Robust Prediction of the Dynamics of Biocircuits using Integrated First-Principles and Data-Driven Models

2021 Fiscal Year Final Research Report

• Project/Area Number: 18H01464
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Review Section: Basic Section 21040:Control and system engineering-related
• Research Institution: Keio University
• Principal Investigator: Hori Yutaka 慶應義塾大学, 理工学部(矢上), 准教授 (10778591)
• Project Period (FY): 2018-04-01 – 2022-03-31
• Keywords: 制御工学 / 生体分子システム / マイクロ流体

Outline of Final Research Achievements

We developed a theoretical framework for analyzing and designing the dynamics of biomolecular systems using a mathematical model, and an experimental platform that efficiently supplements data necessary for constructing such models. Specifically, we developed a modeling framework that combines the first-principles model derived from physical laws with a machine learning model learned from experimental data. The proposed framework enabled robust prediction and analysis of systems' dynamics that takes into account the influence of environmental factors, which was difficult to capture with the first-principles model alone. A microfluidic system was also developed to generate a large number of parallel reaction systems with slightly different environmental factors and use for model identification.

Free Research Field

制御工学

Academic Significance and Societal Importance of the Research Achievements

生体分子システムの持つ様々な不確かさに対してロバスト性の高いモデルベースの反応設計法を提案したことで，多分子・多機能の分子システムのモデルベース開発が容易になる．これにより，反応ネットワークの作用機序をボトムアップ的なアプローチで探求する学術研究や，分子システムの工学応用を目指す研究開発が加速すると期待される．
また，生体分子の反応ネットワークが持つ数理的な構造や不確かさを，制御工学や機械学習の理論と関連させてシステム論的に捉えるためのフレームワークを提案した点は，システム科学の発展においても学術的な意義がある．