To establish a dimensionality reduction and clustering method of flow cytometric data for measuring T cell subpopulations in peripheral blood and ascites of patients with spontaneous bacterial peritonitis (SBP), and to explore the differences in the expression of markers of T lymphocytes in the peripheral blood and ascites.

Total of 4 patients with SBP were recruited in Beijing Ditan Hospital, Capital Medical University from December 2021 to January 2022. The monoclonal antibodies targeting cell surface antigen were used to establish the multi-color flow cytometry analysis panel. The peripheral blood mononuclear cell (PBMC) sample from one healthy donor was applied to adjust optimal detection voltage and fluorescence compensation. This panel was applied to detect peripheral blood and ascites samples from 4 patients with SBP. Uniform manifold approximation and projection (UMAP) and FlowSOM were used to analyze the results from multicolor flow cytometric data.

Using UMAP combined with FlowSOM, CD4⁺ T cells and CD8⁺ T cells were both divided into 12 clusters, respectively. Through the visual UMAP map positioning and statistical analysis, compared with the peripheral blood, the proportion of CD69⁺PD-1⁺CD4⁺ T_EM cells (t = 3.427, P = 0.0416) and CD25^-TIGIT^-CD4⁺T_EM cells (t = 3.203, P = 0.0492) in ascites of patients with SBP were increased, with significant differences.

The dimensionality reduction and clustering method of flow cytometric data can be used to detect the groups and functions of T lymphocyte in peripheral blood and ascites samples of patients with SBP, which has the advantages of rapidity, accuracy, multi-dimensionality and strong visualization.