Back

Deciphering the epitranscriptomic code of RNA degradation with nanopore direct RNA sequencing

Zhang, Z.; Wang, C.-L.; Zhang, Z.-H.; Huang, Y.-F.; Xie, Y.-Y.; Zhong, Z.-D.; Tang, G.-R.; Ren, Z.-H.; Lan, Y.-L.; Kong, J.-W.; Qiao, Z.-S.; Su, T.-W.; Chen, H.-X.; Wang, Q.-Y.; Luo, R.-J.; He, J.-T.; Liu, W.-Q.; Wu, F.; Luo, G.-Z.

2026-05-10 bioinformatics
10.64898/2026.05.09.723953 bioRxiv
Show abstract

The precise regulation of RNA degradation is crucial for gene expression homeostasis, yet how multiple molecular features coordinate on a single transcript remains poorly understood. Here, we use nanopore direct RNA sequencing (DRS) to simultaneously track alternative isoforms, m6A modifications, and poly(A) tail dynamics at single-molecule resolution during a time course of RNA decay. We show that m6A regulates RNA degradation in a stoichiometry-dependent manner, where modification levels quantitatively modulate decay kinetics. Mechanistically, m6A is functionally coupled to deadenylation, promoting accelerated poly(A) tail shortening and coordinated RNA turnover. At the isoform level, we identify regional m6A clusters (RMCs) as structural elements that associate with isoform-selective degradation and remodel protein-coding potential. Furthermore, transcript splicing architecture is associated with distinct m6A deposition patterns, suggesting that gene structure encodes RNA decay kinetics through m6A-mediated regulation. A machine learning model integrating these multi-modal features highlights the central contribution of m6A and deadenylation in shaping RNA decay, while revealing substantial regulatory heterogeneity across transcripts. Collectively, our study deciphers the multi-layered, cooperative principles of RNA degradation and provides an epitranscriptomic perspective for understanding how RNA fate is encoded at the single-molecule resolution.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.