Bayesian Model Averaging for Benchmark Dose Analysis in Developmental Toxicology

To reduce uncertainty due to model selection when a large number of potential candidate models is available, the use of Bayesian Model Averaging (BMA) has emerged as an important tool. As known, the BMA methodology is a coherent approach since we can express the desired quantities as a weighted average of model specific quantities with the weights determined based on how much the data supports each model. In toxicological studies, a wide range of statistical models have been utilized for dose-response modeling and risk assessment with no particular model receiving a universal acceptance. Here, we consider the application of BMA for benchmark dose estimation in developmental toxicity experiments. In such experiments, as in all noncancer studies, the choice of the model can play a crucial role in the final benchmark dose estimates. A Bayesian approach along with the MCMC method is used to fit each individual model used as a component in model averaging and to derive the posterior weights. A simulation study of a developmental toxicity experiment is used to illustrate the methodology