Fig. 4
From: Origin of exon skipping-rich transcriptomes in animals driven by evolution of gene architecture
Relationship between gene structure and sequence composition and ES. a Heatmap representing the distance between distributions of gene architecture values for ES-positive and ES-negative exons, measured with the D statistic of the one-sided Kolmogorov-Smirnov two-sample test (significant if p < 0.01; otherwise grey). D values are recorded as positive/negative (blue/red) according to two one-sided Kolmogorov-Smirnov tests with complementary alternative hypotheses: positive/negative D (blue/red) reflects a positive/negative relationship between ES and the given feature. b Heatmap representing the correlation between intron length and GC content across 65 eukaryotes. Positive correlations (long, GC-rich introns, Spearman’s rho > 0 and p < 0.01) are shown in blue, negative correlations in red. c Heatmap representing genome-level gene architectural traits, for reference: intron density (introns/gene), median exon and intron length (bp), and GC content (percentage). d Difference in intron lengths (y-axis) between homologous sites of S. arctica compared to C. fragrantissima or C. owczarzaki, ranking introns according to their length differential with S. arctica (x-axis). Black arrows indicate the percentage of introns that are longer in S. arctica than in the other species (totals between brackets). e Difference in intron lengths (y-axis) between homologous sites of V. carteri compared to C. reinhardtii or C. variabilis, ranking introns according to their length differential with V. carteri (x-axis). Black arrows indicate the percentage of introns that are longer in V. carteri than in the other species (totals between brackets). f GC content in exons (GCex) and introns (GCin) for each species in the dataset. Note that GCex > GCin in all species except for the glaucophyte Cyanophora paradoxa