Hi community. I'm looking for support to create a workflow in JMP 17.2 for analyzing and plotting data from bioassay qualification study. Specifically, I'd like to base the calculations of relative bias and intermediate precision on equations provided in regulatory guidelines, USP <1033>. @Valerie_Nedbal, @gail_massari
The qualification study design consists of two independent 96WP per run, on each plate full dose response curves of standard reference and 3 samples at varied target relative potency (RP) level (same replication for each): 70, 86, 100, 119 and 142 % RP. The study investigates some robustness factors such as operator, incubation time and selected reagent lot. Specificity is also assessed by testing formulation buffer.
Graph 1: Target RP ratio versus measured RP ratio
Relative bias (RB) calculation with the following steps for each target potency level:
- Geometric mean of all %RP repeats
- Percent RB
- 90% confidence intervals (CI) for average logarithm RP ratio and then calculated CI %RB
Graph 2: %RB with 90% CI for each target RP ratio
Intermediate precision (IP) calculation with the following steps for each target potency level:
- Single factor ANOVA to calculate total variance (between and within groups) for logarithm of RP
- % Geometric CV
- Then, summary from all IP RP levels statistics including kurtosis and skewness calculation.
Separately, it would be great to be able to calculate Cpm and robustness data.
Many thanks in advance.
