Skip to main content
. 2019 Nov 8;15(11):e1008493. doi: 10.1371/journal.pgen.1008493

Fig 4. Recombination contributes to GC content locally but cannot explain the relationship between GC content and Ku incidence.

Fig 4

(a) The mean GC content of genes with evidence for recombination (PHI statistic, [45], see Methods) plotted against the mean GC content of genes without evidence for recombination in a given closely related cluster of organisms (ATGC database [42]). Recombining genes have slightly higher GC content than non-recombining genes (points mostly lie above the dashed x = y line). (b) The difference in GC content for recombining and non-recombining genes within a cluster is smaller for Ku-encoding than Ku-lacking clusters. Clusters classified as Ku-lacking if no members encoding Ku (n = 114) and Ku-encoding if at least one member has Ku (n = 41). Clusters excluded if no evidence for recombination was found for any of their genes.