In this section
@inbook{Triantafyllou:2023:10.1016/B978-0-443-15274-0.50061-5,
author = {Triantafyllou, N and Shah, N and Papathanasiou, MM and Kontoravdi, C},
booktitle = {Computer Aided Chemical Engineering},
doi = {10.1016/B978-0-443-15274-0.50061-5},
pages = {381--386},
title = {Combined Bayesian optimization and global sensitivity analysis for the optimization of simulation-based pharmaceutical processes},
url = {http://dx.doi.org/10.1016/B978-0-443-15274-0.50061-5},
year = {2023}
}
TY - CHAP
AB - We propose an efficient framework that employs Bayesian optimization and global sensitivity analysis for the optimization of detailed pharmaceutical flowsheets. Global sensitivity analysis based on quasi-random sampling is utilized to reduce the dimensionality of the problem by identifying critical process and economic parameters that contribute significantly to the variability of Key Performance Indicators (KPIs) such as batch size and OpEx. Then, Bayesian optimization is performed in the previously identified critical input space based on gaussian process surrogate models and a number of different acquisition functions to find the optimal critical operating conditions that minimize the aforementioned KPIs. We apply this framework to the manufacture of plasmid DNA (pDNA), which is a critical raw material for advanced therapeutics, leading to a surge in demand for pDNA for clinical or commercial use. Optimized manufacturing recipes identified with the proposed framework are projected to achieve an up to 170% increase in the batch size and a 34.7% decrease in the OpEx per batch.
AU - Triantafyllou,N
AU - Shah,N
AU - Papathanasiou,MM
AU - Kontoravdi,C
DO - 10.1016/B978-0-443-15274-0.50061-5
EP - 386
PY - 2023///
SP - 381
TI - Combined Bayesian optimization and global sensitivity analysis for the optimization of simulation-based pharmaceutical processes
T1 - Computer Aided Chemical Engineering
UR - http://dx.doi.org/10.1016/B978-0-443-15274-0.50061-5
ER -