Fig. 3
From: A potentially abundant junctional RNA motif stabilized by m6A and Mg2+
Predicted secondary structures of m6A containing RNA sequences. a Predicted RNA secondary structures were classified into the following motifs: Junctional Watson-Crick m6A–U or mismatch m6A–Y (Y denotes A or C or G.) with the bulge located 5ʹ (5ʹB-[A–U] and 5ʹB-[A–Y]) or 3ʹ ([A–U]-B3ʹ and [A–Y]-B3ʹ) to m6A; or with the bulge located 5ʹ or 3ʹ (5ʹB-[X-A]/[X-A]-B3ʹ, X denotes A or C or U or G) to m6A partner nucleotide. b Distribution of the lowest energy predicted structures for 140,574 m6A sites identified using individual-nucleotide-resolution cross-linking and immunoprecipitation (miCLIP)43 in the human transcriptome (hg19) were predicted using RNAstructure44 and MC-Flashfold39. Secondary structures were predicted for RNA sequences of 41-nt long with m6A (or A) in the middle. Also shown are corresponding predictions for 140,574 unmodified adenine sites selected randomly from the same human transcriptome. c Population of m6A and A sites predicted (MC-Flashfold) to fold into the 5ʹB-[A–U] and 5ʹB-[A–X] conformation within different energetic thresholds relative to the lowest energy structure
