Cell type purification by single-cell transcriptome-trained sorting

Traditional cell type enrichment using fluorescence activated cell sorting (FACS) relies on methods that specifically label the cell type of interest. Here we propose GateID, a computational method that combines single-cell transcriptomics, for unbiased cell type identification, with FACS index sorting, to purify cell types of choice. We validate GateID by purifying various cell types from the zebrafish kidney marrow and the human pancreas without resorting to specific antibodies or transgenes. Overall design: Single cell mRNA sequencing data of unenriched and enriched cells from zebrafish whole kidney marrow and human pancreas using GateID. For each cell, we provide libraries with transcriptome information. An unenriched sample is transcriptome data from a live single cell FACS sort, while an enriched library has transcriptome data for GateID enriched single cells for a given cell type.