Aug 2026· Frontiers in Drug Discovery· Vol 6· 2 citations· 124 references
TL;DR
This comprehensive review examines the state-of-the-art applications of LLMs across the drug development pipeline, spanning target identification, molecular generation, property prediction, and drug repurposing in drug discovery, as well as clinical decision support, patient stratification, treatment personalization, and genomic interpretation in precision medicine.
Abstract
The convergence of artificial intelligence and biomedical research has catalyzed emerging impact in drug discovery and precision medicine. Large language models (LLMs), originally developed for natural language processing, have emerged as powerful tools capable of processing complex biomedical data, from molecular structures to clinical records. This comprehensive review examines the state-of-the-art applications of LLMs across the drug development pipeline, spanning target identification, molecular generation, property prediction, and drug repurposing in drug discovery, as well as clinical decision support, patient stratification, treatment personalization, and genomic interpretation in precision medicine. We analyze the technical methodologies underlying these applications, including multi-modal architectures, knowledge-guided approaches, and retrieval augmented generation systems. Through examination of recent advances, we highlight key achievements such as end-to-end drug discovery pipelines, multi-agent systems for clinical simulation, and knowledge-enhanced models achieving state-of-the-art performance. We critically assess current challenges, including data privacy, model interpretability, hallucination risks, and ethical considerations. Finally, we discuss future directions, emphasizing the potential for federated learning, explainable artificial intelligence, and integrated multiscale approaches to advance the field toward more reliable, transparent, and clinically applicable systems.
Traditional drug development suffers from high costs, low success rates, and patient response variability. Precision drug discovery seeks to overcome these limitations by targeting specific genetic and molecular mechanisms but faces challenges in integrating cross-scale, multimodal biomedical data. Recent advances in artificial intelligence (AI), especially deep learning, provide powerful tools to navigate these complexities. This review surveys representative AI methods across four core stages of precision drug discovery: (a) target identification and validation using omics-, drug-, and structure-based approaches; (b) structure- and sequence-guided virtual screening of active compounds; (c) individualized drug response prediction integrating cell-line, single-cell, and multimodal data; and (d) AI-enabled toxicology and safety modeling to anticipate adverse liabilities and improve translational success. We further highlight a paradigm shift toward autonomous scientific agents capable of causal reasoning and end-to-end experimental guidance. Finally, we discuss persistent challenges, including data bias, limited interpretability, and in silico-to-wet lab translation.
The traditional drug discovery and development process is historically characterized by high attrition rates, escalating financial costs, and decade-long timelines. The emergence of artificial intelligence (AI) and machine learning (ML) has transformed this paradigm by enabling efficient navigation through vast chemical spaces and the integration of complex multi-omic datasets. This evidence-based literature review critically examines the evolution and application of computational technologies across the pharmaceutical pipeline, ranging from early expert systems like DENDRAL, computer-aided drug design (CADD), and quantitative structure-activity relationship (QSAR) modeling to AlphaFold 3 biomolecular complex predictions, computer-assisted synthesis planning (CASP), natural product bioprospecting, and autonomous multi-agent systems. Key advancements in antimicrobial screening, precision oncology, phytochemical characterization, and clinical-stage AI-generated molecules are highlighted. Finally, the translational gap is addressed, emphasizing that AI functions as an advanced decision-support framework requiring rigorous in vitro and in vivo experimental validation, wherein qualified human mediation remains indispensable for therapeutic success.
Leonardo Mairene Muniz· Brazilian Journal of Health...· 0 citations
An operational, end-to-end workflow that explicitly connects computational predictions to medicinal chemistry decision points is provided, addressing a critical gap between computational prediction and clinical translation.
Antonio Lavecchia· Medicinal research reviews (...· 0 citations
Oncologic drug development is lengthy (~14 years) and expensive (~1.2 billion USD) with low clinical trial success rates (4.1%). Big data and artificial intelligence (AI) are widely proposed as tools to address these challenges. In this review, we examine the current performance and future potential of big data and AI applied to preclinical discovery and development, clinical trials, and the regulatory approval process. We first examine the data foundation required for effective AI, including data harmonization, data commons, and analytical tools. We then assess preclinical applications spanning target identification, compound-library curation, virtual ligand screening, generative chemical design, and high-throughput and high-content screening. In clinical development, we consider the use of big data and AI for outcome prediction, trial design, external and synthetic control arms, adaptive monitoring, and in silico trials. Finally, we discuss how post-approval electronic health records can generate real-world data and real-world evidence to support drug repurposing and improve future oncology drug discovery. Big data is conventionally characterized by a series of “Vs.” In this review, we have used seven “Vs” spanning descriptive and constraining properties of big data and a singular outcome. We have proposed an eighth, Vernacular, a constraint defined as the combined alignment of data semantics and terminology, data representation, data exchange, and data governance across heterogeneous, independently generated datasets to promote interoperability and combined analysis. Although cancer data exhibit substantial Volume, Velocity, and Variety, they remain distributed across fragmented repositories that often cannot be readily integrated. We conclude with a discussion of tabulated resources currently available for the application of big data and AI to oncologic therapeutics.
Fakhar U. Singhera, J. Overhulse, Terrence M. Lee et al.· Cancers· 0 citations
This review examines the available literature from human clinical studies, computational drug discovery research, systematic reviews, meta-analyses, and clinical investigations, highlighting the applications of AI in target identification, virtual screening, lead optimization, drug repurposing, ADMET prediction, and precision medicine.
Neha Arora, Yogesh Matta, Monu Kumar et al.· Journal of Pharmaceutical Re...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
