Development of fundamental technologies for data-driven optimisation of cell culture

2023 Fiscal Year Final Research Report

• Project/Area Number: 21K19815
• Research Category: Grant-in-Aid for Challenging Research (Exploratory)
• Allocation Type: Multi-year Fund
• Review Section: Medium-sized Section 62:Applied informatics and related fields
• Research Institution: University of Tsukuba
• Principal Investigator: YING Bei-Wen 筑波大学, 生命環境系, 准教授 (90422401)
• Project Period (FY): 2021-07-09 – 2024-03-31
• Keywords: 細胞培養 / 機械学習 / 培地 / データサイエンス

Outline of Final Research Achievements

This study was to address issues pertaining to the reproducibility and productivity of cell culture. A data science approach was employed to improve cell culture, which is commonly based on individual senses. Through a series of comprehensive experiments, a substantial amount of data was gathered, linking specific culture conditions with the resulting quality of the culture. The dataset was then used to construct the database of ‘environmental information- cell growth’. Machine learning was applied to the experimental data in order to predict which environmental factors contribute to cell culture and to what extent, as well as to identify their relative importance. Active learning (repeated experiments and learning) was employed to enhance the prediction accuracy of the learning model and optimise the culture medium for cell culture.The culture experiments and machine learning methods established in this study can be applied to a variety of cell cultures.

Free Research Field

システム生物学

Academic Significance and Societal Importance of the Research Achievements

細胞培養は基礎研究においても、健康と医療のための産業応用においても、普遍的技術であるにも拘わらず、細胞培養の再現性、安定性、安全性の課題が長年に渡って未解決のままである。本研究はアナログ方式で行われている細胞培養に、デジタル方式である応用情報科学的研究手法を取り入れることにより、細胞増殖に対する高度な制御を実現し、細胞培養の諸問題を解決することに繋がる。本研究は、機械学習・人工知能が職人芸的な細胞培養にある根幹問題を解決する初めての挑戦であり、その成功が生命科学研究において抜本的なトレンドチェンジとなる。本研究で検証される方法論が細胞培養の不確実性を減らし、再生医療の産業化に寄与する。