Online Program Home
  My Program

Abstract Details

Activity Number: 229 - Random Effect/Mixed Models
Type: Contributed
Date/Time: Monday, July 31, 2017 : 2:00 PM to 3:50 PM
Sponsor: Biometrics Section
Abstract #322569 View Presentation
Title: Optimum Designs for Nonlinear Models in the Presence of Multiple Covariates
Author(s): Mahbub AHM Latif* and Barbara Bogacka and Steven G Gilmour
Companies: St. Luke's International University and Queen Mary, University of London and King's College London
Keywords: Pharmacokinetic experiments ; Design of experiments ; Mixed effects model
Abstract:

Pharmacokinetic experiments, conducted in the early stages of drug development process, play an important role in assessing the association between different enzyme activities and substrate under study. Nonlinear (e.g. Michealis-Menten model) mixed effects (NLME) models are commonly used for analyzing these experiments. Enzyme activities, related to the human liver microsomes (HLM) used in the experiments, are considered as covariates for modelling the random parameters of the NLME model, which helps to identify important enzymes for metabolizing the substrate. Outcome of this analysis can make important contributions to the future course of the research with the substrate. Designing such pharmacokinetic experiments requires considering the combination of HLM (enzyme activities) and substrate concentration levels as treatment factor and often a rich design, based on all possible treatments, is used in industrial setup. Depending on the type of the model assumed for metabolism of the substrate, three design criteria are proposed. We can show that optimum designs based on the proposed criteria can get the required information in many fewer runs compared to the rich design.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2017 program

 
 
Copyright © American Statistical Association