Development of a Data and Model Platform to Support the Sharing and Reuse of MD Data

Category

Name and affiliation of Research Director

Outline of R&D

We will construct a molecular dynamics (MD) simulation database to accelerate AI-driven life science research. This database will provide experimental life scientists with an environment to view protein dynamics via web browsers, computational scientists with a platform to register and share simulation data in a standardized manner, and AI/machine learning researchers with ML-Ready training datasets and pre-trained surrogate models. While functioning as a data hub for Japan and Asia in collaboration with the European MDDB, this database will systematically organize multiscale data via enhanced sampling and coarse-grained MD, as well as implement unique added values specialized for AI training, such as benchmark and leaderboard features. By realizing an MD data utilization environment where anyone can access protein dynamics information, we aim to contribute to structural biology and drug discovery research.

Main database(s) subject to research and development

MDDB-AI Hub (Under development)

Period of research and development

Apr 2026 to Mar 2029

Grant Number

JPMJND2603