Table 3. Posterior inferences of the coefficients (on the logit-scale) in the best Bayesian zero-one-inflated beta distribution model on proportion of tissue area with tumor metastasis.
The first model component estimates the mean (linear predictor) in the model, and the second component the probability of zero. The factor levels ‘Experiment 1’, ‘Control’ (anesthetic), ‘C57BL/6’ (mouse strain) and ‘Liver’ (organ) are the baseline values in the model and are included in the intercepts. d—Regression coefficient in the linear predictor for the sum of the two shape parameters in the beta distribution. *—indicates a significant difference when the quantile range does not overlap zero.
Model Component | Coefficient | Mean | 25% quantile | 97.5% quantile |
---|---|---|---|---|
Logit(mean) | Intercept* | -4.338 | -5.133 | -3.568 |
Experiment 2* | 0.918 | 0.531 | 1.246 | |
Experiment 4 | 0.221 | 0.074 | 0.567 | |
Isoflurane | 0.030 | -1.066 | 1.172 | |
Ketamine | 0.527 | 0.584 | 1.685 | |
NSG* | 1.768 | 0.966 | 2.637 | |
Lung* | 2.544 | 1.607 | 3.560 | |
Isoflurane:NSG | 0.284 | -0.639 | 1.067 | |
Ketamine:NSG | 0.135 | -0.699 | 1.158 | |
NSG:Lung | 0.494 | 0.366 | 1.293 | |
Isoflurane:Lung | -0.490 | -1.403 | 0.402 | |
Ketamine:Lung | -0.809 | -1.698 | 0.230 | |
logit(Pr(y = 0)) | Intercept | -0.419 | -1.113 | 0.0463 |
Isoflurane | -0.165 | -1.266 | 0.864 | |
Ketamine | -0.033 | -1.012 | 1.111 | |
Lung* | -2.609 | -4.016 | -1.406 | |
d | 2.131 | 1.880 | 2.454 |