This is an early access version, the complete PDF, HTML, and XML versions will be available soon.
A Graph-Based Deep Learning Framework with Gating and Omics-Linked Attention for Multi-Omics Integration and Biomarker Discovery
College of Medical Information Engineering, Guangdong Pharmaceutical University, Guangzhou 510006, China
*
Author to whom correspondence should be addressed.
†
These authors contributed equally to this work.
Biology 2025, 14(12), 1764; https://doi.org/10.3390/biology14121764
Submission received: 28 October 2025
/
Revised: 26 November 2025
/
Accepted: 5 December 2025
/
Published: 10 December 2025
(This article belongs to the Special Issue Advancing Translational Science Using Bioinformatics and Big Data-Driven Approaches)
Simple Summary
Understanding complex diseases like cancer and Alzheimer’s requires analyzing multiple types of biological data, including gene expression, DNA methylation, and microRNA. Integrating these multi-omics datasets provide a more comprehensive understanding of disease mechanisms. In this study, we present MOGOLA(Multi-Omics integration by Gating and Omics-Linked Attention), a novel deep learning framework that effectively integrates multi-omics data through graph-based learning and attention mechanisms. Our approach captures critical patterns within each omics data and uncovers meaningful relationships across them. Evaluations on several real-world disease datasets, including cancer and Alzheimer’s disease, demonstrate that MOGOLA outperforms existing methods in classification accuracy and successfully identifies key biomarkers with strong biological relevance.
Abstract
Integration of multi-omics data provides a comprehensive perspective on complex biological systems, facilitating advances in disease classification and biomarker discovery. However, the heterogeneity and high dimensionality of omics data present significant analytical challenges. To achieve effective and interpretable multi-omics integration, we propose a novel deep learning framework named MOGOLA(Multi-Omics integration by Gating and Omics-Linked Attention). MOGOLA consists of three core components: (1) A hybrid graph learning module that integrates Graph Convolutional Networks and Graph Attention Networks for intra-omics feature extraction. (2) A gating and confidence mechanism that adaptively weighs feature importance across different omics types. (3) A cross-omics attention-based fusion module that captures inter-omics relationships. Comprehensive evaluations on four benchmark datasets (BRCA, KIPAN, ROSMAP, and LGG) demonstrate that MOGOLA consistently outperforms eleven state-of-the-art approaches. Ablation studies further validate the contribution of each module, while biomarkers identification highlight the framework’s clinical potential. These results show that MOGOLA is a robust and interpretable approach for multi-omics data integration and a contribution to advances in computational biology and precision medicine.
