Genetic Spectrum and Distinct Evolution Patterns of SARS-CoV-2

Abstract

Four signature groups of frequently occurred single-nucleotide variants (SNVs) were identified in over twenty-eight thousand high-quality and high-coverage SARS-CoV-2 complete genome sequences, representing different viral strains. Some SNVs predominated but were mutually exclusively presented in patients from different countries and areas. These major SNV signatures exhibited distinguishable evolution patterns over time. A few hundred patients were detected with multiple viral strain-representing mutations simultaneously, which may stand for possible co-infection or potential homogenous recombination of SARS-CoV-2 in environment or within the viral host. Interestingly nucleotide substitutions among SARS-CoV-2 genomes tended to switch between bat RaTG13 coronavirus sequence and Wuhan-Hu-1 genome, indicating the higher genetic instability or tolerance of mutations on those sites or suggesting that major viral strains might exist between Wuhan-Hu-1 and RaTG13 coronavirus.

Keywords: COVID-19; SARS-CoV-2 genome; clustering; co-infection; evolution; genetic variants.

FIGURE 1

SNVs on about thirty thousand SARS-CoV-2 complete genomes. (A) Circos plot shows distribution, frequency, and co-occurrences of SNV2. From outer to inner circle: coronavirus genome location (nt), gene annotation, occurrence ratios of SNVs at the site (log10 scale, red bars), and connections with high concurrence rates (>0.9) represented by blue lines. The darker the blue lines, the higher concurrence rates. (B) Fifty-four high frequent SNVs with annotated AA changes were detected (in purple) in about twenty-eight thousand patients worldwide. Four major clusters of SNVs and consequent subgroups can be formed to represent patients from different geographical locations.

FIGURE 2

Frequencies of signature SNVs in worldwide and top three countries/areas. (A–D) Four major groups and/or their sub-groups had distinct representing countries/areas, indicating different transmission sources and evolution paths.

FIGURE 3

SNVs migration and evolution patterns over time. (A) SNV T28144C with C18060T and additional C17747T/A17858G spread in United States and other countries/areas from January to March of 2020. (B) SNVs of C14408T and A23403G with C1059T spread in United States and other countries/areas from January to March of 2020. (C) The ratios of four significant groups of SNVs, (A–D) in Figures 1 and 2, varied in different countries/areas with time development. (D) Average temporal ratios of groups (A–D) SNVs show distinct patterns from January to June of 2020.

FIGURE 4

Number of patients carrying significant groups (A–D) of SNVs, indicating potential co-infection by different SARS-CoV-2 strains.

FIGURE 5

Comparison between SARS-CoV-2 SNVs and nucleotides of bats (RaTG13, bat-SL-CoVZC45, bat-SL-CoVZXC21, and RmYN02) coronavirus varying from Wuhan-Hu-1. (A) Percentage of SARS-CoV-2 SNVs on different regions, including SARS-CoV-2 complete genome, and sites where bats coronavirus differ from Wuhan-Hu-1; (B) Percentage of SARS-CoV-2 SNVs which converted to bats nucleotides within the same regions as shown in (A).

FIGURE 6

Structures of SARS-CoV-2 non-structural protein 2 (nsp2) and Spike near specific mutation sites. (A) No clash was predicted between nsp2 GLY212 and nsp2 ASN183 in Wuhan-Hu-1. (B) A clash was predicted between ASP212 and ASN183 after the mutation G212D. (C) There was interaction between T859 and D614. (D) No contact was precited between T859 and G614 after the mutation D614G.