Researchers use artificial intelligence to design potential antiviral compounds
Viral infections can be difficult to treat because viruses evolve rapidly, allowing them to develop resistance to drugs designed to attack viral proteins directly.
Researchers at Korea University have employed artificial intelligence (AI) to discover potential antiviral compounds that target human deoxyhypusine synthase (DHPS), an enzyme crucial for the replication of various viruses. This approach, called host-directed therapy (HDT), differs from traditional methods by focusing on the host's biological pathways rather than directly attacking the virus.
The study, published in the Journal of Cheminformatics on August 5, 2026, outlines a systematic process using machine learning models to predict molecular binding scores and generate new molecules. Among the tested models, Out-of-fold stacking (OOF-Stacking) and RL-Design demonstrated the strongest performance in predicting molecular binding and generating effective compounds, respectively.
Upon experimental validation, six AI-generated fragments, ranging from 140 to 190 daltons, were found to inhibit DHPS without affecting inactive compounds. Two of these fragments achieved notable IC₅₀ values, indicating their potential as early-stage antiviral inhibitors. However, further research is necessary to assess their full potential, including their selectivity, toxicity, and efficacy in animal models.
The researchers emphasize that this work offers a reproducible pathway for transforming AI-designed molecules into experimental validation, potentially making AI-assisted drug discovery more accessible to laboratories with limited computational resources.
Written by urgent.news from Phys.org's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
