Korea Steps Up AI Adoption In Drug Discovery, Repositioning And Surveillance
Executive Summary
South Korea is stepping up the adoption of artificial intelligence to speed up drug development by selecting discovery, repositioning and smart drug surveillance as three initial focus areas for new AI/big data platforms.
South Korea’s Ministry of Science and ICT and Ministry of Health and Welfare have begun a joint project to invest a total of KRW25.8bn ($22m) over the next three years to establish new customized artificial intelligence (AI) and big data platforms to support drug substance discovery, product repositioning and smart drug surveillance systems.
The discovery platform aims to quickly identify new substances using AI that has learned from test results and research papers. The repositioning system will propose new uses by isolating additional effects of drugs that have already verified safety, while the surveillance platform will be able to predict side effects before occurrence by learning about past adverse effect cases.
The project is a follow-up to an earlier official plan announced in South Korea to establish five big data platforms in the medical area: national big data, comprising genomic and other medical information on up to one million people; data-oriented hospitals; information on new drug candidates; biologic patents; and big data gathered from public institutions. These form part of the country’s strategic goal of nurturing the bio-health sector as one of three main economic growth engines.
Reducing Lead Times
If successful, the new joint project is expected to help reduce drug development times by as much as by half, or by up to seven or eight years. The platforms will be open and free for use by researchers and companies.
“The next few years will be a time for a new leap forward in new drug development. The government will support convergence of new drug development and AI to generate successful cases in the bio-health sector,” the science ministry declared.
Although the pharmaceutical sector is seen as a promising industry with potential to generate massive revenues if a blockbuster drug is developed, huge R&D costs and long development times are seen as entry barriers for South Korean companies. Use of AI technology has emerged as one of the strategies to overcome these barriers.
Since last year, the government and pharma industry have been taking a number of moves to step up the adoption of AI and big data in new drug development. (Also see "How Can Korea Grab Window Of Opportunity In AI Drug Development? " - Scrip, 11 Jun, 2019.)
AI Drug Development Platforms
Four teams, including the AI-driven drug discovery startup Arontier, will progress the development of an AI platform for substance discovery. Arontier will use genomic and liver organoid experimental data from the Asan Medical Center to develop a platform for lung and brain cancer therapies. Based on chemical data held by the Korea Research Institute of Chemical Technology, Chung-Ang University plans to develop a specialized platform for degenerative brain disease through prediction of proteins, which is seen as providing cues for the disease treatment.
Other teams will develop a platform for targeted cancer therapies as well as a cloud-based platform using supercomputing infrastructure to research anticancer and fibrosis treatments.
To support the drug repositioning initiative, a research team at the Korea Advanced Institute of Science and Technology will develop a deep learning model for existing mechanisms of action and advance the platform through experimental verification. The Asan Medical Center will also help to set up a drug surveillance support system to rapidly predict and deal with adverse effects through collection of big data on immuno-oncology drugs.
According to the Asan center, the team will expand a common data model and analyze big data based on electronic hospital records, genomic information, patient outcome reports, diagnosis and prescription data. Another facet will involve the exploration of genomic biomarkers to predict the adverse effects of IO therapies, helped by AI algorithms on through clinical data and genomic-based learning.
The project will also develop a smart drug surveillance platform that includes prediction and early detection of adverse events, as well as prompt reporting of adverse effects of IO drugs.