The box plots in panels A, B, and C compare G C content, normalized predicted minimum free energy (M F E), and maximum base pairing of human introns (length less than 5000 nucleotides) to introns from human viruses. The x-axis represents the categories Human and Virus, and the y-axis represents G C content percentage (panel A), M F E in kilocalories per mole per base (panel B), and maximum base pairing (panel C). The data are shown as box-and-whisker plots rather than simple bars. Human introns are represented in teal, and viral introns are represented in pink. Panel A shows no significant difference in G C content between human and viral introns, indicated by n s. Panel B shows that viral introns have significantly higher normalized predicted MFE compared to human introns, indicated by four asterisks. Panel C shows that viral introns have significantly lower maximum base pairing compared to human introns, also indicated by four asterisks. Panel D presents a schematic of the computational experiment. The observed M F E of each intron is compared to the expected MFE derived from shuffled versions of the same intron. The shuffle sequence preserves dinucleotide content (as indicated in the diagram). For each intron, 10 shuffled sequences are generated, the M F E is calculated for each, and the mean M F E of the shuffled ensemble is obtained. Panels E and F plot the minimum M F E of shuffled ensemble versus observed M F E for human introns (panel E) and viral introns (panel F), respectively. The plots show strong correlations, with r squared equals 0.84 for human introns and r squared equals 0.89 for viral introns. Panel G shows the density distribution of Observed MFE minus Mean M F E of the shuffled ensemble for human and viral introns. Viral introns show a significantly different distribution compared to human introns, indicated by four asterisks. All values are approximated based on the figure.
Viral introns are less structured than 90% of human introns. (A–C) Human introns (length <5,000 nt) compared with 126 introns from human viruses according to their distributions of (A) GC content, (B) normalized predicted MFE intron length ≤5,000 nt), and (C) potential for intron base pairing (MBP, see Materials and methods). P values were from t test. ns, not significant; ****P < 0.0001. (D) Schematic describing computational experiments comparing the observed predicted MFE of an intron to the expected value (predicted MFE of an ensemble of 10 shuffles of the same intron). (E) For human introns, each intron’s observed MFE is plotted against the mean MFE derived from the intron’s shuffled ensemble described in D. (F) For human virus introns, each intron’s observed MFE is plotted against the mean MFE derived from the intron’s shuffled ensemble described in D. (G) The distribution of the difference between observed MFE and mean MFE of shuffled ensemble for human and viral introns. P value was from t test. ****P < 0.0001.