SK Biopharmaceuticals Builds AI Model Specialized in Drug Development ... Expands Use in Research and Clinical Development

[Financial News] SK Biopharmaceuticals has completed the development and validation of an artificial intelligence (AI) model specialized in drug development and is expanding its use into real-world R&D. Its strategy is to use AI trained on vast amounts of biomedical data to improve research efficiency, from discovering drug candidates through clinical development.
SK Biopharmaceuticals said on the 6th that it had completed the development and demonstration of the Ministry of Science and ICT-led “AI-Specialized Foundation Model Project” in about 10 months.
The project aims to build a foundational AI model for medical science and advance it for use in three areas: drug development, omics and clinical research. Participants included SK Biopharmaceuticals, Lunit, Trillion Labs, Aizen Science, KakaoHealthcare Corp., and domestic industry-academic and medical institutions.
SK Biopharmaceuticals was responsible for building data and developing and validating AI models for drug development. Using biomedical data on small-molecule compounds, proteins and genomes, among other areas, it secured training data totaling about 7.14 million records and 4.4 billion tokens. The data covers 66 drug-development tasks, including drug properties and toxicity, interactions between drugs and targets, drug responses, and predictions of clinical trial outcomes.
The AI model co-developed with Trillion Labs underwent performance validation on these 66 tasks. The company said that applying not only public data but also nonpublic drug data held by pharmaceutical companies confirmed the model’s potential for use in research settings.
The company also tested the potential for AI to support clinical development. It built a workflow system to extract information on chemicals used in research and assist with expert reviews and the preparation of submission documents. It also developed a digital twin model using publicly available clinical data on Alzheimer’s disease. The model uses virtual patients to predict changes in cognitive function for individuals and updates its predictions when new clinical information is added.
SK Biopharmaceuticals plans to use the AI model development capabilities it gained through the project and computing infrastructure powered by NVIDIA B200 graphics processing units (GPUs) to drive the AI transformation of its own R&D. Through its TR-AX (AI Transformation of Translational Research) initiative, the company aims to expand AI use across research into diseases, therapeutic targets and drug candidates, as well as clinical development.
SK Biopharmaceuticals President Donghoon Lee emphasized the importance of AI models that can be used in real-world research settings, saying, “We will ensure that AI takes root as a key competitive advantage that supports researchers’ decision-making and increases the speed and likelihood of success in drug development.”
[email protected] Jung Sang-hee Reporter