Skip to content
CAAIL

CAAIL is in beta. We're re-checking earlier classifications and reorganizing the research areas for v2. See the roadmap Tell us what you think

CHO Reference (Chinese Hamster Ovary)

The Chinese Hamster Ovary (CHO) cell line is the dominant mammalian host for recombinant-protein biomanufacturing. CHO is not a cellular agriculture species, but it is the closest mature analogue for cell-ag process modeling: its genome-scale metabolic models are the most-developed mammalian-cell GEM ecosystem available, and its biomass parameterization, perfusion-process methodology, and model-reduction techniques translate directly to the cultivated-meat GEMs (porcine, avian, salmonid) catalogued on the per-species pages of this directory. This page collects the CHO reference reconstructions as a biopharma-adjacent substrate.

Genome-scale metabolic models

iCHO1766 / iCHO2048 / CHOmpact — Chinese Hamster Ovary (biopharma-adjacent reference)

The CHO cell line is the mammalian biopharma workhorse, and its GEM family is the most-developed mammalian GEM ecosystem available: Hefzi et al.’s iCHO1766 (2016, Cell Systems) is the consensus reconstruction; iCHO2048s (Gutierrez et al. 2020, Nature Communications) adds the secretory pathway; CHOmpact (2023, Biotechnology and Bioengineering) and follow-on Bayesian-flux-estimation pipelines (2025) produce reduced models for digital-twin work. CHO is not itself a cellular agriculture species, but its biomass parameterization, perfusion-process methodology, and reduction techniques translate directly to the cell-ag GEMs on the per-species pages and to those still being built. iCHO1766 is downloadable from BiGG Models, where it is listed as iCHOv1, and from CHOgenome.org, which also carries the cell-line-specific CHO-K1, CHO-S and DG44 models. The iCHO2048s models for individual secreted products (MATLAB .mat) and the Python script that builds them are in the authors’ GitHub repository (LewisLabUCSD/MammalianSecretoryRecon, which has no licence file). CHOmpact’s reactions, stoichiometric matrix and biomass composition are Supplementary Tables 1 to 3 of its paper’s supplementary workbook (10.1002/bit.28459), as tables rather than an SBML file.

Reference: Papers.md #85 (Hefzi et al. 2016, Cell Systems).

Curation source: This entry is long-standing CAAIL curation, migrated from the prior flat Datasets.md. CHO is a biopharma-adjacent reference rather than a cultivated-meat species, so it is not drawn from the Todhunter et al. 2024 supplemental.

Report an issue

Cell culture media training datasets

CHO-GS(-/-) media-component → critical quality attributes (Gangwar et al. 2024)

A supervised-learning training table covering CHO-GS(-/-) cell culture in chemically defined media of varying metal-ion composition and the resulting critical quality attributes (CQAs). Generated by Gangwar, Balraj, & Rathore (2024, Applied Microbiology and Biotechnology 108(1), 308) to train an end-to-end explainable-AI framework for media-component selection and CQA prediction; the tree-ensemble feature-attribution analysis (XGBoost / Gradient Boosting / Decision Trees / Random Forest / CatBoost with SHAP) is what places Papers.md #170 in the matrix’s Ensemble Learning × Media Optimization cell. The training data are in the paper’s supplementary file (ESM1): Table S1 gives Cu, Fe, Zn, Mn, Mg, Co and Ni concentrations (ppm) with acidic and basic charge-variant percentages for 34 formulations, and Table S2 a further 42 rows varying Fe and Zn; there is no repository accession.

cited by22Report an issue

Multi-omics & metabolomics datasets

These CHO datasets pair antibody-producing cell-culture runs with omics readouts: multi-omic time courses and clone-level metabolomic biomarker statistics. They are the measurement layer that constrains and validates the mammalian-cell GEMs above, and the same profiling logic transfers to the cell-ag livestock GEMs (porcine, avian, salmonid) on the per-species pages: which nutrients limit growth, how metabolism reprograms across a run, and which metabolic signatures separate a productive line from an unproductive one.

CHO multi-omic culture-pH profiling dataset (Lee et al. 2021)

A multi-omics profiling of a CHO cell-culture system that resolves how culture pH shifts cell growth, antibody titer, and product quality (Lee, Kok, Lakshmanan, et al. 2021, Biotechnology and Bioengineering). The paired transcriptomic, proteomic, and metabolomic layers make this a parameterization and validation substrate for mammalian-cell GEMs and for process models that couple pH set-points to CQAs, the same growth-versus-quality trade-off cell-ag bioprocess models have to capture. The Supporting Information workbook (bit27899-sup-0004-6_Suppl_Tables.xlsx) gives per-sample normalized values and differential-expression results for about 16,700 transcripts, 1,500 proteins, 380 metabolites and 150 lipids; there is no repository deposit. Published: 10.1002/bit.27899.

cited by40Report an issue

CHO antibody-producer multi-omic dataset (Gopalakrishnan et al. 2024)

A multi-omic characterization of antibody-producing CHO cell lines that maps metabolic reprogramming and pinpoints nutrient-uptake bottlenecks across production (Gopalakrishnan, Johnson, Valderrama-Gomez, et al. 2024, Metabolic Engineering, from the Lewis lab). The nutrient-limitation and flux-reprogramming signals are directly usable for constraining GEM flux states and for designing feeds, methodology that carries over to media and feed design for cultivated-meat cell lines. Its supplementary workbook (mmc1.xlsx) gives phase-specific uptake and secretion rates per metabolite for each cell line and replicate (ST2), mean fluxes for about 6,700 reactions (ST4) and per-sample expression for about 1,770 genes (ST5). Published: 10.1016/j.ymben.2024.07.009.

cited by14Report an issue

CHO clone-selection productivity biomarkers (Barberi et al. 2025)

Productivity-biomarker statistics from a metabolomic-dynamics study supporting machine-learning selection of productive CHO cell lines during biopharma process development (Barberi, Benedetti, Diaz-Fernandez, et al. 2025, AIChE Journal). The time-resolved metabolomic trajectories behind them, used to classify high- from low-producing clones, are a methodology template for metabolomics-guided line selection in cultivated-meat cell-line development, where picking a stable high-performing line early is a core bottleneck. The metabolomic data themselves are confidential, as the paper’s data statement says; its supplementary material lists about 90 productivity-biomarker ions with m/z, time point, VIP, covariance, correlation and p value (SM1_extracellular_productivity_biomarkers), along with the numerical data behind Figures 2 to 4. Published: 10.1002/aic.18602.

cited by3Report an issue

Bioprocess characterization datasets

This section holds CHO bioprocess measurements at the resolution constraint-based and kinetic models need: process-mode uptake and secretion rates that parameterize genome-scale models. The cell-ag GEMs on the per-species pages take measurements of the same kind.

CHO process-mode uptake and secretion rates (Venkatarama Reddy et al. 2026)

Uptake and secretion rates from a CHO-VRC01 flux study spanning fed-batch and perfusion process modes across bioreactor pH, media supplementation, and feeding strategies (Venkatarama Reddy et al. 2026, Metabolic Engineering), paired with a kinetically constrained dynamic flux balance analysis model that regresses kinetic parameters across amino-acid metabolism and cell growth. Raw data and model code are available from the authors on request; the supplementary file (mmc1.docx) gives, in Table S1, uptake and secretion rates per metabolite at pH 6.75, 7 and 7.25 across three culture phases, the kind of rates that parameterize GEMs for cell-ag digital twins. Published: 10.1016/j.ymben.2026.102552 (preprint 10.64898/2026.01.11.698917).

cited by1Report an issue

Data on request

Studies whose own data availability statement says their data are shared only on request. They have no public deposit or supplementary results table, so they are not counted as datasets; they are listed so the work stays findable.

StudyPaperData availability statement
Pseudo Perfusion of Chinese Hamster Ovary (CHO) Cells as a Reliable Platform for Data Generation to Model and Guide Continuous Perfusion Biomanufacturing (Malinov et al. 2026)DOI”Data will be made available upon reasonable request.”
Comprehensive modeling of cell culture profile using Raman spectroscopy and machine learning (Tanemura et al. 2023)DOI”The data that support the findings of this study are available from the corresponding author upon reasonable request. Additional data are available in the supplementary material of this article.”

Further reading

Linked external resources are independent of TUCCA and Tufts University and remain under their own licenses.