Data-driven and physics informed modeling of Chinese Hamster Ovary cell bioreactors

Tianqi Cui; Tom Bertalan; Nelson Ndahiro; Pratik Khare; Michael Betenbaugh; Costas Maranas; Ioannis G. Kevrekidis

doi:10.1016/j.compchemeng.2024.108594

Data-driven and physics informed modeling of Chinese Hamster Ovary cell bioreactors

Tianqi Cui
, Tom Bertalan
, Nelson Ndahiro
, Pratik Khare
, Michael Betenbaugh
, Costas Maranas
, Ioannis G. Kevrekidis

Research output: Contribution to journal › Article › peer-review

Abstract

Fed-batch culture is an established operation mode for the production of biologics using mammalian cell cultures. Quantitative modeling integrates both kinetics for some key reaction steps and optimization-driven metabolic flux allocation, using flux balance analysis; this is known to lead to certain mathematical inconsistencies Here, we propose a physically-informed data-driven hybrid model (a “gray box”) to learn models of the dynamical evolution of Chinese Hamster Ovary (CHO) cell bioreactors from process data The approach incorporates physical laws (e.g.

Original language	English (US)
Article number	108594
Journal	Computers and Chemical Engineering
Volume	183
DOIs	https://doi.org/10.1016/j.compchemeng.2024.108594
State	Published - Apr 2024
Externally published	Yes

Keywords

Black/gray-box identification
Differentiable optimization
Machine Learning
Metabolic flux analysis
Neural ODEs

ASJC Scopus subject areas

General Chemical Engineering
Computer Science Applications

Access to Document

10.1016/j.compchemeng.2024.108594

Cite this

@article{bf63b41da8a4435f9a97ee954cc0a666,

title = "Data-driven and physics informed modeling of Chinese Hamster Ovary cell bioreactors",

abstract = "Fed-batch culture is an established operation mode for the production of biologics using mammalian cell cultures. Quantitative modeling integrates both kinetics for some key reaction steps and optimization-driven metabolic flux allocation, using flux balance analysis; this is known to lead to certain mathematical inconsistencies Here, we propose a physically-informed data-driven hybrid model (a {\textquotedblleft}gray box{\textquotedblright}) to learn models of the dynamical evolution of Chinese Hamster Ovary (CHO) cell bioreactors from process data The approach incorporates physical laws (e.g.",

keywords = "Black/gray-box identification, Differentiable optimization, Machine Learning, Metabolic flux analysis, Neural ODEs",

author = "Tianqi Cui and Tom Bertalan and Nelson Ndahiro and Pratik Khare and Michael Betenbaugh and Costas Maranas and Kevrekidis, \{Ioannis G.\}",

note = "Publisher Copyright: {\textcopyright} 2024 Elsevier Ltd",

year = "2024",

month = apr,

doi = "10.1016/j.compchemeng.2024.108594",

language = "English (US)",

volume = "183",

journal = "Computers and Chemical Engineering",

issn = "0098-1354",

publisher = "Elsevier B.V.",

}

TY - JOUR

T1 - Data-driven and physics informed modeling of Chinese Hamster Ovary cell bioreactors

AU - Cui, Tianqi

AU - Bertalan, Tom

AU - Ndahiro, Nelson

AU - Khare, Pratik

AU - Betenbaugh, Michael

AU - Maranas, Costas

AU - Kevrekidis, Ioannis G.

PY - 2024/4

Y1 - 2024/4

N2 - Fed-batch culture is an established operation mode for the production of biologics using mammalian cell cultures. Quantitative modeling integrates both kinetics for some key reaction steps and optimization-driven metabolic flux allocation, using flux balance analysis; this is known to lead to certain mathematical inconsistencies Here, we propose a physically-informed data-driven hybrid model (a “gray box”) to learn models of the dynamical evolution of Chinese Hamster Ovary (CHO) cell bioreactors from process data The approach incorporates physical laws (e.g.

AB - Fed-batch culture is an established operation mode for the production of biologics using mammalian cell cultures. Quantitative modeling integrates both kinetics for some key reaction steps and optimization-driven metabolic flux allocation, using flux balance analysis; this is known to lead to certain mathematical inconsistencies Here, we propose a physically-informed data-driven hybrid model (a “gray box”) to learn models of the dynamical evolution of Chinese Hamster Ovary (CHO) cell bioreactors from process data The approach incorporates physical laws (e.g.

KW - Black/gray-box identification

KW - Differentiable optimization

KW - Machine Learning

KW - Metabolic flux analysis

KW - Neural ODEs

UR - https://www.scopus.com/pages/publications/85185172566

UR - https://www.scopus.com/pages/publications/85185172566#tab=citedBy

U2 - 10.1016/j.compchemeng.2024.108594

DO - 10.1016/j.compchemeng.2024.108594

M3 - Article

AN - SCOPUS:85185172566

SN - 0098-1354

VL - 183

JO - Computers and Chemical Engineering

JF - Computers and Chemical Engineering

M1 - 108594

ER -

Data-driven and physics informed modeling of Chinese Hamster Ovary cell bioreactors

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this