Fig. 7. Experimental verification of canonical and noncanonical upstream open reading frames (uORFs).

a The scheme showing how to determine the effect of uORF variations in the human population on the translation efficiency (TE) of downstream coding sequences (CDSs). With the mRNA-Seq and Ribo-Seq data of 60 human lymphoblastoid cell lines¹¹⁵, we calculated the translation efficiency of CDS for each gene and obtained the genotypes of each subject from the 1000Genomes Project. For each uORF variant, we performed a linear regression between the number of non-uORF alleles and the log₂(TE) of downstream CDS across the 60 cell lines. b The distribution of slopes in the linear regression between genotypes (the number of non-uORF alleles) and CDS translational efficiency among 60 cell lines. Center line, median; box limits, upper and lower quartiles; whiskers, 1.5 times the interquartile range. The number of variants in each category was shown in the plot. Exact P values of two-sided Wilcoxon signed-rank tests were shown in the plot. The ratio of relative luciferase intensity (log₂) between the reporters with the uORF allele or the non-uORF allele for each variant of canonical uORFs (c) or noncanonical uORFs (d). The bars are displayed in blue or red when the relative intensity of uORF-allele is significantly lower or higher than that of the non-uORF allele (one-sided Wilcoxon rank-sum tests, P_adj < 0.1), respectively. Measures of center, mean; error bars, standard errors. n = 4 or 5 independent biological replicates for each variant (details are presented in source data). Source data are provided as a Source Data file.