“…Over the last decades, the increasing amount of antibody-antigen crystal structures has enabled the first quantitative insight into the physicochemical features of antibody-antigen interaction (Benjamin et al, 1984;Berzofsky, 1985;Burkovitz et al, 2013;Dalkas et al, 2014;Lawrence and Colman, 1993;MacCallum et al, 1996;Ofran et al, 2008;Peng et al, 2014;Raghunathan et al, 2012;Sela-Culang et al, 2013;Sivalingam and Shepherd, 2012). For example, it has been observed repeatedly that paratopes localize mostly, but not exclusively, to CDRs (Kunik et al, 2012a), and that certain amino acids are preferentially enriched or depleted in the antibody binding regions (Mian et al, 1991;Nguyen et al, 2017;Ramaraj et al, 2012;Sela-Culang et al, 2013;Wang et al, 2018). For epitopes, several analyses have shown that their amino-acid composition is essentially indistinguishable from that of other surfaceexposed non-epitope residues if the corresponding antibody is not taken into account (Kringelum et al, 2013;Kunik and Ofran, 2013;Ponomarenko and Bourne, 2007).…”