Table 1.
Parameters of the ex vivo model | |||
---|---|---|---|
Parameter | Unit | Description | Prior distribution |
μ′ | h | mean of the normal distribution defining the initial parasite age distribution | U(0, 24) |
σ′ | h | SD of the normal distribution defining the initial parasite age distribution | U(0, 20) |
q | (unitless) | a fraction parameter | U(0, 1) |
c1′ | ng | maximum level of intraerythrocytic PvLDH | U(0, 0.01) |
c2′ | h | parasite age where the intraerythrocytic PvLDH level reaches c1/2 | U(0, 72) |
δ′ | h−1 | ex vivo death rate of parasites | U(0, 0.5) |
r′ | h−1 | ex vivo rupture rate of dead parasite | U(0, 1) |
λ′ | h−1 | ex vivo decay rate of PvLDH | N(0.0106, 4.1565 × 10−4) within (0, 0.1) |
Lm0 | ng | initial level of PvLDH in culture media | U(0, 1.755) |
h | ex vivo length of life cycle | Fixed to be 72 |
The prior distribution for the population mean of each model parameter is a uniform distribution with biologically plausible boundaries specified in the table. Note that the ex vivo length of asexual life cycle was fixed to be 72 h, as explained in the main text. The prior for λ′ was determined based on a modelling study of an independent in vitro experiment where decay data of a human PvLDH were collected and fitted by an exponential decay model (Eq. 1; see the first section of Methods for details). We assigned an upper limit of Lm0 of 1.755 ng because the initial PvLDH level in culture media was below the limit of quantification of 3.9 ng/mL, which is equivalent to 1.755 ng based on the culture media volume of 0.45 mL. Priors for other parameters were set to be uniform distributions with bounds selected either based on biological plausibility (e.g., 0 as a lower bound) or to be some appropriate values to avoid any resolvable truncation of the posterior distribution (e.g., those upper bounds shown in the table).