Structurally profiling the baseline repertoire snapshots [31].

A full table containing the values for all 41 baseline datasets is available in the Supporting Information (S2 Table). In order, the columns show: the dataset label, the number of VH and VL reads within each snapshot, the number of FREAD-modellable VH and VL reads (once clustered at 90% sequence identity), the number of predicted modellable Fvs resulting from these VH-VL pairings, and the number of distinct structures (cluster centres) identified in each dataset. Mod. = Modellable, SIC = Sequence Identity Clustered.