Finding the Sweet Spot: Cross-Product Analytical Optimization in Biopharmaceutical Development (2026-EU-30MP-2788)

In biopharmaceutical development, a recurring challenge is determining how analytical insights can be leveraged to establish and optimize methods. This case study focuses on a specific chromatographic technique, with the underlying approach having been successfully applied in other areas such as formulation and bioprocess development.

To address this task, data from individual design of experiments (DOEs) were either merged by introducing the target as an additional factor or combined afterward through model integration to generate unified predictive functions. A key challenge was ensuring consistent evaluation across projects to enable comparability of models. JMP’s Generalized Regression was instrumental in building robust, interpretable models.

Based on these models, optimization targets and ranges were defined. Using JMP’s Profiler and Simulation functionalities, mean predictions and Monte Carlo simulations were conducted to explore the design space with respect to key performance criteria. Strategic visualizations revealed optimal parameter combinations across several input and output variables simultaneously, making complex data tangible for subject matter experts.

As an outcome, similar input-output trends were identified, enabling the derivation of generalizable process knowledge. In some cases, uncertainty quantification allowed for risk assessment of proposed set points, supporting informed decision making in method development. For more complex scenarios, surrogate modeling techniques such as artificial neural networks (ANNs) and other machine learning approaches were applied to uncover deeper patterns.

This case study demonstrates how JMP supports multidisciplinary collaboration and enables strategic statistical thinking in pharmaceutical development.