ConvexGating is learning cell-sorting strategies in flow cytometry automatically and interpretably with artificial intelligence. While the method was experimentally validated in CD8+ subtypes and adipose tissue progenitor cells, it was also transferable to Cytometry by Time of Flight and Cellular Indexing of Transcriptomes and Epitopes by Sequencing data.
Why it matters
The importance of the study lies not only in automating the gating process that determines the criteria by which cells are separated in flow cytometry, but also in making this process interpretable. This feature makes it possible to assess how strategies obtained from different samples, particularly CD8+ subtypes and adipose tissue progenitor cells, were developed. The experimental validation of the method shows that the approach is not limited to the computational level, while its transferability to different types of single-cell data demonstrates that its application is not restricted to a single measurement platform. Nevertheless, the fact that it can be adapted to different data types leaves open the question of whether the same results and ease of interpretation can be achieved on every platform.