large spatial datasets. Overall, RankMap provides a practical and scalable solution for reference-based cell type annotation.