Table 1 Predictive accuracy of the different models to simulate the effects of diurnal temperature variations on sporozoite prevalence during SMFAs

From: Modelling the effects of diurnal temperature variation on malaria infection dynamics in mosquitoes

Model name

Method to simulate changes in β

Methods to estimate the HMTP

Fitted h

Log-likelihood

AUC

1

Mean

i

*

−737.33

0.75

2

Fluctuating

i

*

−543.59

0.77

3

Mean

ii (maximum)

10.0

−409.8

0.82

4

Fluctuating

ii (maximum)

10.0

−305.93

0.82

5

Mean

ii (mean)

0

−652.47

0.77

6

Fluctuating

ii (mean)

0

−528.83

0.78

7

Mean

ii (minimum)

0

−652.47

0.77

8

Fluctuating

ii (minimum)

0.7

−512.6

0.79

9

Mean

iii (maximum)

0

−652.47

0.77

10

Fluctuating

iii (maximum)

0

−528.83

0.78

11

Mean

iii (mean)

23.8

−543.22

0.78

12

Fluctuating

iii (mean)

20.6

−381.17

0.79

13

Mean

iii (minimum)

0.1

−654.91

0.77

14

Fluctuating

iii (minimum)

3.5

−556.04

0.79

  1. A range of methods are used to calculate the HMTP (c) and the transition rate between exposed mosquito states (β), which are compared by their ability to predict laboratory sporozoite prevalence given mosquitoes are exposed to different temperature profiles. h is the number of hours post infection that are used to estimate the HMTP. * indicates h was not estimated because the HMTP (c) does not vary with h for this model.
Back to article page