High-throughput screening methods are used routinely to support process development in a rapid and cost-effective manner. In order to apply these methods also for process characterization, robotic scale-down models need to be established that are comparable to large-scale commercial processes. For this reason, the systematic adoption of high-throughput techniques in late-stage development has come with the need to better understand and predict scale differences.<br/>The qualification of high-throughput techniques as a scale-down model is commonly performed on an empirical base. Scale offsets are thereby mostly attributed to differences in system geometry. Recent studies have also shown that the scale offset can vary, depending on which process parameters are screened. To get deeper insights on how scale differences affect process performance, scale offsets have also been described on a mechanistic level using established chromatography models. This approach requires an extensive system characterization to effectively account for scale differences and to support scale-up. Yet a thorough system characterization can hardly be conducted in high-throughput process development.<br/>In this study we present an approach that combines Bayesian statistics and mechanistic modeling to perform a risk-based scale-up of a high-throughput chromatography system. The mechanistic model, which is calibrated using high-throughput data, will be used to predict large-scale performance based on the robotics screening readouts. Bayesian statistics is applied to account for the experimental and systematical uncertainty during calibration of the mechanistic model. In this way, the prediction uncertainty of the mechanistic model reflects the uncertainty in the scale-down model and can provide deeper insights on how scale differences affect process performance. This model-based assessment of the prediction uncertainty of high-throughput methods will therefore allow to evaluate the suitability of the scale-down model for process characterization.

Risk-based scale-up of high-throughput chromatography systems using Bayesian statistics and mechanistic modeling

Macromolecular Chemistry: The Second Century

[Author Benefits](https://scimeetings.acs.org/?utm_source=pubs_content_marketing&utm_medium=website&utm_campaign=0320_MCF_NPI_Launch_Spring_Homepage&ref=pubs_content_marketing)

[How to use SciMeetings](https://storage.googleapis.com/dl.morressier.com/acs/SciMeetings-ACSSpring2020-HowToGuide_M3020-Final.pdf)

[Frequently Asked Questions](https://storage.googleapis.com/dl.morressier.com/acs/SciMtng_ACSNMSp2020_FAQ_M3120_Final.pdf)

ACS Spring 2020 National Meeting & Expo

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.