Single-cell sequencing data can identify a person

An individual’s genome is identifiable. But datasets of RNA sequences from single cells are like a blurry photocopy of the genome. The sequence numbers are small, and the data are noisy and represent only the subset of genes expressed in a given cell. That’s why “there was a lot of debate” over whether one could identify a person based on data from a single-cell RNA sequencing experiment, says Columbia University bioinformatics professor Gamze Gürsoy. Researchers treat single-cell sequencing as biochemically anonymous and share datasets publicly. The technique catalogs RNA from individual cells to detect differences between cell types; as it has become more affordable, studies with hundreds or thousands of research participants have become available online. Gürsoy and her lab found that they could robustly link single-cell transcriptomes with known genomes using two such datasets from studies of autoimmune disorders ( Cell 2024, DOI: 10.1016/j.cell.2024.09.012 ). The researchers used statistical