Drug discovery and development is a time-consuming process that is anything but mundane. In some cases, a majority of drug components may be rejected due to toxicity issues. Additionally, drug repositioning - the process of identifying new targets for existing or abandoned drugs - is a crucial aspect of drug discovery. By enabling researchers to minimize the number of wet-lab analyses, computational prediction of the binding affinity between chemical compounds and protein targets significantly enhances the chances of identifying lead compounds. In recent years, machine learning (ML) and deep learning approaches have been utilized to predict drug-target interactions, thus reducing the time and cost involved in drug discovery endeavors. Proteins that are targeted by drugs are typically classified into four main groups: enzymes, ion channels, G-protein-coupled receptors, and nuclear receptors. Drug repurposing principles can be broadly categorized as either drug-based or disease-based. In drug-based repurposing, a hypothesis is analyzed to determine whether a drug can effectively treat multiple diseases based on the similarity between them. Conversely, disease-based repurposing involves identifying new uses for existing drugs based on their known targets. Computational methods, such as machine learning models, are often utilized to predict possible drug-target interactions. This study aims to explore various drug repurposing methods and their applications using machine learning models in drug discovery and development, given the abundance of biological data and computational resources available to researchers.

Bioinformatics, Artificial Intelligence, Drug Discovery, Drug Development, Drug Repurposing