Table 1.

Model parameters.

Parameter	Units	Median	95% CI	Information
${\beta }_{{\mathrm{P}}_0}$	(d−1)	1.26	[1.24, 1.28]	Estimated
1/δE	(d)	4.6		15,26
1/δP	(d)	2		15,26
1/η	(d)	5		52
1/γI	(d)	14		11
1/αI	(d)	25		11
βA/βP	(1)	0.022	[0.020, 0.030]	Estimated
βI/βA	(1)	1		11,12
${\beta }_{{\mathrm{P}}_1}/{\beta }_{{\mathrm{P}}_0}$	(1)	0.89	[0.87, 0.92]	Estimated
${\beta }_{{\mathrm{P}}_2}/{\beta }_{{\mathrm{P}}_1}$	(1)	0.72	[0.70, 0.73]	Estimated
$\mathrm{Mean}{\beta }_{{\mathrm{P}}_3}/{\beta }_{{\mathrm{P}}_2}$	(1)	0.50	[0.48, 0.51]	Estimated
$\mathrm{Standard}\mathrm{deviation}{\beta }_{{\mathrm{P}}_3}/{\beta }_{{\mathrm{P}}_2}$	(1)	0.038	[0.025, 0.053]	Estimated
Δt0	(d)	35		11
ω	1	2.42	[2.33, 2.52]	Estimated

The posterior distribution of the parameters marked as estimated was sampled through the DREAM_ZS implementation of the Markov chain Monte Carlo algorithm⁵³. For all estimated parameters, we used uninformative priors within biologically meaningful boundaries. Following our previous application¹¹, we assumed σ = 0.25, r_S = 0.5, ζ = 0.45, and r_E = r_P = r_A = r_R = r_S, whereas r_I = r_Q = r_H = 0. Moreover, γ_Q = γ_I = γ_H, γ_A = 2γ_I, and α_H = α_I.