N6-methyladenosine (m 6 A) is the most abundant modification on mRNA and is implicated in critical roles in development, physiology, and disease. A major limitation has been the inability to quantify m 6 A stoichiometry and the lack of antibodyindependent methodologies for interrogating m 6 A.Here, we develop MAZTER-seq for systematic quantitative profiling of m6A at single-nucleotide resolution at 16%-25% of expressed sites, building on differential cleavage by an RNase. MAZTER-seq permits validation and de novo discovery of m 6 A sites, calibration of the performance of antibody-based approaches, and quantitative tracking of m 6 A dynamics in yeast gametogenesis and mammalian differentiation. We discover that m6A stoichiometry is ''hard coded'' in cis via a simple and predictable code, accounting for 33%-46% of the variability in methylation levels and allowing accurate prediction of m 6 A loss and acquisition events across evolution. MAZTER-seq allows quantitative investigation of m 6 A regulation in subcellular fractions, diverse cell types, and disease states. (A) Distribution of cleavage efficiencies (y axis) in RNA extracted from WT ime4D/D strains with versus without m 6 A-IP treatment. (B) Clustered pairwise correlation of the samples in (A). (C) Empirical false-detection rates per confidence group. (D) Distribution of m 6 A-seq sites across the confidence groups defined via MAZTER-seq. (E) Distribution of m 6 A-seq scores from Schwartz et al. (2013) by MAZTER-seq confidence groups. (F) Sequence logos for sites identified via MAZTER-seq shown separately for the indicated confidence groups. (G) Higher-confidence sites are closer to the end of the transcript. Distributions of 3 0 end distances by confidence group. (H) SCARLET-based readouts of methylation levels at each of the indicated sites.

No abstract