No strong evidence was found that HA-1 antigenic drift
was due to vaccination. The genetic variation observed is thought
to result from non-sterilizing immunity. Positive selection did
occur in a low number of isolates, with evidence of episodic
diversifying selection. Molecular variation was found to mediate
partial cross-protectivity. The array of antigenic site variations was
determined, explaining the loss of immune-potency of a single
vaccine seed. Virus surveillance in poultry farms is essential for
pinpointing appropriate vaccine seeds, and these data will be
invaluable in understanding human infection.