Table 4. Model selection results for multiple regression models testing the relationship between viral RNA concentrations in lung tissues as a function of virus subtype, sex, and day post inoculation and interactions between the three variables.
| Model | K | Adj. R2 | AICc | ΔAICc | AICc weight |
| Subtype+Sex+DPI+Sex*DPI | 9 | 0.62 | 374.3 | 0.0 | 0.46 |
| Subtype+Sex+DPI+Subtype* Sex+Sex * DPI | 13 | 0.63 | 377.1 | 2.8 | 0.11 |
| Subtype+Sex+DPI+Subtype*DPI+Sex *DPI | 13 | 0.63 | 377.1 | 2.8 | 0.11 |
| Subtype+Sex+DPI | 8 | 0.61 | 377.2 | 2.9 | 0.11 |
| Subtype+DPI | 7 | 0.60 | 378.6 | 4.3 | 0.05 |
| Subtype+Sex | 7 | 0.60 | 379.0 | 4.7 | 0.04 |
| Subtype+Sex+DPI+Subtype*DPI | 12 | 0.62 | 379.9 | 5.6 | 0.03 |
Only models with a ΔAICc<6 are shown. K is the number of parameters. Adj. R2 is the R2 value adjusted for the number of parameters in the model; it indicates the amount of variation explained in the model. AICc is Akaike's information criterion adjusted for small sample size. ΔAICc values indicate the difference between a given model and the best model. The AIC weight shows the relative support for each model.