Bridging Sequence and Graph Structure for Epigenetic Age Prediction

2026-05-11

AI Summary

Existing methods struggle to jointly model the co-methylation graph structure of DNA methylation sites and their sequence context within a unified framework. This work proposes a sequence–graph fusion architecture that integrates eight handcrafted DNA sequence-derived statistical features—such as CpG density—via a lightweight gated modulation mechanism to adaptively scale methylation signals before feeding them into a graph convolutional network for epigenetic age prediction. By simultaneously capturing both sequence context and graph topology, the approach achieves a mean absolute error of 3.149 years on 3,707 blood samples, outperforming the strongest graph-based baseline by 12.8%. Furthermore, it reveals the dynamic importance of key sequence features during aging, demonstrating superior performance over end-to-end CNN-based representations.

Abstract

Epigenetic clocks based on DNA methylation have emerged as powerful tools for estimating biological age, with broad applications in aging research, age-related disease studies, and longevity science. Despite advances across machine learning approaches to epigenetic age prediction, spanning penalised linear regression, deep feedforward networks, residual architectures, and graph neural networks, no existing method jointly models co-methylation graph structure and site-specific DNA sequence context within a unified framework. We propose a unified sequence--graph integration framework for epigenetic age prediction that addresses this gap, integrating eight-dimensional DNA sequence statistical features through a lightweight gated modulation mechanism that adaptively scales each site's methylation signal according to its sequence-determined biological relevance prior to graph convolution. Evaluated on 3,707 blood methylation samples against a comprehensive set of baselines, our method achieves a test MAE of 3.149 years, a 12.8\% improvement over the strongest graph-based baseline. Biologically informed statistical features outperform CNN-based sequence encoding, demonstrating that handcrafted sequence features are more effective than end-to-end learned representations in this data regime. Post-hoc interpretability analysis identifies CpG density and local adenine frequency as features with age-dependent importance shifts, consistent with known mechanisms of age-related hypermethylation at CpG-dense promoter regions. Our code is at https://github.com/yaoli2022/graphage-seq.

Problem

• epigenetic age prediction
• DNA methylation
• graph structure
• sequence context
• co-methylation

Innovation

• sequence-graph integration
• epigenetic age prediction
• gated modulation
• DNA methylation
• graph neural networks