Construction of a Multimodal Database Supporting AI-Driven RNA Drug Discovery

Category

Name and affiliation of Research Director

Outline of R&D

We will construct an integrated database to support AI-driven RNA drug discovery by consolidating and standardizing multimodal data across five levels and 17 categories—including RNA structure, modifications, interactions, expression, and disease relevance—from the entire human transcriptome. We will develop a new deep learning model trained on public data, to predict binding affinities between RNA and small molecules. We will implement a precomputed architecture to enable rapid search and visualization of druggability and off-target risks for target sites. By centralizing previously dispersed data resources and significantly streamlining processes from target discovery to safety assessment, we aim to accelerate next-generation drug development for cancer and intractable diseases while enhancing international competitiveness.

Main database(s) subject to research and development

RNATDB (Under development)

Period of research and development

Apr 2026 to Mar 2029

Grant Number

JPMJND2602