Estimating susceptibility of rodenticide by molecular dynamics and machine learning using 3D structures of vitamin k epoxide reductase

Name: Takeda, Kazuki

Organization : Kitasato University, School of Veterinary Medicine

Email : takeda@vmas.kitasato-u.ac.jp

Name: Yasuo, Nobuaki

Organization : Tokyo Institute of Technology, Department of Computer Science

Name: Watanabe, Kanami

Organization : Kitasato University, School of Veterinary Medicine

Name: Sato, Ryo

Organization : Kitasato University, School of Veterinary Medicine

Name: Kamata, Ryo

Organization : Kitasato University, School of Veterinary Medicine

Name: Sekijima, Masakazu

Organization : Tokyo Institute of Technology, Department of Computer Science

LCSH: Rodenticides -- Susceptibility -- Experimentation

LCSH: Machine learning

LCSH: Vitamin K

Abstract: Anticoagulant chemicals such as warfarin have been widely used as pharmaceuticals or rodenticides. Their mode of action is inhibition of reduction of vitamin K epoxide (inactive) to vitamin K (active) by vitamin K epoxide reductase (VKOR). One of pharmacological/toxicological interest of anticoagulant chemicals is the resistance or hypersusceptibility of VKOR to them that is conferred by intra/inner species genetic diversity of VKOR. It is necessary to assess the susceptibility of VKOR variants for proper use of the anticoagulants. In this study, we attempted to construct a method to evaluate the susceptibility from the polymorphisms by molecular dynamics simulation and machine learning. The human and rat VKOR 3D structures were constructed using Alphafold2. Mutagenesis and the structure optimization, molecular docking with warfarin, and molecular dynamics simulations were performed using Schrödinger Suites. Molecular docking showed that many mutants had same score, indicating the effect of a single amino acid mutation cannot be properly assessed by the docking. On the contrary, the binary discriminant model of resistance/susceptibility using features obtained from molecular dynamics showed relatively high F1 score, even for simple support vector machine. These results suggest that molecular dynamics are useful for assessing the chemical susceptibility of proteins depended on their structure

King Mongkut's University of Technology North Bangkok. Central Library

Address: BANGKOK

Email: library@kmutnb.ac.th

Created: 2022

Modified: 2022-11-25

Issued: 2022-11-25

บทความ/Article

application/pdf

BibliograpyCitation : In King Mongkut's Institute of Technology Ladkrabang. Pure and Applied Chemistry International Conference (PACCON 2022) (pp.971-975). Bangkok : King Mongkut's Institute of Technology Ladkrabang

eng

©copyrights King Mongkut's University of Technology North Bangkok

ลำดับที่.	ชื่อแฟ้มข้อมูล	ขนาดแฟ้มข้อมูล	จำนวนเข้าถึง	วัน-เวลาเข้าถึงล่าสุด
1	PACCON 2022pp.971-975.pdf	1.38 MB

ใช้เวลา

0.047138 วินาที