5 Conclusion and Outlook
In this chapter, the mDoE concept for the combination of mathematical process
models with DoE was described. The most commonly used designs were examined,
and one representative study was described in detail. The role of a digital twin was
discussed and applied to a medium optimization case study using mDoE. A mathematical process model as a starting point for digital twin was adapted to four
experiments, and widely distributed boundary values for a DoE were evaluated
using model predictions instead of laboratory experiments. The reduced experimental spaces were experimentally performed (DoE) and compared to the simulated DoE
(mDoE). The same optimal conditions were found, and the further development in
different steps of a process development workflow was described. Finally, the
development of a digital twin and its use in mDoE can be seen as a useful tool in
decision-making for process development and optimization with DoE in QbD.
Statistical DoE can still be used for initial screening studies and can also lead to
process optimization in several rounds. Compared to conventional DoE, mDoE
supplies a more knowledge-based development of bioprocesses. Due to the mathematical model in mDoE, challenges in DoE can be avoided. The mathematical model
can be used for simulating the entire time trajectory with, e.g., metabolite formation
uptake. Hence, not only endpoints of experiments are examined. Thus the knowledge about the process can be increased. Furthermore, domain knowledge is
required and can be captured as additional constraints to the system, leading to a
focused screening or optimization of bioprocesses using the mathematical model as a
digital twin in mDoE.
Currently, the mDoE approach is tested for algae, yeasts, and cell culture. Further
applications of digital twins and mDoE can be seen in the field of cell therapeutics,
e.g., in the treatment of previously untreatable diseases as tumor diseases, brain
insult, and chronic infections. Since, the production of cells is still mainly performed
in static culture systems (e.g., T-flasks), it is difficult to provide a sufficient quantity
of patient-specific cells. A digital twin in combination with mDoE could be used to
build up an understanding of the process and, e.g., scale-up to enable fast and
efficient proliferation of stem and immune cells.
References
1. Nelson AL, Dhimolea E, Reichert JM (2010) Development trends for human monoclonal
antibody therapeutics. Nat Rev Drug Discov 9:767–774
2. Walsh G (2014) Biopharmaceutical benchmarks 2014. Nat Biotechnol 32:992–1000
3. Kretzmer G (2002) Industrial processes with animal cells. Appl Microbiol Biotechnol
59:135–142
4. Walsh G (2018) Biopharmaceutical benchmarks 2018. Nat Biotechnol 36:1136–1145
5. Chen C, Le H, Goudar CT (2016) Integration of systems biology in cell line and process
development for biopharmaceutical manufacturing. Biochem Eng J 107:11–17
56
K. B. Kuchemüller et al.
In this chapter, the mDoE concept for the combination of mathematical process
models with DoE was described. The most commonly used designs were examined,
and one representative study was described in detail. The role of a digital twin was
discussed and applied to a medium optimization case study using mDoE. A mathematical process model as a starting point for digital twin was adapted to four
experiments, and widely distributed boundary values for a DoE were evaluated
using model predictions instead of laboratory experiments. The reduced experimental spaces were experimentally performed (DoE) and compared to the simulated DoE
(mDoE). The same optimal conditions were found, and the further development in
different steps of a process development workflow was described. Finally, the
development of a digital twin and its use in mDoE can be seen as a useful tool in
decision-making for process development and optimization with DoE in QbD.
Statistical DoE can still be used for initial screening studies and can also lead to
process optimization in several rounds. Compared to conventional DoE, mDoE
supplies a more knowledge-based development of bioprocesses. Due to the mathematical model in mDoE, challenges in DoE can be avoided. The mathematical model
can be used for simulating the entire time trajectory with, e.g., metabolite formation
uptake. Hence, not only endpoints of experiments are examined. Thus the knowledge about the process can be increased. Furthermore, domain knowledge is
required and can be captured as additional constraints to the system, leading to a
focused screening or optimization of bioprocesses using the mathematical model as a
digital twin in mDoE.
Currently, the mDoE approach is tested for algae, yeasts, and cell culture. Further
applications of digital twins and mDoE can be seen in the field of cell therapeutics,
e.g., in the treatment of previously untreatable diseases as tumor diseases, brain
insult, and chronic infections. Since, the production of cells is still mainly performed
in static culture systems (e.g., T-flasks), it is difficult to provide a sufficient quantity
of patient-specific cells. A digital twin in combination with mDoE could be used to
build up an understanding of the process and, e.g., scale-up to enable fast and
efficient proliferation of stem and immune cells.
References
1. Nelson AL, Dhimolea E, Reichert JM (2010) Development trends for human monoclonal
antibody therapeutics. Nat Rev Drug Discov 9:767–774
2. Walsh G (2014) Biopharmaceutical benchmarks 2014. Nat Biotechnol 32:992–1000
3. Kretzmer G (2002) Industrial processes with animal cells. Appl Microbiol Biotechnol
59:135–142
4. Walsh G (2018) Biopharmaceutical benchmarks 2018. Nat Biotechnol 36:1136–1145
5. Chen C, Le H, Goudar CT (2016) Integration of systems biology in cell line and process
development for biopharmaceutical manufacturing. Biochem Eng J 107:11–17
56
K. B. Kuchemüller et al.