Structure-aware deep learning enhances m6A prediction and reveals cell type-associated RNA structural signatures

Mingze Sun, Di Zhang, Zhiyuan Li and Yihan Lin

PLOS Computational Biology, 2026, vol. 22, issue 8, 1-19

Abstract: N6-methyladenosine (m6A), the most abundant mRNA modification in eukaryotes, plays essential roles in gene regulation and disease pathogenesis. Computational prediction of m6A sites offers a scalable alternative to costly experimental approaches, yet current methods rely predominantly on linear sequence features. This overlooks potentially informative RNA structural context, which is associated with local methylation patterns and may provide complementary predictive information beyond linear sequence motifs. To incorporate this complementary information, we propose SMART-m6A (Sequence–structure Multifeature Attention RNA Transformer for m6A), a deep learning framework that integrates sequence and structural information through parallel convolutional feature extraction and structure-guided attention for multifeature fusion. SMART-m6A achieves superior predictive performance compared to existing methods, with particularly clear advantages in sequence-ambiguous candidates. Beyond prediction accuracy, learned attention patterns reveal strong concordance with experimentally validated m6A-binding protein recognition sites and identify potentially novel regulatory motifs. Through systematic ablation studies and targeted structural-input perturbation analyses, we show that sequence and structure provide complementary predictive information, and that sites with greater prediction sensitivity to structural perturbation exhibit distinct local structural profiles between cell lines. Collectively, this work demonstrates the predictive value of sequence-derived structural features in m6A modeling and provides a multifeature deep learning framework for accurate and interpretable structure-aware epitranscriptomic prediction.Author summary: Chemical modifications of RNA provide a crucial regulatory layer beyond the genetic code. Among them, N6-methyladenosine (m6A) is the most abundant internal modification in eukaryotic messenger RNA and has been implicated in diverse biological processes, including development, immune regulation, and tumorigenesis. Understanding how m6A is deposited across transcripts is therefore essential for deciphering epitranscriptomic regulation and its roles in health and disease. Although high-throughput technologies can map m6A sites genome-wide, these experiments remain expensive and technically demanding, motivating the development of computational prediction methods. However, most existing algorithms rely primarily on linear RNA sequence features and largely ignore RNA secondary structure, which has been associated with RNA-binding protein recognition and m6A-related regulatory contexts. Here, we introduce SMART-m6A, a structure-aware deep learning framework that integrates RNA sequence and secondary structure through a multifeature neural architecture with structure-guided attention. By jointly modeling these complementary sources of information, SMART-m6A achieves improved predictive accuracy across multiple benchmark datasets. Importantly, the model is not only predictive but also biologically informative: it identifies structural features associated with m6A deposition, reveals that cell type–specific m6A sites exhibit greater structural variability, and uncovers sequence motifs linked to known RNA-binding proteins. Together, our results indicate that sequence-derived structural features are informative for m6A prediction and highlight the value of structure-aware machine learning approaches for understanding epitranscriptomic mechanisms.

Date: 2026
References: Add references at CitEc
Citations:

Downloads: (external link)
https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1014649 (text/html)
https://journals.plos.org/ploscompbiol/article/fil ... 14649&type=printable (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:plo:pcbi00:1014649

DOI: 10.1371/journal.pcbi.1014649

Access Statistics for this article

More articles in PLOS Computational Biology from Public Library of Science
Bibliographic data for series maintained by ploscompbiol ().