In industrial biotechnology, process optimization can be an important step during process development. A common objective is to maximize the product yield or other process-related responses. By seeking to only maximize the desirability of the responses, one might however find conditions that are not stable. A way to take into account stability is to use the Simulation Experiment tool found in the JMP Prediction Profiler. The simulation can be performed once a model describing the relationship between factor settings and responses is established. It will then find the condition(s) around which the responses vary the least.
This work presents the results obtained after applying the Simulation Experiment on a Custom Design DoE model.
The differences in results between the approach maximizing desirability and the approach using the simulation are shown. In this case study, both approaches led to selecting very similar factor settings i.e. the settings that led to maximum product yield corresponded also to a very stable operating point.
The Simulation Experiment made it possible to find the most stable point, which would have otherwise been extremely difficult and time-consuming considering the high amount of interaction effects and the complexity of the model.
Weitere Informationen: https://community.jmp.com/t5/Discovery-Summit-Munich-2020/Industrial-Biotech-Case-Study-Customized-D...
