Fig. 5: Synonymous codons may influence m⁶A deposition.

a The saturation synonymous codon swap strategy. For each codon in codons (−15 to +15 codons, −45 nt to +45 nt) around the m⁶A site, the codon was substituted by each of its synonymous codons. The delta changes of m⁶A probability value (ΔProbability) after swap was calculated by iM6A. b Positional plot of ΔProbability (cutoff = 0.1) for saturation synonymous codon swap. Red color dots were those codon swap events that increased m⁶A probability; Green color dots were those codon swap events that decreased m⁶A probability. c The systematic and quantitatively determination of the effect on m⁶A deposition for all codons by linear regression (See details in Methods). Total number of m⁶A sites used was 1473. (The full list of effect values for each synonymous codon are listed in Supplementary Data 3). d Bar plot of effect values for top 10 enhancer codons and top 10 silencer codons, its corresponding amino acids were also labeled. e Synonymous codon pair in enhancing and silencing m⁶A deposition. The amino acids (the first column) were coded by the synonymous codons (the second and third columns, respectively), which the enhancer codons enhancing m⁶A deposition, and the silencer codons silencing m⁶A deposition. f Scatter plot for the effect correlation for all synonymous codons between the study of human synonymous codons and the study of mouse synonymous codons. The effect of each synonymous codon was determined by the slope of its linear regression equation, and each gray dot was a synonymous codon. g, h Positional plot for the frequency of Top 20 enhancer codons (g) or silencer codons (h) in the sequences around the m⁶A sites. The plots were compared between higher m⁶A probability sites (red color, probability ≥ 0.7) and lower m⁶A probability sites (the exact RAC motif matched control, probability < 0.01). Data were presented as mean ± SEM. (Using other top number of enhancer or silencer codons generated similar results).