Extended Data Fig. 1: Additional analyses of exhausted intratumoral SCC T cells.
From: Comparative analysis of cell–cell communication at single-cell resolution
a) UMAP projection of all T cells from the dataset published by Ji, et al.29, colored by author-annotated T cell subtype. b) Dot plot depicting average and percent expression of the exhaustion signature score by author-annotated T cell subtypes. c) ROC curves depicting the ability of each cluster from the single-cell T cell object (left) or Scriabin generated T cell-CD1C+ DC CCIM (right) to be classified as exhausted or non-exhausted. Each line corresponds to a single cluster. The diagonal black line corresponds to an AUC = 0.5, where there is no predictive power of classification. AUC = 0, the cluster can be perfectly classified as non-exhausted; AUC = 1, the cluster can be perfectly classified as exhausted.
