The competition for AI-based drug development is shifting from molecular design to securing patient and clinical data. However, the lack of a robust medical data collection and utilization system in South Korea means that a national framework is essential for AI-driven drugs to translate into actual therapeutic outcomes.
Seok Cha-ok, a professor at Seoul National University, emphasized this need during a press conference at the '2026 Korean Society for Biotechnology Fall Conference and International Symposium' held at the Jeju International Convention Center on October 2. He stated, "The system-level and individual-level medical data necessary to predict and verify actual therapeutic effects are absolutely lacking in the country. It is crucial to establish a national system for collecting and utilizing medical data."
Professor Seok explained that AI drug development can be divided into molecular design and the prediction and verification of therapeutic effects. At the molecular level, he noted that data is no longer a decisive variable.
He said, "In molecular-level modeling, the type and scale of data do not differ significantly between models. AlphaFold also utilized publicly available data similar to previous competing groups, and now AI-generated distilled data plays a crucial role." He added, "At this stage, molecular design is more about competition in ideas and computational resources like GPUs rather than data competition."
The challenge arises in the next phase. To predict whether the designed substances will be effective in actual patients, data at the cellular, tissue, and human levels is required. Professor Seok stressed, "This area has not yet reached a stage where AI models can lead the industry as a whole. For further advancement, a national data system must be in place."
He believes that data integration is essential to tackle diseases that cannot be treated with existing methods. "Galax is currently applying molecular design technology to biologically validated diseases and targets. To create new treatments that were previously impossible, it is insufficient to rely solely on molecular-level data; AI integration that combines biological and medical data is necessary," he stated.
The potential of AI-driven drugs is already becoming a reality. Professor Seok noted, "Antibodies designed from scratch using AI are already entering the drug development phase, and among the substances currently in development, many candidates could become new drugs in a few years."
The changes brought about by AI are occurring not only in drug development but across nearly all fields of science and technology. Therefore, he advocates for support to be spread across the entire scientific and technological landscape rather than focusing on specific areas. Regarding talent development, he emphasized the need to design new interdisciplinary curricula that select core knowledge that must be taught in each discipline, highlighting the importance of 'dual-skilled talent' who understand both AI and specialized fields.
Gregory Stephanopoulos, a professor at the Massachusetts Institute of Technology (MIT) who pioneered the field of metabolic engineering, delivered a lecture on the future direction of biotechnology at the conference. In a subsequent press conference, he predicted that the combination of AI and automation could lead to significant advancements in biomanufacturing.
When asked if an AI-based leap similar to AlphaFold is possible in metabolic engineering, he replied, "We still need to wait and see," but added, "The combination of AI and automation could lead to substantial progress."
He explained, "The design, construction, and testing cycle, where robots test various candidates and continuously retrain machine learning models with that data, will be largely automated. This allows for exploration from known areas into data-scarce regions." He noted that such changes could manifest in strain performance improvement, metabolic engineering, and drug candidate design.
However, he was cautious about how AI will transform the industry. He stated, "While the potential for AI to bring significant changes to the industry is high, it is still uncertain which fields will experience these changes first. Drug development, where there is a high unmet medical need, is a promising area."
He also emphasized the role of field experts. Professor Stephanopoulos remarked, "There are no universal solutions for AI applications, such as in the hydrogenation stage of biorefinery processes or the depolymerization stage of polymers. Each field expert must find solutions tailored to their processes. There is a clear gap between computer scientists and bio-experts, and efforts are needed to understand each other's fundamental principles."
This conference, themed 'AI STAR (AI Integrated Sustainable Transformations for Advanced Robust Biotechnology),' invites top experts from home and abroad to discuss key issues in the biotechnology (BT) field. The international sessions cover topics such as convergent bioengineering, sustainable aviation fuel (SAF) production technologies based on biofoundries, and technologies and applications for controlling cellular microenvironments.
The domestic sessions include sessions led by sector committees, as well as discussions on digital transformation-based cell culture engineering, greenhouse gas reduction biotechnology, cell and gene engineering for next-generation cancer treatment, protein nanocage engineering, life cycle assessment (LCA) of sustainable bioresources, and AI-based biomanufacturing symposiums.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.