BMA‐Mod: A Bayesian model averaging strategy for determining dose‐response relationships in the presence of model uncertainty

Abstract

Successful pharmaceutical drug development requires finding correct doses. The issues that conventional dose-response analyses consider, namely whether responses are related to doses, which doses have responses differing from a control dose response, the functional form of a dose-response relationship, and the dose(s) to carry forward, do not need to be addressed simultaneously. Determining if a dose-response relationship exists, regardless of its functional form, and then identifying a range of doses to study further may be a more efficient strategy. This article describes a novel estimation-focused Bayesian approach (BMA-Mod) for carrying out the analyses when the actual dose-response function is unknown. Realizations from Bayesian analyses of linear, generalized linear, and nonlinear regression models that may include random effects and covariates other than dose are optimally combined to produce distributions of important secondary quantities, including test-control differences, predictive distributions of possible outcomes from future trials, and ranges of doses corresponding to target outcomes. The objective is similar to the objective of the hypothesis-testing based MCP-Mod approach, but provides more model and distributional flexibility and does not require testing hypotheses or adjusting for multiple comparisons. A number of examples illustrate the application of the method. K E Y W O R D SBayesian model averaging, likelihood principle, MCMC, nonnormal distributions, predictive distributions

Factors Affecting the Willing to Join in Coffee Crop Insurance in Dak Lak Province, Vietnam: A Novel Application of Bayesian Model Averaging Approach

“…According to 1 showed that the correlation matrix in the model of factors affecting the willingness to join crop insurance. Gould, 2018), the LR statistical test, according to the Chi-square distribution, is 167.47 and the probability value Pr (> Chi2) <0.0001 so the Ho hypothesis is rejected. The five factors included in the Logistic model are important for insurance.…”

“…According to (Wang, D., Zhang, W., Bakhai, A, 2004), a Bayesian solution to the model's uncertainty has been proposed and applied recently (A. Lawrence Gould, 2018). This method selects a subset of all possible models (max K = 2 p , ignoring interactions between explanatory variables) (Krzysztof Drachal, 2018) .…”

Optimal adaptive allocation using deep reinforcement learning in a dose‐response study

“…, where s i is an element of s, and 𝛼 (1) , 𝛽 (1) , 𝛼 (2) , 𝛽 (2) , 𝛼 (3) , 𝛽 (3) are the parameters of the DNN. We estimate 𝜋 * using reinforcement learning.…”

Flexible Fitting of PROTAC Concentration–Response Curves with Changepoint Gaussian Processes

“…In addition, some parts of a curve can be fixed, while others modeled more flexibility. A parametric framework offers two approaches to fitting when the shape of a curve is not known in advance: model averaging 7 and model selection. Both approaches start with a set of parametric models and then undergo two stages.…”