A Novel LASSO type Bayesian Weighted Quantile Sum Regression for highly correlated multi-group mixture analysis: current context, importance and future directions

➔ The underlying philosophy is that even when the individual concentrations of single exposures are all below the regulatory level, the cumulative exposure might be substantially higher or may be associated with health endpoints. [...] What’s yet to be done These Bayesian Hierarchical Models offer a lot of flexibility but that might lead to data being severely overfitted, particularly when the exposures are highly correlated in the presence of multiple groups How to address this problem ? Introduce statistical regularization or penalty terms Motivation for this method What’s the Novelty? ● We propose adding Penalty term or stati. [...] What’s the mathematical innovation here ? Modelling the Laplace (double-exponential) distribution as a scale mixture of a normal distribution with an exponential mixing density (Kyung et al., 2010) That’s great but does this method work in real life ? Yes it does! We tested the LASSO type Group BWQS method addressing the aim To evaluate the association between prenatal environmental chemical expos. [...] Then what are the potential advantages/limitations of using either method ? Let’s do some simulations! Simulated Correlation Matrix for Exposures The outcome is a linear combination of the Groups of exposures with pre- fixed coefficients Scenario 1: Two Major Contributing Exposures in each group Heatmap of Estimated Weights Penalized Group-Mixture BWQS Hierarchical BKMR Measures of the Performance. [...] Depending on the research question and the data structure, other exposure mixture approaches may be suitable (e.g., if there is only one major contributing exposure in a group, both Group BWQS and Hierarchical BKMR may perform equivalently).