Annotation of T-cell subclusters across different single-cell technologies. (A) Clinical course showing white blood cell (WBC) count per nanoliter with transient partial response to nivolumab (C2D1, C4D1) and subsequent disease progression with unsuccessful re-exposure to nivolumab (relapse, C6D1, and progression). (B) Percentage of AML and T cells in peripheral blood quantified with scRNA-seq (27 777 cells), scATAC-seq (28 713 cells), and CyTOF (677 322 cells). (C) UMAP representation and clustering (scRNA-seq) of T/NK cells (n = 2711). Annotation of clusters based on surface marker expression (E) and single-cell gene expression (F). (D) Clonal expansion of T cells defined as number of T cells with the same clonotype based on identical complementarity defining region 3β (CDR3β) sequence. Gray indicates no CDR3β sequence detected. (E-F) Surface marker expression obtained through CITE-seq (E) and single-cell gene expression (F) of single T/NK cells across clusters. (G) UMAP representation and clustering of T cells (CyTOF) based on protein expression of lineage-defining surface markers. The heatmap shows the median expression of markers used for cluster annotation. (H) Protein expression of TBX21, TIGIT, CD28, and GZMB on T cells (CyTOF). (I) Chromatin accessibility of PDCD1 gene (encoding PD-1 protein) across T-cell subpopulations defined by scATAC-seq (supplemental Figure 1I). exh., exhausted; sen., senescent; CM, central memory; EM, effector memory; TEMRA, effector memory T cell re-expressing CD45RA.