Learning Multi-Relational Graph Representations for DNA Methylation-Based Biological Age Estimation

2026-05-07

AI Summary

This study addresses a key limitation in existing DNA methylation–based age prediction methods, which typically overlook the biological relationships among CpG sites. To overcome this, the authors propose a novel multi-relational graph neural network that explicitly models three types of biological associations—co-methylation, genomic co-localization, and gene-level connectivity—by constructing corresponding graph structures. These multi-source representations are adaptively integrated through a learnable gating mechanism. Evaluated on large-scale datasets, the method achieves significantly improved prediction accuracy, demonstrating stronger correlation with chronological age. Moreover, it exhibits heightened sensitivity in detecting epigenetic age acceleration across multiple disease cohorts and enhances model interpretability by leveraging biologically informed graph structures.

Abstract

Aging clocks aim to estimate biological age, a measure of physiological state distinct from chronological age, from observable biomarkers, and are widely used for health assessment and disease analysis. DNA methylation is a particularly informative biomarker due to its stability and strong association with aging, and recent learning-based approaches have improved predictive performance. However, most existing methods treat CpG sites as independent features, overlooking the complex and heterogeneous biological relationships among them. We propose RelAge-GNN, a multi-relational graph neural network framework for DNA methylation-based age prediction. Our method constructs three complementary graphs capturing co-methylation patterns, genomic co-localization, and gene-level associations among CpG sites. Each graph is modeled by an independent GNN branch, and a learnable gating mechanism adaptively fuses the resulting representations. Experiments on large-scale datasets show that RelAge-GNN achieves competitive accuracy and stronger correlation with chronological age compared to state-of-the-art methods. Moreover, the model exhibits improved sensitivity in detecting age acceleration across diverse disease cohorts, highlighting its potential utility for disease characterization. Finally, through post hoc interpretability analyses, we quantify the contributions of different relational structures and CpG sites, providing biologically meaningful insights and suggesting potential directions for aging-related research. Our code is available at: https://anonymous.4open.science/r/RelAge-GNN-F1E3/.

Problem

• DNA methylation
• biological age estimation
• CpG sites
• multi-relational graph
• aging clocks

Innovation

• multi-relational graph neural network
• DNA methylation
• biological age estimation
• CpG site relationships
• age acceleration