Fig. 3: Changes in C→U preference over the course of SARS-CoV-2 divergence.
From: Mutational spectrum of SARS-CoV-2 during the global pandemic
a Phylogenetic tree of the closest natural relatives of SARS-CoV-2, which was adapted from GISAID. Substitutions between ancestral and descendant sequences were collected in the root-to-leaf direction. The substitution ratio of C→U over U→C is marked by color on each edge (reddish for more C→U and greenish for more U→C). b–i The substitution spectra for 12 major classes are shown for the spread among humans (b) and for each divergence lineage indicated in the middle of branches (c–i). There are more C→U than U→C changes in human SARS-CoV-2 (b), while the reverse pattern is shown in the divergence of bat coronaviruses (c–i). Although there is uncertainty in the mutation calls for old lineages due to a high degree of mismatch between sequences, one upper branch suggests more C→U than U→C changes (i). The status of the ancestral sequence requires the use of information outside the node. The pangolin coronavirus from Guanxi (asterisks) is a consensus sequence of the following six sequences from the same study: P3B, P2V, P5E, P5L, P1E, and P4L. For divergence among these sequences, mutation counts per class and the C→U/U→C ratio are not displayed because their mutation count is not sufficiently large to reliably represent them.
