Figure 3. 5’ UTR regulatory elements shape translation initiation in early development.

(A) Effect of 5´ UTR RNA structures on ribosome recruitment. Scatter plot of RNA folding scores calculated by MXfold2 (x axis) and mean log₂ transformed RRS values (y axis) for the HMW fraction at 10 hpf. (B) Effect of 5´ UTR length on ribosome recruitment. Scatter plot of 5´ UTR length in nucleotides (x axis) and mean log₂ transformed RRS values (y axis) for the HMW fraction at 10 hpf. Each dot represents one 5´ UTR (n = 17,906) and adjusted R² values for best-fit to a nonlinear generalized additive model and corresponding p-value are presented. (C) Effect of upstream open reading frame (uORF) number on ribosome recruitment. Violin plots of mean log₂ transformed RRS values for the HMW fraction at 10 hpf for 5´ UTRs binned by number of predicted uORFs. Center line of boxplots represent median, box limits represent the upper and lower quartiles. (D) Mean nucleotide frequency at positions −4 to −1 of the reporter canonical sfGFP translation start site for 5´ UTRs with lowest RRS values (bottom 10%, n = 1,791; left panel) and highest RRS values (top 10%, n = 1,791; right panel) for the HMW fraction at 10 hpf. Position weight matrixes of poor and strong TIS consensus context derived from mean nucleotide frequencies are presented. Grey AUG represents the translation start site of sfGFP. See STAR Methods, Figure S3 and Table S3.