Scaling of SARS-CoV-2 Genomic Divergence with Cumulative Infections During the COVID-19 Pandemic

The COVID-19 pandemic constitutes the largest natural experiment in viral evolution ever observed. Unlike previous analyses that have used time as the independent variable, this study uses cumulative infections — a more biologically meaningful measure of replication opportunity that is not confounded by variation in transmission intensity.  Analysis of 1,280 SARS-CoV-2 genomes from three years of global transmission paired with cumulative infection estimates from the Institute for Health Metrics and Evaluation (IHME) found that substitutions relative to Wuhan-Hu-1 accumulated at a rate of 8.9 per billion human infection events. The observed pandemic scaling relationship suggests that producing ~1,000 fixed nucleotide differences would require replication opportunities comparable to roughly 100 billion human infections. An intentionally conservative model based on confirmed COVID-19 cases, well-established as an underestimate of total infections, and substitution data that have not been filtered to remove extreme values likely to represent sequencing artifacts or transient polymorphisms predicts that billions of infections would be required. These results provide an empirical benchmark for the replication opportunity required to generate large numbers of substitutions under pandemic-scale human transmission suggesting that the divergence observed among known sarbecoviruses would require either very large host populations, extended evolutionary timescales, or both under conditions comparable to pandemic-scale human transmission.