Fig. 6: The Kozak sequence contextual characteristics that influence upstream open reading frame (uORF) translation.

a Relationship between the Kozak scores and normalized translational initiation signals of uORF start codons (uoAUGs) in human HEK293 cells, mouse MEF cells, and fly S2 cells. In each sample, we ranked uORFs based on increasing Kozak scores and divided them into 50 bins (100 bins for S2 cells) with equal numbers of uoAUGs. The median Kozak score and normalized TIS signal for each bin were used to calculate Spearman’s correlations (ρ) and two-sided P values. The linear fit was indicated with a blue line. b The distribution of Spearman’s correlation coefficients between the coding sequence (CDS) Kozak scores and the number of uORFs for that gene in eukaryotes in different taxa. In the left panel, each dot represents one species. The right panel shows that in humans, genes that have multiple uORFs tend to have weaker Kozak sequence context around the start codon of CDSs. P_adj, two-sided P value after correction for multiple testing; NS, not significant. Box plots showing the distribution of the Euclidian distance of the position weight matrix of Kozak sequences (PWMK) for cAUGs (c) and uoAUGs (d) between species within the same taxa (brown) or species in different taxa (green). Differences in distances were compared with two-sided Wilcoxon rank-sum tests. Exact P values (no correction for multiple testing were made) and the number of pairwise distances in each group were shown in the plot. Center line, median; box limits, upper and lower quartiles; whiskers, 1.5 times the interquartile range. Source data are provided as a Source Data file.