Robust Design Strategies for Bioassay and Drug Synergy

• Activity Number: 259
• Type: Topic Contributed
• Date/Time: Tuesday, August 10, 2004 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Timothy E. O'Brien*+
• Companies: Loyola University Chicago
• Keywords: optimal design ; generalized nonlinear modeling ; drug synergy ; relative potency ; interaction ; oncology

Abstract

The validity of biomedical and scientific research depends upon the accuracy of the statistical models and the efficiency of the experimental designs used by these researchers. Optimal design theory provides researchers with the means to select the "optimal" settings for their experiment in the sense of yielding a design or plan that will require the fewest repetitions to yield accurate results. However, since statistical models are often not known with complete certainty, these designs may be inappropriate and often have no ability to signal significant lack of fit. Thus, so-called robust designs are desired to estimate model parameters and accurately make predictions, on the one hand, and to highlight model misspecification, on the other. The focus of this paper is to outline recent developments in modeling and design strategies for the detection of relative potency and synergy for situations where the assumed (nonlinear) model function, the initial parameter estimates, or the error structure is not completely known.