One of my first experiences of analysing biomedical data, following my training in applied statistics, was to perform differential gene expression analysis of microarray data. Nowadays, microarray technologies have been largely superseded by three major biotechnological advances: RNA sequencing (RNA-seq), single-cell RNA sequencing (scRNA-seq) and spatial ‘omics’. The 2020s has witnessed a massive rise in the take-up of spatial omics technologies, particularly at subcellular resolution, but what biotechnological and methodological advances were necessary for this spatial omics revolution to occur?

One key enabling approach — combinatorial labelling of probes to capture the presence of RNAs in situ in single cells — was described by Levsky et al. in 2002. In this study, the authors used their new approach of ‘single-cell gene expression profiling’ to assess the transcription of 10–11 genes in approximately 2,500 single-cell nuclei from cell lines in situ. The use of combinations of fluorescent colours for labelling each gene was key to overcoming the limit of 3–4 optical colour channels in standard fluorescence in situ hybridization. By representing each gene as a barcoded combination of four fluorescent colours positive for at least two colours, the authors were able to assign individual ‘pseudo-colours’ to each gene being profiled.