A university and three corporations are forming a trio to carry out a large government project worth 47 billion won to develop antibody new drugs with artificial intelligence (AI). They set a bold goal of developing 10 antibody drug candidates over two years.
Korean bio corporation PROTEINA said on the 5th that it was finally selected as the lead research and development institution for the state project "Development and validation of antibody biopharmaceuticals using AI models," overseen by the Ministry of Health and Welfare. Samsung Bioepis will take part in the project, along with a research team led by Professor Baek Min-kyung of the Department of Biological Sciences at Seoul National University. Baek co-developed "RoseTTAFold," a protein structure prediction AI, with David Baker of the University of Washington, who won the Nobel Prize in chemistry last year.
The core of this project is the development of a "biobetter" with improved efficacy over existing drugs. Put simply, it is to develop an improved new drug that expands the pool of patients eligible for dosing or reduces side effects by adding new technology to existing medicines. The three institutions aim to discover 10 antibody drug candidates over 27 months, move three into preclinical trials, and complete an investigational new drug (IND) application for one.
At first, the industry expected that an AI new drug development corporation would be selected as the lead institution for this project. But the result was different. Instead of an AI platform, PROTEINA, which has technology to analyze "protein-protein interaction" data, made the final list. The company developed technology that precisely analyzes how proteins bind to each other and act.
Yoon Tae-young, a professor in the Department of Biological Sciences at Seoul National University, founded PROTEINA in 2016 while at the Korea Advanced Institute of Science and Technology (KAIST). The core technology is the "SPID" platform. It is a technology that precisely observes, like taking pictures with a camera, the moment proteins bind and exchange signals, and quantifies and analyzes it. Yoon said its strength is the ability to detect protein binding with up to 100 times higher sensitivity than conventional methods.
SPID is seen not just as an analysis device but as a comprehensive solution covering the entire new drug development process. That is because it can be used broadly from candidate discovery to mechanism-of-action analysis, target protein binding confirmation, antibody structure design, biomarker discovery, and companion diagnostic device development.
PROTEINA also developed the AI antibody design platform "AbGPT-3D" with Professor Baek Min-kyung's team. When an antigen protein on the surface of a pathogen or diseased cell is identified, AbGPT-3D designs the structure of an antibody that binds to it, generates a sequence optimized for that structure, and evaluates developability, offering a three-step AI function platform. PROTEINA said this technology shortens verification that used to take months to within two weeks and can analyze the amino acid sequences that make up more than 5,000 antibody proteins per week.
An industry official who took part in the review of this state project said, "SPID is a technology that allows you to visually confirm the actual action of drug candidates," and noted, "Combined with Seoul National University's AI design capability, we judged it would help quickly and accurately predict the efficacy of antibody new drugs."
Samsung Bioepis also plays an important role in this project. Its task is to complete candidates as clinical trial antibody drugs and oversee the entire preparation process for clinical trials. It will establish cell lines for antibody drug production with automated equipment, optimize culture and purification processes, and handle preclinical testing and the IND application.
The idea is to make full use of Samsung Bioepis' experience in developing antibody drugs. Samsung Bioepis will lead clinical trials and commercialization of antibody candidates that succeed through this project, and PROTEINA will receive royalties accordingly.
Kim Yun-cheol, an executive director at Samsung Bioepis, said, "Being selected for this state project recognizes our process optimization capabilities built through biosimilar development," and added, "We will use AI design and analysis platforms to discover new candidates and accelerate innovative new drug development."
Yoon Tae-young, CEO of PROTEINA, said, "Based on a joint research system, we will achieve challenging goals that were impossible with existing technology within 27 months," and added, "With this project, we will fully introduce AI design technology across the entire antibody drug development process and create a proof case for global AI-based antibody new drug development."