[23p-1BN-1~13] 12.6 Nanobiotechnology

5:00 PM - 5:15 PM JST(8:00 AM - 8:15 AM UTC)Presentation by Applicant for JSAP Young Scientists Presentation Award

[23p-1BN-13] Calculation of protein residue interaction parameters by machine learning

〇Sota Matsuoka¹, Hideo Doi¹, Yusuke Tachino¹, Koji Okuwaki^1,2, Yoshinori Hirano³, Yuji Mochizuki^1,4 (1.Rikkyo Univ., 2.JSOL Corp., 3.Keio Univ., 4.Univ. Tokyo)

Keywords:

Machine Learning,FMO,DPD

Effective interaction parameters (χ) in dissipative particle dynamics (DPD) simulations can be calculated empirically using FCEWS (FMO-based Chi-parameter Evaluation Workflow System). However, FCEWS has a significant cost of use, and reducing this cost is an important issue. Therefore, we attempted to improve the calculation of χ using machine learning by implementing a preprocessing system called pre_fcews. Currently, we focused on chignolin, which consists of 10 amino acid residues, and were able to confirm the folding structure of chignolin in the DPD using the predicted value of χ.