Fig. 1

Study workflow. The PBMCs from a whole blood venous sample of COVID-19 patients were divided into two aliquots. One aliquot was directly subjected to TCRβ sequencing (pre-stimulation repertoires). The other aliquot was first stimulated with S-peptides pool, in the presence of IL-2 for 12 days, to promote the expansion of S-specific cells, and then was also subjected to TCRβ sequencing (post-stimulation repertoires). TCRβ repertoire analysis involved a filtering step to exclude confounding sequences not shared between pairs of pre- and post-repertoires. The mapping of S epitopes associated with clonotypes was carried out by recognizing in the repertoires the TCRβ experimentally associated with S epitopes (public databases MIRA [20] and VDJdb [21]), as well as inferring the S-specificity of clonotypes using the GLIPH algorithm [12]