MicroRNAs in oncology: a translational perspective in the era of AI
Cet article examine le rôle des microARNs dans la biologie des cancers ainsi que leur intérêt comme outils diagnostiques ou cibles thérapeutiques puis analyse les opportunités offertes par l'intelligence artificielle et les algorithmes d'apprentissage automatique pour découvrir des signatures ou des stratégies thérapeutiques basées sur les microARNs
Over the past three decades, knowledge of microRNA (miRNA) biology has advanced from the initial discovery of their regulatory functions to the finding of abnormal activity in leukaemias, and then to a comprehensive understanding of the roles of miRNAs in both normal physiology and most diseases, with cancer being extensively studied. miRNA dysregulation contributes to tumorigenesis, with certain miRNAs acting as either tumour suppressors or oncogenic factors in a context-dependent manner. A subset of miRNAs have shown promise as tumour biomarkers and therapeutic targets in preclinical studies, with several miRNA-based diagnostic tools and treatments progressing to clinical trials. Artificial intelligence (AI) and machine learning techniques began to be introduced into cancer research and oncology a decade ago and are now on the verge of revolutionizing biomarker identification and clinical trials. In this Review, we highlight important roles of miRNAs in cancer biology and their potential as diagnostic tools and therapeutic targets. In particular, we discuss emerging challenges and opportunities presented by AI-driven data analysis and combinatorial strategies, and how advances in these areas have addressed previous doubts on the clinical translation of miRNA-based biomarkers and therapeutics.
Nature Reviews Clinical Oncology , article en libre accès, 2026