Determinative Role of Non-structural Protein Genes in Genus Identification of Coronaviruses

Interrelationship of coronavirus genus with key fragments of viral genome was investigated. Genes of structural proteins (S-gene of spike protein and N-gene of nucleocapsid protein) and ORF1ab of polyprotein pp1ab, that in infected cell is split into 16 non-structural proteins, were considered as such fragments. Statistical method based on averaged codon distribution in the genes of genus prototype variants was applied in the work to recognize genus of coronavirus. High reliability of this method has been demonstrated earlier in recognizing the 15 species and serotypes of the flaviviruses, such as viruses of yellow fever, dengue fever, various encephalitides, etc. For each key fragment of the coronavirus genome the numerical experiments on identification of genus for the 3242 viral genomes from the GenBank have been done. The highest reliability (98
^%
) was achieved, when ORF1ab frequency codon characteristics were used. It appeared to be that in recognizing genus of Gammacoronavirus, basing on spike protein gene, about half of the 345 genomes of this genus were identified as Betacoronavirus (84.6
^%
) and Alphacoronavirus (15.4
^%
). Analogous phenomenon of significant error appeared in determinating Alphacoronavirus genus, basing on nucleocapsid protein gene, also. However, these significant errors may be a consequence of the coronavirus genome plasticity in the result of homologous recombinations between the viral genomes.