Fig. 5: GLIPH prediction of TCR specificity shows convergence toward shared specificities in autoimmune and non-autoimmune follicular T cells.
a Schematic of computational pipeline to assign clonotype groups from scRNA-seq data, followed by assigning clonotypes to TCR specificity groups using the GLIPH2 algorithm. b Scatter plot comparing specificity group size between WT and mixed 564Igi chimeras. Specificity groups are colored according to the number of unique clonotypes that belong to a given specificity group and sized according to number of samples in which the given specificity group is observed. Select specificity groups are labeled with their arbitrary name in blue. c Mappings of indicated specificity groups onto UMAP visualization of transcriptomic data from WT (left) or mixed 564Igi (right) chimeras. d Motif analysis of 14 amino acid long CDR3s of TCRα (left) and TCRβ (right) chains of all cells belonging to indicated specificity group. e Unweighted network analysis of clonotype assignment to indicated specificity groups. Specificity groups are depicted as blue circles, clonotypes are colored according to condition (black, WT; red, 564Igi; green, both) and sized according to number of cells belonging to given clonotype. f Scatter plot comparing TCR specificity group size between WT and mixed 564Igi chimeras and colored according to most prevalent cluster amongst cells belonging to given specificity group (top left), percentage of cells within given specificity group assigned to the TFR cluster (top right), geometric mean of clone sizes of clonotypes within given specificity group (bottom left), and Shannon diversity index of clonotype expansion within given specificity group (bottom right).
