Achim, K. et al. Nat. Biotechnol. 33, 503–509 (2015).

Satija, R. et al. Nat. Biotechnol. 33, 495–502 (2015).

Single-cell RNA-seq allows tissues to be deconstructed into their component parts. Two new methods set out to 'reconstruct' tissues by projecting profiled cell types back onto the organismal map, leveraging spatial information for a small set of landmark genes from RNA in situ hybridization data. Satija et al. generate a discretized map of landmark gene expression and develop the computational tool Seurat, which cleans up RNA-seq signal on the basis of gene coexpression and assigns cells to the discretized map according to a statistical model of landmark gene expression. They sequence zebrafish blastula cells and map even rare cell types. Achim et al. offer a complementary approach that requires in situ data for only a small number of highly expressed genes to map single cells on the basis of RNA-seq data. The authors use the method to map cells from the brain of a marine annelid.