Table 3.
Degrees of freedom (df), log-Likelihood (log(L)), AICc, ΔAICc (AICcmodel − AICcmin model), and AICc w for each linear model in the analysis of experiment iii ordered in descending fit
| Model description: sqrt(Gametocyte Density) ~ + (1|mouse) | df | log(L) | AICc | ΔAICc | AICc w |
|---|---|---|---|---|---|
| Age | 5 | − 99.37 | 209.6 | 0.000 | 0.409 |
| Age + treatment + age * treatment | 8 | − 96.15 | 210.4 | 0.850 | 0.267 |
| Age + treatment | 6 | − 99.08 | 211.4 | 1.785 | 0.168 |
| Null | 3 | − 103.5 | 213.3 | 3.759 | 0.062 |
| Age + samp.time | 7 | − 99.52 | 214.7 | 5.088 | 0.032 |
| Treatment | 4 | − 103.3 | 215.1 | 5.469 | 0.027 |
| Age + samp.time + treatment + age*treatment | 10 | − 96.38 | 216.1 | 6.556 | 0.015 |
| Age + samp.time + treatment | 8 | − 99.22 | 216.6 | 7.000 | 0.012 |
| Samp.time | 5 | − 103.7 | 218.3 | 8.743 | 0.005 |
| Samp.time + treatment | 6 | − 103.5 | 220.2 | 10.57 | 0.002 |
The response variable for each model is the square root-transformed gametocyte density and the random effect is “mouse”. “Samp.time” refers to sampling time (4-, 8- or 12-HPI), “age” refers to gametocyte age (19-, 25-, or 31-h old) and treatment is either TNF-α or control. The null model includes only “mouse” as a random effect