To examine the variation at the identified amino acid positions
in this study, 1130 full-length H9N2 NA sequences (as of February
16, 2015), derived from GenBank, were analyzed using Multiple
Sequence Comparison by Log-Expectation (MUSCLE) version
3.4.31 with default settings. The variation was
expressed as the percentage of viruses carrying a different residue
at the critical amino acid positions identified in this study.