<(From top left) Professor Woo Youn Kim (KAIST), Dr. Jeheon Woo (KISTI), Dr. Seonghwan Kim (KAIST), and Jun Hyeong Kim (PhD candidate)>
Whether a smartphone battery lasts longer or a new drug can be developed to treat incurable diseases depends on how stably the atoms constituting the material are bonded. The core of 'molecular design' lies in finding how to arrange these countless atoms to form the most stable molecule. Until now, this process has been as difficult as finding the lowest valley in a massive mountain range, requiring immense time and costs. Researchers at KAIST have developed a new technology that uses artificial intelligence to solve this process quickly and accurately.
KAIST announced on February 10th that Professor Woo Youn Kim's research team in the Department of Chemistry has developed 'Riemannian DenoisingModel (R-DM),' an artificial intelligence model that understands the physical laws governing molecular stability to predict structures.
The most significant feature of this model is that it directly considers the 'energy' of the molecule. While existing AI models simply mimicked the shape of molecules, R-DM refines the structure by considering the forces acting within the molecule. The research team represented the molecular structure as a map where higher energy is depicted as hills and lower energy as valleys, designing the AI to move toward and find the valleys with the lowest energy.
R-DM completes the molecule by navigating this energy landscape, avoiding unstable structures to find the most stable state. This applies the mathematical theory of 'Riemannian geometry,' resulting in the AI learning the fundamental law of chemistry: 'matter prefers the state with the lowest energy.'
Experimental results showed that R-DM achieved up to 20 times higher accuracy than existing AI models, reducing prediction errors to a level nearly indistinguishable from precise quantum mechanical calculations. This represents the world's highest level of performance among AI-based molecular structure prediction technologies.
This technology can be utilized in various fields, including new drug development, next-generation battery materials, and high-performance catalyst design. It is expected to serve as an 'AI simulator' that will dramatically speed up research and development by significantly shortening the molecular design process, which previously took a long time. Furthermore, it has great potential in environmental and safety fields, as it can quickly predict chemical reaction paths in situations where experiments are difficult, such as chemical accidents or the spread of hazardous substances.
Professor Woo Youn Kim stated, "This is the first case where artificial intelligence has understood the basic principles of chemistry and judged molecular stability on its own. It is a technology that can fundamentally change the way new materials are developed."
This study was led by Dr. Jeheon Woo from the KISTI Supercomputing Center and Dr. Seonghwan Kim from the KAIST Innovative Drug Discovery Research Group as co-first authors. The research results were published on January 2nd in the world-renowned academic journal Nature Computational Science.
※ Paper Title: Riemannian Denoising Model for Molecular Structure Optimization with Chemical Accuracy, DOI: 10.1038/s43588-025-00919-1
Meanwhile, this research was conducted with the support of the Chemical Accident Prediction-Prevention Advanced Technology Development Project of the Korea Environmental Industry & Technology Institute, the Science and Technology Institute InnoCore Project of the Ministry of Science and ICT, and the Data Science Convergence Talent Cultivation Project conducted by the National Research Foundation of Korea with support from the Ministry of Science and ICT.
