Prediction of rare events associated with biomolecules using mean force dynamics

2021 Fiscal Year Final Research Report

• Project/Area Number: 17K05620
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Biological physics/Chemical physics/Soft matter physics
• Research Institution: National Institute of Advanced Industrial Science and Technology
• Principal Investigator: Morishita Tetsuya 国立研究開発法人産業技術総合研究所, 材料・化学領域, 主任研究員 (10392672)
• Co-Investigator(Kenkyū-buntansha): 米澤 康滋 近畿大学, 先端技術総合研究所, 教授 (40248753)
• Project Period (FY): 2017-04-01 – 2022-03-31
• Keywords: 自由エネルギー / レア・イベント / 分子動力学 / 機械学習

Outline of Final Research Achievements

We have proposed an approach that incorporates time dependence in principal component analysis to identify each of step-by-step structural changes in molecular dynamics (MD) simulations (TDPCA). The time dependence allows for re-optimization of the principal components according to the structural development, which can be exploited for enhanced sampling in MD simulations. We have also developed algorithms that allow for reconstructing free-energy profiles along collective variables (CVs) by extending the logarithmic mean-force (LogMFD) dynamics. The logarithmic parallel-dynamics (LogPD) algorithm was proposed to efficiently reconstruct free-energy profiles using multiple replicas of the system based on Crooks-Jarzynski nonequilibrium work relation. Also, isokinetic dynamics was incorporated in LogMFD/PD to enable uniform sampling along the CVs. These new approaches allow for on-the-fly free-energy estimation, which can be implemented using an open-source library, PLUMED.

Free Research Field

分子動力学計算による化学物理現象の解明及び計算手法開発

Academic Significance and Societal Importance of the Research Achievements

本研究では、分子シミュレーションが抱える時間スケールに関する問題に取り組み、従来では実現不可能な長時間現象を、限られた時間内において実現できる手法開発を推進した。これにより、複雑な構造を有する生体分子やナノ物質の原子レベルの長時間挙動を予測することが可能となり、ナノ材料設計や分子材料設計、ひいては創薬におけるシミュレーションによる開発促進に繋がることが期待される。