東京大学とちとせ研究所、抗体医薬品製造プロセスのデジタル設計に新展開～自動化実験データをフル活用した培養工程の数理モデル構築～

2026-02-12 東京大学

自動化実験と数理モデリングの融合によるデジタル設計

<関連情報>

• https://www.t.u-tokyo.ac.jp/press/pr2026-02-12-002
• https://aiche.onlinelibrary.wiley.com/doi/10.1002/aic.70231

小規模自動化実験プラットフォームを用いたmAb生産における細胞培養の最適化のためのハイブリッドモデリング Hybrid modeling for optimizing cell cultivation in mAb production using a small-scale automated experimental platform

Kosuke Nemoto, Yuki Yoshiyama, Mizuki Morisasa, Junshin Iwabuchi, Yusuke Hayashi, Sara Badr, Hirokazu Sugiyama
AIChE Journal  Published: 11 February 2026
DOI:https://doi.org/10.1002/aic.70231

Abstract

This work presents the optimization of cell cultivation for monoclonal antibody (mAb) production. We developed a hybrid model describing the effects of multiple process variables on antibody productivity and impurity generation. An automated platform with 12 × 250 mL bioreactors was set up. Fed-batch cultivation experiments were conducted using a newly developed CHO-MK cell line. Six process variables were varied over a wide range, and experimental data under 29 different cultivation conditions were used for model development. An average R² value of 0.86 across all components and conditions was achieved with model fitting. Additional validation experiments under 15 cultivation conditions confirmed the model’s high prediction accuracy. The model was then applied to identify optimal conditions that maximize mAb yield while constraining impurity concentrations. Based on the results, a design recommendation was proposed and validated experimentally. This study demonstrates that integrating model-based simulations with automated experiments is effective for optimizing cell cultivation conditions.