|
Abstract:
|
Cell-based biological assays(also called bioassays) are an indispensable part of the quality assessment in the pharmaceutical industry. They determine the potency of drug substances in a sample through Relative Potency against a standard preparation. Block designs are an efficient tool to assay multiple preparations simultaneously and mitigate multiple sources of bias, including the cell-culture plates used, locations within plates, and the operators. In the design of bioassays, orthogonality is desired between preparations and the nuisance factors. However, the treatment/control setup of a bioassay and the multiple block effects makes the construction of a block design difficult. Another challenge is that, in practice, when absolute orthogonality is not feasible, we need a design that achieves “near-orthogonality”. As a solution to these challenges, we proposed an Extended Youden Design with mathematical constructions. We also developed an efficient algorithm implementation with R. Numerical results showed that Extended Youden Design is the optimal design for the given scenario.
|
