Scientists at the University of Edinburgh have developed a new method which will revolutionise the way we understand and categorise single cells. Dr Ava Khamseh Traditionally, analysis of individual cells has been problematic because of the ways they are clustered, which disallows many states of cells to be discovered. This is because clustering allows cells to be labelled once, mainly by type, only.A new method, called Stator, finely resolves states and subtypes without clustering, even among apparently equivalent cells. It works by finding high-order gene expression dependencies in sparse single cell sequencing data.The method itself is driven by the data – not by subjective choices – and can find up to seven genes with dependent expression or non-expression in the same cells. Led by Professor Chris Ponting and Dr Ava Khamseh at the Institute of Genetics and Cancer, the study highlights how in the first use cases, Stator discovered sub-phases of the cell cycle, the future fate of mouse embryonic cells, rare cell sub-types and cycling transformed cells whose expression signatures are prognostic of cancer. With Stator, no more does a cell need to be placed at a single location in expression space. It now allows many cell types, sub-types and states to be discovered, even among relatively few cells. Dr Ava Khamseh Lecturer in Biomedical AI Link to paper in Molecular Systems Biology. The paper is published by . Tags 2025 Publication date 03 Jan, 2025
