Fig. 4: Divergent rules governing m⁶A deposition in plant and mammalian genomes.

a-c Internal exons of human (a), rice (b), and Arabidopsis (c) transcripts were grouped into 100 bins of equal size based on their length, and the average m⁶A level for each bin was plotted against exon length, represented by blue dots. d-f ‘m⁶A density’ of each bin of the internal exons of human (d), rice (e), and Arabidopsis (f) transcripts were shown against exon length, represented by yellow dots. m⁶A density was calculated as the m⁶A level within each exon, normalized by its length and multiplied by 1,000. g-i All internal exons were aligned at their internal exon junction sites in human (g), rice (h), and Arabidopsis (i) genomes, and the overall m⁶A level per sliding window in the flanking regions was shown against the distance to exon junction sites, represented by brown dots. j-l Distribution of ‘m⁶A likelihood’ near the internal exon junction in human (j), rice (k), and Arabidopsis (l) transcripts were shown in dark green line, with 95% confidential intervals shadowed. m-o Similar to panel (a-c) but the average m⁶A level per exon in the last exons of human (m), rice (n), and Arabidopsis (o) genomes were shown. p-r Similar to panel (d-f) ‘m⁶A density’ in the last exons of human (p), rice (q), and Arabidopsis (r) genomes. s-u Similar to panel (g-i) m⁶A level per sliding window flanking last exon junction site in human (s), rice (t), and Arabidopsis (u) genomes. v-x Similar to panel (j-l) Distribution of ‘m⁶A likelihood’ near the last exon junction in human (v), rice (w), and Arabidopsis (x) genomes. y, Diagram showing the inhibition mode in humans and activation mode in plants contributes to distinct m⁶A distribution pattern. For j-l and v-x data are presented as median values.