Méthodes de deep learning pour la prédiction de structure secondaire des ARNs longs
Résumé
Le rôle essentiel des ARNs a été démontré dans divers processus biologiques et maladies. Toutefois, on ignore encore la fonction de nombreux ARNs. Une meilleure connaissance de leur rôle pourrait permettre de découvrir de nouveaux biomarqueurs ou cibles thérapeutiques et ainsi d’améliorer l’efficacité des traitements médicaux. Cependant, la val
idation expérimentale de leur fonction est très coûteuse, ce qui pose un frein à l’étude de
leurs rôles. Il est possible de pallier ce problème grâce à des outils informatiques. En particulier, l’apprentissage profond est aujourd’hui fréquemment utilisé pour l’étude des ARNs. Il permet de découvrir efficacement des motifs récurrents dans de larges jeux de données.
Contenu complet
1.1 Context
Precision medicine aims to tailor treatments to an individual’s unique molecular and clin ical profile [Jameson, 2015]. A growing aspect of this personalized approach involves exploring the role of RNA molecules in health and disease [Esteller, 2011]. While much attention has historically been directed toward protein-coding sequences, it is now appar ent that non-coding RNAs (ncRNAs) are integral to a wide range of biological processes, playing key regulatory roles in cancer and neurological, cardiovascular and developmental diseases [Calin, 2006; Prasanth, 2007; Lewin, 2009; Cech, 2014; Slack, 2019]. Further more, alterations in RNA secondary structure can disrupt normal cellular processes. The discovery of disease-causing mutations in RNAs is uncovering a wealth of new therapeu tic targets, while advances in RNA biology and chemistry are enabling the development of novel RNA-based tools for developing therapeutics [Cooper, 2009; Aznaourova, 2020]. The secondary structure of RNA has emerged as an important information for understand ing RNA function and exploiting it for therapeutic benefit [Wan, 2014; Assmann, 2023; Bose, 2024]. RNA secondary structure is made of complementary bases within the same RNA molecule that pair up, forming stems, loops, bulges, and more complex configura tions. These secondary structure patterns often serve as binding interfaces for proteins, small molecules, or other RNAs. For example, a hairpin loop might enable an ncRNA to recruit a protein complex that modifies gene expression. As such, identifying the RNA secondary structure allows researchers to understand how an RNA might act within the cell. In the context of precision medicine, discovering robust biomarkers is crucial for early diagnosis and patient stratification. Certain ncRNAs exhibit secondary structures that ei ther stabilize or interfere with diseases, and RNA molecules may show altered expression