[Courtesy of the Korea Pharmaceutical and Bio-Pharma Manufacturers Association]
As pharmaceutical companies are hesitant to introduce AI, the government should set effective utilization policies and industrial directions, said Kim Hwa-jong, a professor of computer and communications engineering who heads the Korea AI Center for Drug Discovery and Development (KAICD).
"The government's role is the most important. First of all, it is a representative area where private companies are not investing, and it is necessary to provide a method for actual data sharing," Kim said, calling for an ecosystem where AI venture companies can provide solutions and make money. "The introduction of AI is essential in terms of a big direction of technological change."
Kim suggested that AI technology can be used extensively in the development of new drugs from the discovery of candidate substances to clinical trials. "New AI drug developers are focusing on introducing AI at the stage of discovering new drugs, such as discovering new targets that cause diseases and designing new compounds virtually."
In the preclinical and clinical stages, Kim said that AI can be introduced for experimental design, toxicity prediction, test automation, test optimization, factory manufacturing, and pharmacovigilance. "AI development is only in the beginning stage, and the scope and availability of AI technology in the future are expected to gradually accelerate through the interaction of analysis and data exchanges."
In particular, AI is more likely to produce results in the development of new drugs where it is difficult for humans to understand intermediate processes, Kim said, adding AI models will better predict biological, chemical, and medical phenomena that humans rarely understand. "AI algorithms used in the development of new drugs can accelerate the development stage of new drugs as they can explore nearly infinite substances and increase the probability of clinical success while spending less cost and time."
The biggest obstacle, Kim said, is a limit to the performance of AI because data collected from various medical institutions cannot be exchanged. "A method of effectively sharing data held by medical institutions, public institutions, and pharmaceutical companies, respectively, is needed, and our center is trying to establish measures to solve such problems."
KAICD aims to help domestic companies incorporate AI into the process of developing new drugs, find more effective new drug candidates, and reduce the time and cost of developing new drugs. The center also helps pharmaceutical and bio companies find AI solutions. "Because AI models need to be applied to actual data to know their performance, it is difficult to determine the level of AI solution companies in advance. So KAICD serves as networking to connect companies that need AI solutions with solution providers."
(This story is based on an interview conducted by Aju Business Daily reporter Jeon Hwan-wook.)
Copyright ⓒ Aju Press All rights reserved.
