Initiation of Basic Research for Drug Repurposing Utilizing Fugaku

株式会社GEXVal
2023年1月20日

Initiation of Basic Research for Drug Repurposing Utilizing Fugaku

GEXVal Inc. (hereinafter "GEXVal") has started research on AI (Artificial Intelligence) analysis technology for drug repurposing using Fugaku.

This research is part of joint program with Dr. Shingo Tsuji, Specially Appointed Professor at the Research Center for Advanced Science and Technology, the University of Tokyo, and Dr. Koichi Hosomi, Professor, Faculty of Pharmaceutical Sciences, Kindai University, which started in July 2022 for drug repurposing.

The data-driven process of the Repurposing Engine^®, GEXVal's drug discovery technology platform, utilizes AI analysis technology to discover hidden relationships among drug candidates, their targets, and diseases. In order to utilize and integrate the large amount of diverse life science data and real-world data obtained from various platforms, a large number of dimensions and tremendous computation analysis power are required. In this research, we will use Fugaku, which has world-class computing power, to verify large-scale network embedding in graph neural network tools and optimize the algorithm.

GEXVal strives to generate extra value to deliver cost-effective, innovative treatment options to improve health and QoL of patients and families affected by rare and intractable diseases.

Fugaku
Fugaku is a supercomputer installed and operated at RIKEN Advanced Institute for Computational Science in Japan. It has the world's highest scores for calculation speed performance, execution performance of actual applications, AI performance, and big data analysis performance (As of June 2020)*.
*Reference: https://www.r-ccs.riken.jp/fugaku/about/

Graph neural network
Diverse data is represented as a graph with nodes (nodes) and edges (lines), and the data is analyzed as a graph neural network, which is one of the methods of deep learning. In drug discovery, it is a promising method for discovering relationships between drug candidate substances and their targets and diseases.

Rare diseases
Diseases with a very limited number of patients (Japan: less than 50,000; US: less than 200,000; Europe: less than 5 out of 10,000, etc.). There are more than 7,000 rare diseases in the world, totaling about 350 million people are said to be affected.

Challenges of rare diseases
Therapeutic drugs are available for only about 5% of rare diseases, and many of the drugs approved in recent years, such as cell therapy, gene therapy, and antibody drugs, are extremely expensive and have a considerable impact on medical health policy. There are serious access-to-medicine challenges, with many not readily available to all patients.

Challenges of the pharmaceutical industry
The success rate in the research and development of new drugs is very low, and pharmaceutical companies face productivity challenges. Despite the significant amount of investment in research and development, many new drug-candidate substances have been discontinued without being fully exploited.

Drug repurposing
Also known as “drug reprofiling” or “drug repositioning,” drug repurposing is a method for discovering and developing the potential of already marketed drugs or drug candidate substances under research and development for diseases different than initially assumed. As data on safety, pharmacokinetics, manufacturing methods, etc. have been accumulated to some extent, it is possible to shorten the development timeline and reduce development costs, and it is expected to contribute to the improvement of research and development productivity.

Repurposing Engine^®
This is GEXVal's unique innovation engine for strategic drug repurposing research. We will find the best answer for R&D strategy aiming at a higher probability of success. We reexamine drug candidate substances whose development has been discontinued and develop and implement the necessary strategies to contribute to the medical field.

有关本事宜请咨询：info@gexval.com