Table 2. Estimated timing and frequency of potential punctuated antigenic changes.

Dates (mm/dd/yy) of potential punctuated antigenic changes are identified based on above 15% increases in susceptibility; the standard deviations are computed based on estimates from all model-inference runs indicating punctuated antigenic changes around the same time. The time intervals between subsequent antigenic changes are estimated based on four different distributions; the best-fit estimates are shown in bold. For A(H1N1), numbers in the parentheses show results when dates during the pandemic are included.

Strain	Timing of change		Time intervals between antigenic changes
	Date	SD (days)	Mean (years)	SD (years)	Distribution	AIC
A(H1N1)	1/21/00	4.2	4.6 (3.7)	4.6 (3.7)	exponential	69.45
	6/11/06	0	2.9 (2.0)	0.01 (0.01)	log-norm	70.62
	1/25/09	0	4.5 (3.8)	4.1 (4.2)	gamma	71.35
	7/6/09	3.5	4.5 (3.9)	4.2 (4.7)	Weibull	71.42
	9/13/09	2.2
	1/5/10	4
	1/14/11	2.6
A(H3N2)	5/15/98	6.2	1.8	1.2	gamma	166.26
	1/10/99	4.2	1.8	1.1	Weibull	166.29
	1/9/00	1.9	1.4	0.01	log-norm	166.31
	6/27/00	6.3	1.8	1.8	exponential	167.15
	1/14/03	8.8
	7/11/04	0
	2/27/05	14.8
	6/20/07	3.7
	7/28/10	4
	5/10/12	8.1
	1/3/15	3.7
	6/8/15	5.7
B	2/6/00	0	6.2	6.2	exponential	54.30
	1/27/02	0	3.2	0.01	log-norm	55.64
	1/6/03	7.2	6.2	6.7	Weibull	56.26
	1/13/18	0	6.2	6.4	gamma	56.29