Cao, J. et al. Science 361, 1380–1385 (2018).

It has recently become possible to interrogate multiple molecular classes from an individual cell, yet this achievement is typically low throughput. Cao et al. now use single-cell combinatorial indexing to jointly profile gene expression and chromatin accessibility from large numbers of cells. In their sci-CAR method, pooled nuclei are distributed in wells, and well-specific barcodes are incorporated in first-strand cDNA and delivered to genomic regions by Tn5 transposition. Mixing and redistribution of intact nuclei allows a new well-specific barcode to be added to second-strand cDNA and genomic fragments during amplification, such that all molecules receive the unique barcode combination corresponding to their nucleus of origin. The authors sequenced thousands of cells from mouse kidney and three time points of a dexamethasone-treated cancer cell line, and were able to classify cell types, observe accessibility and expression dynamics, and link regulatory regions with target genes.