Difference between revisions of "ERF for Frambozadrine in rats"
From Testiwiki
m (links added) |
(first results added) |
||
Line 83: | Line 83: | ||
== Result == | == Result == | ||
+ | |||
+ | It is not clear which of the plausible methods for estimating the result is the best. The discussion is ongoing. {{disclink|Which method is the best for dose-response estimation?}} | ||
+ | |||
+ | ===Non-parametric Bayesian estimation=== | ||
+ | |||
+ | Males and Females combined | ||
+ | |||
+ | {| {{prettytable}} | ||
+ | !Dose levels | ||
+ | !MLE | ||
+ | !Prior | ||
+ | !Posterior mean | ||
+ | !Variance | ||
+ | |---- | ||
+ | |0 | ||
+ | |0.053 | ||
+ | |0.125 | ||
+ | |0.0571 | ||
+ | |0.0634 | ||
+ | |---- | ||
+ | |1.2 | ||
+ | |0.133 | ||
+ | |0.25 | ||
+ | |0.1052 | ||
+ | |0.0348 | ||
+ | |---- | ||
+ | |1.8 | ||
+ | |0.102 | ||
+ | |0.375 | ||
+ | |0.13 | ||
+ | |0.0241 | ||
+ | |---- | ||
+ | |15 | ||
+ | |0.091 | ||
+ | |0.5 | ||
+ | |0.154 | ||
+ | |0.021 | ||
+ | |---- | ||
+ | |21 | ||
+ | |0.064 | ||
+ | |0.625 | ||
+ | |0.1799 | ||
+ | |0.0551 | ||
+ | |---- | ||
+ | |82 | ||
+ | |0.511 | ||
+ | |0.75 | ||
+ | |0.4969 | ||
+ | |0.2762 | ||
+ | |---- | ||
+ | |109 | ||
+ | |0.687 | ||
+ | |0.875 | ||
+ | |0.687 | ||
+ | |0.2778 | ||
+ | |---- | ||
+ | |} |
Revision as of 11:15, 24 October 2007
Scope
Frambozadrine dose-response function in rats describes the long-term health impact(s) caused by frambozadrine as a function of dose in rats. This dose-response function applies only to continuous long-term exposures of frambozadrine (like in chronic studies).
Contents
Definition
Causality
Upstream variables not defined.
Data
Toxicological data about frambozadrine in rats.
Dose(mg/kg-day) | Total no rats | Hyperkeratosis |
---|---|---|
Male | ||
0 | 47 | 2 |
1.2 | 45 | 6 |
15 | 44 | 4 |
82 | 47 | 24 |
Female | ||
0 | 48 | 3 |
1.8 | 49 | 5 |
21 | 47 | 3 |
109 | 48 | 33 |
Plausible dose-response functions
- Generalized dose-response function
- Multistage model (first order)
- Multistage model (second order)
- Weibull model
Formula
Methods for estimating dose-responses
- Bootstrap method
- Probabilistic inversion
- Non-parametric Bayesian estimation
- Bayesian model averaging
Unit
probability of impact
Result
It is not clear which of the plausible methods for estimating the result is the best. The discussion is ongoing. D↷
Non-parametric Bayesian estimation
Males and Females combined
Dose levels | MLE | Prior | Posterior mean | Variance |
---|---|---|---|---|
0 | 0.053 | 0.125 | 0.0571 | 0.0634 |
1.2 | 0.133 | 0.25 | 0.1052 | 0.0348 |
1.8 | 0.102 | 0.375 | 0.13 | 0.0241 |
15 | 0.091 | 0.5 | 0.154 | 0.021 |
21 | 0.064 | 0.625 | 0.1799 | 0.0551 |
82 | 0.511 | 0.75 | 0.4969 | 0.2762 |
109 | 0.687 | 0.875 | 0.687 | 0.2778 |