a. The non-specific rate (left) of MeRIP using normal or competitive washing. The non-specific IP rate was calculated from the LC/MS-MS data using the ratio of m(6)A/A in the input sample divided by that in the IP sample. Meanwhile, ssRNA with or without m(6)A was used to validated the signal to noise ratio of these two methods (right). Enrichment of IPed versus input RNA is normalize against ssRNA without m(6)A. Two independent experiments were performed with bar representing the average value (see source data in Supplementary Table 5). b. Similarity (Pearson’s correlation) of gene expression levels between each pair of samples. The FPKM of genes from input were used. The samples were hierarchically clustered (see source data in Supplementary Table 5). c. Fraction of overlapped m(6)A peaks called by MACS2 and MeTPeak/moving-window. d. Sequence logo and p value of deduced consensus motif of m(6)As in each tissue. n= 36461, 27866, 45503, 26163, 24180, 29174, 38444, 21780 m(6)As in brain, heart, kidney, liver, lung, muscle, placenta and stomach respectively, binomial distribution test (see methods for details). e. Distribution of m(6)A peaks surrounding the CDS in mRNA regions in each tissue. f. The fraction of m(6)A peaks in heart, kidney, liver, lung, muscle, placenta, and stomach in the 5′ UTR, CDS, intron, stop codon, 3′ UTR, promoter, and intergenic regions.