Multivariate Analyses of Codon Usage of SARS-CoV-2 and other betacoronaviruses

Haogao Gu; Daniel Chu; Malik Peiris; Leo L.M. Poon

portant for viral entry. The 302 genetic variability is extreme in spike gene 27 , and this highly mutable gene may possess 303 valuable information about recent evolution history. In our results, the synonymous codon 304 usage of SARS-CoV-2 in spike gene was distinct from those of RaTG13 and other 305 phylogenetic relatives ( Figure 3A) , which was not observed in orf1ab or nucleocapsid gene. 306 Although the codon usage in spike of SARS-CoV-2, RaTG13 and P1E were similar at amino 307 acid level, the difference at synonymous codon usage level indicates that they are unlikely to 308 share a very recent common ancestor. It is more likely that SARS-CoV-2, RaTG13 and P1E 309 might have undergone different evolution pathways for a certain period of time. The amino 310 . CC-BY-NC-ND 4.0 International license author/funder. It is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.02.15.950568 doi: bioRxiv preprint acid usage of SARS-CoV-2 in membrane was clustered with bat SARSr-CoV, however the 311 synonymous codon usage of SARS-CoV-2 was still distinct to these bat coronaviruses. 312 Notably, in membrane gene, pangolin P1E had a more similar synonymous codon usage to 313 SARS-CoV-2 than RaTG13. These findings suggest that there may be different selection 314 forces between genes. Our result supports different evolutionary background or currently 315 unknown host adaption history in SARS-CoV-2. The codon usage of SARS-CoV-2 in 316 nucleocapsid gene was similar to bat SARSr-CoV both at amino acid level and synonymous 317 level, suggesting that no highly significant mutation happened in this gene. 318 Codon usage can be shaped by many different selection forces, including the influence from 319 host factors. Some researchers have hypothesised that the codon usage in SARS-CoV-2 320 maybe directly correlated to the codon usage of its host 28 . However our recent study on 321 influenza A viruses implied that these may not be the most influential factors shaping the 322 codon usage of a viral genome 19 . Our analysis took advantage of the existing genomes of 323 Betacoronavirus to study the complex host effect on codon usage, which warrants more 324 accurate but relatively conserved estimation. 325