| In environmental risk assessment of a toxic chemical, main focus is in understanding induced target organ toxicity and/or carcinogenicity. Mathematical models based on systems of ordinary differential equations with biologically relevant parameters are tenable methods for describing the disposition of chemicals in target organs. In evaluation of a toxic chemical, dose response assessment often addresses only toxicodynamics [TD] of the chemical, while its toxicokinetics [TK] do not enter into consideration. The primary objective of this research is to integrate both TK and TD in evaluation of toxic chemicals while performing dose response assessment. Population models with hierarchical setup and nonlinear predictors, for TK concentration and TD effect measures are considered. A one compartment model with biologically relevant parameters, such as organ volume, uptake rate and excretion rate or clearance, is used to derive the TK predictor while a two parameter Emax is used as a predictor for TD measures. Hierarchical modeling setup allows inclusion of both intra- and inter-subject variations in the model. Inference of the model parameters with nonnegative constraint has been carried out by Bayesian approaches using MCMC techniques. The inference procedure was adapted to address assay's Limit of Detection (LOD) values as well.; Exposure assessment is another important task in environmental risk assessment and should be done prior to formulate dose response relationship. Understanding presence (or absence) of an agent, and its concentration and distribution is the primary focus of an exposure assessment. Predictive models and environmental monitoring can be used to determine the levels of exposure at different points. Spatial prediction and interpolation of pollutant concentrations is considered here as a mode of exposure assessment. Models based on spatial covariance matrices are considered to construct a predictor of pollutant concentrations at points of interests. Bayesian approach with MCMC techniques is implemented to estimate model parameters. The whole process; known as kriging , is applied to predict the arsenic concentration in ground water used for personal usage. |
