Fig. 3: TCR and CD28 titration identifies genes sensitive to each signal.

A In the linear model, we required a linear increase of gene expression along with stimulus intensity (incremental fold-change ≥ 1.5 in gene expression), separately evaluating naive and memory cells. Genes that did not follow the linear model were tested for the switch model. Here, we assumed an ‘on-and-off’ mode of expression where a gene is significantly upregulated (fold-change ≥ 2) in response to the presence of either CD28 or TCR. In both of these models, we used all seven conditions, e.g. when testing for CD28-sensitive genes, we grouped the TCR alone stimulation with the resting since neither received CD28 signal. A gene was classified in one of the two categories without overlap and prioritised for the linear model. B Comparison of the number of genes in naive and in memory cells that are TCR or CD28 sensitive. C Hallmark pathways enriched for TCR-sensitive and CD28-sensitive genes in naive and memory cells using fgsea. D Pathway enrichment plot for G2M checkpoint. E Gene expression profiles of selected switcher genes, genes that change stimulus sensitivity between the two cell types. The number below the stimulus name represents the significance of the difference between naive and memory T cells derived using t-test.