Accelerating Drug Discovery through AI-Based Repurposing and Design

Abstract

The conventional drug discovery pipeline is constrained by high development costs, extended timelines, and high attrition rates, highlighting the need for more efficient and adaptive discovery strategies. Artificial intelligence (AI) has emerged as a transformative approach to address these challenges by enabling data-driven drug repurposing and molecular optimization. The study presents an integrated AI-driven framework that combines AI-based drug repurposing with generative molecular design to accelerate early-stage drug discovery. The proposed framework employs machine learning and deep learning techniques to identify novel disease–drug associations from large-scale biomedical data, including drug- target interaction networks, clinical evidence, and biological knowledge graphs. These repurposing strategies facilitate the prioritization of approved or investigational compounds for new therapeutic indications, thereby reducing developmental risk and timelines. Subsequently, generative AI models, such as variational autoencoders, transformer-based architectures, and reinforcement learning methods are utilised to optimize the chemical structures of repurposed leads. These models efficiently explore chemical space while simultaneously improving target affinity, drug-likeness, ADMET properties, and synthetic feasibility through multi-objective optimization. By integrating repurposing and generative molecular design into a unified workflow, the framework bridges knowledge-driven and structure-based drug discovery paradigms. This synergistic approach enhances lead prioritization and addresses limitations of standalone repurposing or de novo design strategies. The proposed pipeline is particularly relevant for complex and unmet disease areas, including cancer and infectious diseases, where rapid therapeutic development is critical. Overall, this work highlights the potential of integrated AI methodologies to streamline drug discovery, reduce late-stage failures, supporting the development of effective and affordable therapeutics.