About Drug Repurposing

Pharmacologist James Black once said, "The most fruitful basis for the discovery of a new drug is to start with an old drug." Drug repurposing is a method of developing new therapeutic uses for existing drugs whose safety and pharmacokinetics have been proven in humans. This can significantly reduce the time, cost, uncertainty, and side effects associated with drug development.

Recent studies have increasingly used artificial intelligence (AI) computational methods to systematically predict new drug targets or drug repurposing candidates. This is because computer-simulated approaches are faster and less costly than experimental high-throughput screening and can be used as an initial filtering step for evaluating thousands of compounds.

Here, we cite literature on the repurposing of small-molecule drugs. The authors of the article used artificial intelligence/machine learning (AI/ML) techniques to predict off-target interactions of drugs, as well as to identify and optimize existing small molecule drugs for new therapeutic applications through cross-species transcriptomics information. This approach combines chemical similarity-based and target-based prediction methods with machine learning approaches to predict potential off-target interactions.

Fig. 1. The workflow of computational drug repurposing. (Rao M.; et al. 2023)

Research Methods

The research team developed a computational framework that combines AI/ML and chemical similarity-based target prediction methods with transcriptomic information across species. The framework uses eight different target prediction methods, including three machine learning methods, to analyze a dataset of 2,766 FDA-approved drugs.

Research Results

• The study identified 27,371 off-target interactions involving 2013 protein targets, with an average of approximately 10 interactions per drug. Sixty-three percent of these interactions have been confirmed in vitro.
• The researchers explored the chemical space of drugs, analyzing the physicochemical properties of 2,766 FDA-approved drugs and finding that fewer than half of the drugs followed the traditional "drug-like chemical space" paradigm.
• Compared to the drugs in the dataset, the researchers identified 150,620 structurally similar compounds that may have potential therapeutic or toxicity targets.
• The researchers found that drugs targeting G protein-coupled receptors (GPCRs), enzymes and kinases predicted the highest number of interactions.

This literature provides a scientific basis for discovering additional therapeutic potentials for existing drugs through AI/ML technology and transcriptomics data. By predicting off-target interactions, small molecule drug development success rates can be increased and new perspectives and approaches are provided in the field of repurposing existing drugs.

Unlike traditional massive experiments and accidental discoveries, AI technology is more rapid and efficient in the development of new uses for existing drugs, and has great potential in the treatment of rare diseases and unmet medical needs. Recognizing the unique advantages of AI technology in the field of drug repurposing, Protheragen-ING AI-Pharma is committed to using advanced machine learning algorithms to find new therapeutic opportunities for existing drugs.

Our AI model can quickly analyze large amounts of biomedical data to predict and identify drug performance in different disease contexts. If you are interested in our services or have a question, please feel free to contact us for more details.

Original Article:

Rao M.; et al. (2023). Artificial Intelligence/Machine Learning-Driven Small Molecule Repurposing via Off-Target Prediction and Transcriptomics. Toxics. 2023, 11(10): 875.

Services Related in the Article:

Drug Repositioning and Repurposing
AI-powered Drug Discovery and Design
AI-assisted Drug High Throughput Screening