MACHINE LEARNING IN DRUG DISCOVERY AND PHARMACOLOGICAL RESEARCH: APPLICATIONS, ACHIEVEMENTS, AND CHALLENGES

Abstract

This article presents a systematic literature review examining machine learning (ML) applications across the drug discovery and development pipeline, from target identification and molecular design through preclinical testing, clinical trial optimisation, and pharmacovigilance. Publications from 2015 to 2024 indexed in PubMed, Scopus, Web of Science, and Nature journals were analysed. The review addresses the theoretical foundations of ML in computational pharmacology, principal application domains, documented achievements including AI-designed molecules entering clinical trials, challenges including data quality limitations and model interpretability, and the implications for pharmaceutical research capacity development. The analysis demonstrates that ML applications in drug discovery have produced measurable reductions in development timelines — particularly in virtual screening (reducing compound screening time by 90–99%) and toxicity prediction (improving prediction accuracy to 70–85%) — while enabling exploration of chemical space at scales impossible through traditional experimental approaches. The first AI-discovered drug candidates entering clinical trials represent landmark achievements with transformative implications for pharmaceutical research.