[16p-PB02-1~15] 12.6 Nanobiotechnology

[16p-PB02-14] Machine learning assisted improvement of evaluation scheme of FMO-DPD

〇Sota Matsuoka¹, Hideo Doi¹, Koji Okuwaki¹, Ryo Hatada¹, Sojiro Minami¹, Ryosuke Suhara¹, Yuji Mochizuki^1,2 (1.Rikkyo Univ., 2.Univ. Tokyo)

Keywords:machine learning, FMO, DPD

The effective interaction (χ) parameters in dissipative particle dynamics (DPD) simulations can be evaluated by using the fragment molecular orbital (FMO) based Chi-parameter Evaluation Workflow System (FCEWS). The total cost of FCEWS usage is completely dominated by FMO calculations, thus its reduction is of special interest. In this study, we have tried to improve the efficiency of evaluations with machine learning technique and developed a workflow system (pre_fcews).