Fig. 2

Unbiased clustering of memory T helper cells using FlowSOM algorithm. PBMCs of psoriasis and PsA patients and healthy controls were isolated and used for Flow cytometry. Doublets and dead cells were excluded. A Living CD3⁺CD4⁺CD45RO⁺CD25^low/int cells were gated for each subject and exported, arcsinh-transformed, down-sampled to 6000 cells per sample, and aggregated per group. The 3 aggregated files were used to generate a FlowSOM tree and cells were clustered into 100 nodes. The nodes were clustered in 12 metaclusters with hierarchical clustering, indicated by the background colors. B Table of metaclusters and the known labels for the different memory T helper subsets. C Median expression of the 6 individual surface markers (CCR6, CD25, CCR4, CXCR3, CCR10, and CLA) plotted in the generated FlowSOM tree