Belarusian Scientists Leverage AI for Drug Repurposing, Target Identification, and De Novo Molecule Design
核心洞察
Belarusian researchers at the United Institute of Informatics Problems are applying AI to drug repurposing, screening existing medications for new therapeutic uses including COVID-19 (搜索).
Computational methods are being used to identify novel biological targets in the body, addressing limitations of current drugs that lose effectiveness over time.
Generative AI models are designing potential new drug molecules, with the most promising candidates against tuberculosis (搜索) and viral infections now entering testing.
Researchers at the United Institute of Informatics Problems of the National Academy of Sciences of Belarus (搜索) (NASB) are deploying artificial intelligence across three key domains of drug discovery: drug repurposing, biological target identification, and generative design of novel therapeutic molecules. Anna Karpenko, Research Fellow at the Mathematical Cybernetics Laboratory and Chairwoman of the Council of Young Scientists at the institute, detailed these efforts at a press conference in Minsk on 15 July.
Drug Repurposing Against COVID-19 (搜索)
The Mathematical Cybernetics Laboratory focuses on determining whether existing drugs already approved for specific diseases can be effectively used against other illnesses through computational methods, including AI. In a notable collaboration, the institute completed a successful development project with a university in Shanghai, screening large databases to detect potential drugs against COVID-19 (搜索). "We had a successful development project implemented with one of the universities in Shanghai. As part of the project we screened large databases to detect potential drugs against COVID-19," Karpenko stated.
Identifying New Biological Targets
A second research priority involves identifying the necessary targets in the body that drugs need to act upon. Karpenko noted that the human body is not yet fully researched, and drugs currently on the market do not always act effectively and safely. Furthermore, not all medications remain as effective as they were decades ago. "In recent years we've seen noticeable progress in this field thanks to computer methods," she explained, emphasizing that searching for new targets and determining their applications remain topical tasks.
Generative AI for Molecular Design
The third area of application involves generative models to design potential drug molecules. AI is used to predict new structures that can effectively act on targets in the body. "The task is to find substances that can effectively act on targets in the body, and determine whether these substances can become new medications," Karpenko emphasized.
Supercomputing Infrastructure and Therapeutic Focus
The research department's Republican Collective Use Supercomputer Center provides the computational capacity required to build more complex and deeper models and achieve effective results. The institute's R&D products span several therapeutic areas: COVID-19 (搜索), tuberculosis (搜索), and oncological diseases. According to Karpenko, the best results have been achieved with regard to tuberculosis and viral infections, where new molecules have been identified and are currently undergoing testing.