Singapore Researchers Develop Faster Method to Identify Treatments for Diabetic Wounds

Researchers at the National University of Singapore have developed a new drug discovery approach for diabetic wound healing that combines artificial intelligence with molecular modeling. Among the most promising candidates identified by the system was folic acid, a widely used vitamin supplement.

The study was conducted by an interdisciplinary team led by Professor Georgia Pastorin, bringing together experts in biomedical engineering and computer science. The findings were published in the journal ACS Nano Medicine.

Diabetic wounds, including diabetic foot ulcers, remain one of the most challenging complications of diabetes because they involve multiple disrupted biological processes, including inflammation, tissue repair, and cell growth. This complexity has made the search for effective treatments particularly difficult.

In the new study, researchers employed a multi-step strategy. Artificial intelligence was used to analyze scientific literature and identify potential biological connections, while computational chemistry techniques modeled interactions between drug molecules and proteins involved in wound healing. The most promising findings were then validated through laboratory experiments.

The platform analyzed approximately 2,989 existing drugs and 8,739 proteins associated with wound repair. Through a series of computational screening steps, the researchers narrowed the field to 35 promising drug candidates and 50 key protein targets.

One of the standout candidates was folic acid. Although commonly used as a nutritional supplement, its potential role in treating diabetic ulcers has received relatively little attention. In laboratory experiments using skin cells, folic acid significantly accelerated wound closure compared with untreated controls.

According to the researchers, the strength of the new method lies in its division of tasks: artificial intelligence identifies promising relationships from scientific publications, computational chemistry evaluates molecular interactions, and laboratory testing confirms whether the predicted biological effects occur in practice.

The team estimates that this integrated approach can reduce the time required for early-stage drug discovery by more than 70% compared with conventional methods. They believe the platform could also be adapted to accelerate the search for treatments for other complex diseases beyond diabetic wound healing.