EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

GMFLDA: Improved Prediction of lncRNA-Disease Association via Graph Convolutional Network

Kwangsu Kim, Jihwan Ha

IEEE Access · 2025 · 13 · 85330-85341

DOI: 10.1109/access.2025.3568461 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Access, 2025. Verfasst von Kwangsu Kim und 1 weiteren Autor:innen. Die Arbeit wurde bisher 27-mal zitiert (Crossref) und nennt 69 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Research Foundation of Korea, Korean Government.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The rapid expansion of diverse networks has created a growing need to integrate multiple heterogeneous structures to effectively capture both inter- and intra-entity relationships. This integration helps preserve the intrinsic meaning of complex biological interactions. Network-based approaches have been highly effective in predicting node labels and uncovering hidden associations between entities. These methods have been widely applied in areas such as user-item recommendations, gene-gene interactions, and lncRNA-disease association prediction. In this study, we present GMFLDA, an advanced machine learning framework for inferring lncRNA-disease associations (LDA) by synergizing graph convolutional networks (GCNs) with deep matrix factorization. Specifically, GCNs are leveraged to extract and encode high-fidelity feature representations of lncRNAs and diseases, while deep matrix factorization, implemented via a multi-layer perceptron, facilitates the discovery of potential disease-associated lncRNAs. Our model exhibits outstanding predictive performance, achieving AUCs of 0.9183 and 0.9057 in leave-one-out and five-fold cross-validation experiments, respectively. Extensive comparative evaluations demonstrate that GMFLDA surpasses four state-of-the-art methods, underscoring its superior predictive capability. We anticipate that GMFLDA will serve as a powerful computational tool for biomarker discovery, significantly accelerating the identification of disease-associated lncRNAs while mitigating the time and cost constraints of traditional wet-lab experiments.

Förderung

  • National Research Foundation of Korea
  • Korean Government

Zitieren

APA

Kim, K., & Ha, J. (2025). GMFLDA: Improved Prediction of lncRNA-Disease Association via Graph Convolutional Network. IEEE Access, 13, 85330-85341. https://doi.org/10.1109/access.2025.3568461

Vancouver

Kim K, Ha J. GMFLDA: Improved Prediction of lncRNA-Disease Association via Graph Convolutional Network. IEEE Access. 2025;13:85330-85341. doi:10.1109/access.2025.3568461

BibTeX
@article{Kim2025,
  author = {Kim, Kwangsu and Ha, Jihwan},
  title = {GMFLDA: Improved Prediction of lncRNA-Disease Association via Graph Convolutional Network},
  journal = {IEEE Access},
  year = {2025},
  volume = {13},
  pages = {85330--85341},
  doi = {10.1109/access.2025.3568461}
}
Kostenlos starten →