Mapping transcriptionally equivalent populations across single cell RNA-seq datasets

Comparison of single cell RNA sequencing (scRNA-seq) data across individuals, tissues and perturbations is important but challenging due to technical differences in sample processing, number of cells, read depth and technical variation. To address these challenges we developed scID (single cell identification), which uses the linear discriminant analysis framework to identify transcriptionally related cells across scRNA-seq datasets. Through extensive characterization of published data we demonstrate that despite differences in the sequencing depth, cell coverage, tissue composition and technology used, scID outperforms existing approaches for identifying equivalent cell populations across datasets.