Application of Design Space Mapping Accounting for Model Uncertainty and Common Cause Variability In Drug Substance Development

This talk will focus on the methodology for mapping probabilistic design space based on first principles models and illustrate it with applications within drug substance development. The uncertainties in a model can be classified into two categories (1) Uncertainty in model parameters fitted from experimental data (2) Common cause variability in input factors of the model.  The incorporation of these uncertainties present some statistical and computational challenges in the context of first principles models which will be highlighted in the talk. The application of design space mapping will be illustrated using case studies on reaction kinetics and reactive distillation unit operations.