Strong evidence for the evolution of decreasing compositional heterogeneity in SARS-CoV-2 genomes during the pandemic

The rapid evolution of SARS-CoV-2 during the pandemic was characterized by the fixation of a plethora of mutations, many of which enable the virus to evade host resistance, likely altering the virus' genome compositional structure (i.e., the arrangement of compositional domains of varying lengths and nucleotide frequencies within the genome). To explore this hypothesis, we summarize the evolutionary effects of these mutations by computing the Sequence Compositional Complexity (SCC) in random stratified datasets of fully sequenced genomes. Phylogenetic ridge regression of SCC against time reveals a striking downward evolutionary trend, suggesting the ongoing adaptation of the virus's genome structure to the human host. Other genomic features, such as strand asymmetry, the effective number of K-mers, and the depletion of CpG dinucleotides, each linked to the virus's adaptation to its human host, also exhibit decreasing phylogenetic trends throughout the pandemic, along with strong phylogenetic correlations to SCC. We hypothesize that viral CpG depletion (throughout C➔U changes), promoted by directional mutational pressures exerted on the genome by the host antiviral defense systems, may play a key role in the decrease of SARS-CoV-2 genome compositional heterogeneity, with specific adaptation to the human host occurring as a form of genetic mimicry. Overall, our findings suggest a decelerating evolution of reduced compositional complexity in SCC, whereas the number of K-mers and the depletion of CpG dinucleotides are still increasing. These results indicate a genome-wide evolutionary trend toward a more symmetric and homogeneous genome compositional structure in SARS-CoV-2, which is partly still ongoing.