Hybrid Deep Modeling of a GS115 (Mut+) <i>Pichia pastoris</i> Culture with State–Space Reduction
Hybrid modeling workflows combining machine learning with mechanistic process descriptions are becoming essential tools for bioprocess digitalization. In this study, a hybrid deep modeling method with state–space reduction was developed and showcased with a <i>P. pastoris</i> GS115 Mut+...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-07-01
|
| Series: | Fermentation |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2311-5637/9/7/643 |
| _version_ | 1827733073148510208 |
|---|---|
| author | José Pinto João R. C. Ramos Rafael S. Costa Rui Oliveira |
| author_facet | José Pinto João R. C. Ramos Rafael S. Costa Rui Oliveira |
| author_sort | José Pinto |
| collection | DOAJ |
| description | Hybrid modeling workflows combining machine learning with mechanistic process descriptions are becoming essential tools for bioprocess digitalization. In this study, a hybrid deep modeling method with state–space reduction was developed and showcased with a <i>P. pastoris</i> GS115 Mut+ strain expressing a single-chain antibody fragment (scFv). Deep feedforward neural networks (FFNN) with varying depths were connected in series with bioreactor macroscopic material balance equations. The hybrid model structure was trained with a deep learning technique based on the adaptive moment estimation method (ADAM), semidirect sensitivity equations and stochastic regularization. A state–space reduction method was investigated based on a principal component analysis (PCA) of the cumulative reacted amount. Data of nine fed-batch <i>P. pastoris</i> 50 L cultivations served to validate the method. Hybrid deep models were developed describing process dynamics as a function of critical process parameters (CPPs). The state–space reduction method succeeded to decrease the hybrid model complexity by 60% and to improve the predictive power by 18.5% in relation to the nonreduced version. An exploratory design space analysis showed that the optimization of the feed of methanol and of inorganic elements has the potential to increase the scFv endpoint titer by 30% and 80%, respectively, in relation to the reference condition. |
| first_indexed | 2024-03-11T01:05:54Z |
| format | Article |
| id | doaj.art-a8dade0c8dcd492e99020e271c3d7318 |
| institution | Directory Open Access Journal |
| issn | 2311-5637 |
| language | English |
| last_indexed | 2024-03-11T01:05:54Z |
| publishDate | 2023-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Fermentation |
| spelling | doaj.art-a8dade0c8dcd492e99020e271c3d73182023-11-18T19:16:20ZengMDPI AGFermentation2311-56372023-07-019764310.3390/fermentation9070643Hybrid Deep Modeling of a GS115 (Mut+) <i>Pichia pastoris</i> Culture with State–Space ReductionJosé Pinto0João R. C. Ramos1Rafael S. Costa2Rui Oliveira3LAQV-REQUIMTE, Department of Chemistry, NOVA School of Science and Technology, NOVA University Lisbon, 2829-516 Lisbon, PortugalLAQV-REQUIMTE, Department of Chemistry, NOVA School of Science and Technology, NOVA University Lisbon, 2829-516 Lisbon, PortugalLAQV-REQUIMTE, Department of Chemistry, NOVA School of Science and Technology, NOVA University Lisbon, 2829-516 Lisbon, PortugalLAQV-REQUIMTE, Department of Chemistry, NOVA School of Science and Technology, NOVA University Lisbon, 2829-516 Lisbon, PortugalHybrid modeling workflows combining machine learning with mechanistic process descriptions are becoming essential tools for bioprocess digitalization. In this study, a hybrid deep modeling method with state–space reduction was developed and showcased with a <i>P. pastoris</i> GS115 Mut+ strain expressing a single-chain antibody fragment (scFv). Deep feedforward neural networks (FFNN) with varying depths were connected in series with bioreactor macroscopic material balance equations. The hybrid model structure was trained with a deep learning technique based on the adaptive moment estimation method (ADAM), semidirect sensitivity equations and stochastic regularization. A state–space reduction method was investigated based on a principal component analysis (PCA) of the cumulative reacted amount. Data of nine fed-batch <i>P. pastoris</i> 50 L cultivations served to validate the method. Hybrid deep models were developed describing process dynamics as a function of critical process parameters (CPPs). The state–space reduction method succeeded to decrease the hybrid model complexity by 60% and to improve the predictive power by 18.5% in relation to the nonreduced version. An exploratory design space analysis showed that the optimization of the feed of methanol and of inorganic elements has the potential to increase the scFv endpoint titer by 30% and 80%, respectively, in relation to the reference condition.https://www.mdpi.com/2311-5637/9/7/643hybrid modelingdeep learningADAM method<i>Pichia pastoris</i> GS115 Mut+single-chain antibody fragment (scFv)bioprocess digitalization |
| spellingShingle | José Pinto João R. C. Ramos Rafael S. Costa Rui Oliveira Hybrid Deep Modeling of a GS115 (Mut+) <i>Pichia pastoris</i> Culture with State–Space Reduction Fermentation hybrid modeling deep learning ADAM method <i>Pichia pastoris</i> GS115 Mut+ single-chain antibody fragment (scFv) bioprocess digitalization |
| title | Hybrid Deep Modeling of a GS115 (Mut+) <i>Pichia pastoris</i> Culture with State–Space Reduction |
| title_full | Hybrid Deep Modeling of a GS115 (Mut+) <i>Pichia pastoris</i> Culture with State–Space Reduction |
| title_fullStr | Hybrid Deep Modeling of a GS115 (Mut+) <i>Pichia pastoris</i> Culture with State–Space Reduction |
| title_full_unstemmed | Hybrid Deep Modeling of a GS115 (Mut+) <i>Pichia pastoris</i> Culture with State–Space Reduction |
| title_short | Hybrid Deep Modeling of a GS115 (Mut+) <i>Pichia pastoris</i> Culture with State–Space Reduction |
| title_sort | hybrid deep modeling of a gs115 mut i pichia pastoris i culture with state space reduction |
| topic | hybrid modeling deep learning ADAM method <i>Pichia pastoris</i> GS115 Mut+ single-chain antibody fragment (scFv) bioprocess digitalization |
| url | https://www.mdpi.com/2311-5637/9/7/643 |
| work_keys_str_mv | AT josepinto hybriddeepmodelingofags115mutipichiapastorisiculturewithstatespacereduction AT joaorcramos hybriddeepmodelingofags115mutipichiapastorisiculturewithstatespacereduction AT rafaelscosta hybriddeepmodelingofags115mutipichiapastorisiculturewithstatespacereduction AT ruioliveira hybriddeepmodelingofags115mutipichiapastorisiculturewithstatespacereduction |