Bioassay case study applying the maximin D‐optimal design algorithm to the four‐parameter logistic model. (31st July 2015)

• Record Type: Journal Article
• Authors: Coffey, Todd
• Is Part Of: Pharmaceutical statistics. Volume 14:Number 5(2015:Sep./Oct.)
• Page Start: 427
• Page End: 432
• Publication Date: 2015-07-31
• Subjects: Pharmacy -- Statistical methods -- Periodicals; Pharmacy -- Statistics -- Periodicals; 615.10727
• DOI: 10.1002/pst.1702
• Languages: English
• ISSNs: 1539-1604

Abstract

<abstract abstract-type="main" id="pst1702-abs-0001"> <title> <x xml:space="preserve">Abstract</x> </title> <p id="pst1702-para-0001">Cell‐based potency assays play an important role in the characterization of biopharmaceuticals but they can be challenging to develop in part because of greater inherent variability than other analytical methods. Our objective is to select concentrations on a dose–response curve that will enhance assay robustness. We apply the maximin D‐optimal design concept to the four‐parameter logistic (4PL) model and then derive and compute the maximin D‐optimal design for a challenging bioassay using curves representative of assay variation. The selected concentration points from this 'best worst case' design adequately fit a variety of 4PL shapes and demonstrate improved robustness. Copyright © 2015 John Wiley &amp; Sons, Ltd.</p> </abstract>