High domain conservation in NS1 groups, subgroups and lineages.

Pattern of prevalent (at least 85%) amino acid residues in indicated positions for each class in db4535. The motif conservation (%) is described in the Results section. F103 and M106 are indispensable for CPSF30 binding [33], [34]; FMKDD motif results in strong binding, while FMREG results in weak binding [32]. Prolines in 212 and 215 as well as positive (K) aa in 217 residues are indispensable for crk/crkL binding [15];*only 4.33% of sequences have both P212 and P215 in A5.2, while letters in parentheses represent consensus residues for 230aa sequences (18.67%). Lysines 217, 219 and 221 are susceptible to sumoylation, while M222V mutation inhibits sumoylation [35]. Amino acids 227–230 constitute a PDZbm [2]. The bar (|) notation represents the most common residues in that position in decreasing order of frequency.