m⁶A-FORM: An m⁶A-focused Foundation Model for Decoding m⁶A Regulatory Function

N6-methyladenosine (m⁶A) regulates mRNA fate through site-specific methylation, reader recognition and downstream effects on RNA stability and decay. However, current computational approaches focus mainly on site prediction, leaving unresolved the broader challenge of inferring m⁶A regulatory context and function from epitranscriptomic profiles. Here we present m⁶A-FORM, an m⁶A-focused foundation model for for regulatory discovery. Pretrained on 24.9 million RNA sequence windows from 22.5 million MeRIP-seq regions across 143 human studies, m⁶A-FORM learns reusable representations of m⁶A-associated transcript contexts. We adapt this encoder to single-nucleotide m⁶A discovery, regulator-binding prediction, YTHDF2-associated decay prediction and tissue-scale epitranscriptomic mapping. m⁶A-FORM predicts binding of 19 m⁶A readers, writers and erasers and identifies sequence and RBP-context features associated with YTHDF2-mediated RNA degradation. Applied to 67 datasets from 24 human tissues, it identifies tissue-conserved m⁶A sites linked to stronger methylation, reader binding, RBP occupancy and decay propensity.