Semi-supervised segmentation of RNA 3D structures using density-based clustering
Computational and Structural Biotechnology Journal · 01.01.2025 · 27 · 3966-3984
DOI: 10.1016/j.csbj.2025.08.037 ↗Kurz gefasst
Fachartikel in der Zeitschrift Computational and Structural Biotechnology Journal, online erschienen am 13.09.2025. Verfasst von Quoc Khang Le und 3 weiteren Autor:innen. Beteiligte Einrichtung: Université Evry Paris-Saclay, IBISC, Evry-Courcouronnes 91020, France. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 59 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
A growing body of evidence shows that the biological activity of RNA molecules is not only due to their primary and secondary structures, but also to their spatial conformation. This is analogous to proteins, where investigating function, folding, or evolution often requires dividing the three-dimensional (3D) structure into subparts that can be studied individually. These independent substructures, known as protein "3D domains", are geometrically defined as compact and spatially separate regions of the polypeptide chain. In RNA macromolecules, however, and to the best of our knowledge, no equivalent 3D-based concept has yet been formulated. We present RNA3DClust, an application of the Mean Shift clustering algorithm to the RNA 3D structure partitioning problem. For this work, a dedicated post-clustering procedure was developed to address the peculiarities of delimiting 3D domains in RNA conformations. Tuning and benchmarking RNA3DClust required us to create reference datasets of RNA 3D domain annotations and to devise a new scoring function-the Chain Segment Distance (CSD)-for assessing segmentation quality. Importantly, we show that the domain decompositions produced by RNA3DClust are consistent with those based on RNA biological function and evolution. Finally, the emerging interest in long non-coding RNAs (lncRNAs) and their likeliness of containing folded regions has motivated us to generate an additional reference dataset of lncRNA predicted conformations. The resulting delineations of 3D domains by RNA3DClust illustrate the potential of our method for analyzing lncRNA 3D structures. Source code and datasets are freely available for download on the EvryRNA platform at: https://evryrna.ibisc.univ-evry.fr.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Université Evry Paris-Saclay, IBISC, Evry-Courcouronnes 91020, France.
Zitieren
APA
Le, Q. K., Angel, E., Tahi, F., & Postic, G. (2025). Semi-supervised segmentation of RNA 3D structures using density-based clustering. Computational and Structural Biotechnology Journal, 27, 3966-3984. https://doi.org/10.1016/j.csbj.2025.08.037
Vancouver
Le QK, Angel E, Tahi F, Postic G. Semi-supervised segmentation of RNA 3D structures using density-based clustering. Computational and Structural Biotechnology Journal. 2025;27:3966-3984. doi:10.1016/j.csbj.2025.08.037
BibTeX
@article{Le2025,
author = {Le, Quoc Khang and Angel, Eric and Tahi, Fariza and Postic, Guillaume},
title = {Semi-supervised segmentation of RNA 3D structures using density-based clustering},
journal = {Computational and Structural Biotechnology Journal},
year = {2025},
volume = {27},
pages = {3966--3984},
doi = {10.1016/j.csbj.2025.08.037}
}
EUVIMED