AI-Driven Drug Discovery: Integrating Machine Learning and Multi-Omics for Pharmacological Target Identification and Drug Repurposing

Artificial intelligence (AI) combined with multi-omics data is creating new ways to identify therapeutic targets and find new uses for existing medicines. Traditional drug discovery takes many years, costs a great deal, and often fails before a medicine reaches patients. AI methods can examine large genomic, transcriptomic, proteomic, and metabolomic datasets and help rank disease-related molecules that may be suitable drug–targets. This review describes common machine-learning (ML) and deep-learning (DL) methods, explains the value of each omics layer, and discusses approaches for combining these data to identify targets. It also reviews drug-repurposing methods such as transcriptomic signature reversal, network analysis, and knowledge-graph link prediction, with examples from cancer, infectious disease, and cardiometabolic conditions. Key issues include data quality, model interpretation, regulatory expectations, and the need for experimental confirmation. Future work may benefit from models that combine biological mechanisms with ML, repeated measurements over time, and precision-medicine approaches. These tools can guide research, but predictions must be tested in laboratory studies and clinical trials.