Deep learning modeling mA deposition reveals the importance of downstream cis-element sequences.
Journal:
Nature communications
PMID:
35581216
Abstract
The N-methyladenosine (mA) modification is deposited to nascent transcripts on chromatin, but its site-specificity mechanism is mostly unknown. Here we model the mA deposition to pre-mRNA by iM6A (intelligent mA), a deep learning method, demonstrating that the site-specific mA methylation is primarily determined by the flanking nucleotide sequences. iM6A accurately models the mA deposition (AUROC = 0.99) and uncovers surprisingly that the cis-elements regulating the mA deposition preferentially reside within the 50 nt downstream of the mA sites. The mA enhancers mostly include part of the RRACH motif and the mA silencers generally contain CG/GT/CT motifs. Our finding is supported by both independent experimental validations and evolutionary conservation. Moreover, our work provides evidences that mutations resulting in synonymous codons can affect the mA deposition and the TGA stop codon favors mA deposition nearby. Our iM6A deep learning modeling enables fast paced biological discovery which would be cost-prohibitive and unpractical with traditional experimental approaches, and uncovers a key cis-regulatory mechanism for mA site-specific deposition.