Back

Virus evolution affected early COVID-19 spread

Corcoran, D.; Urban, M. C.; Wegrzyn, J.; Merow, C.

2020-09-30 epidemiology
10.1101/2020.09.29.20202416 medRxiv
Show abstract

As the SARS-Cov-2 virus spreads around the world afflicting millions of people, it has undergone divergent genetic mutations. Although most of these mutations are expected to be inconsequential, some mutations in the spike protein structure have been hypothesized to affect the critical stage at which the virus invades human cells, which could affect transmission probability and disease expression. If true, then we expect an increased growth rate of reported COVID-19 cases in regions dominated by viruses with these altered proteins. We modeled early global infection dynamics based on clade assignment along with other demographic and meteorological factors previously found to be important. Clade, but not variant D614G which has been associated with increased viral load, enhanced our ability to describe early COVID-19 growth dynamics. Including clade identity in models significantly improved predictions over earlier work based only on weather and demographic variables. In particular, higher proportions of clade 19A and 19B were negatively correlated with COVID-19 growth rate, whereas higher proportions of 20A and 20C were positively correlated with growth rate. A strong interaction between the prevalence of clade 20C and relative humidity suggests that the impact of clade identity might be more important when coupled with certain weather conditions. In particular, 20C an 20A generate the highest growth rates when coupled with low humidity. Projections based on data through April 2020 suggest that, without intervention, COVID-19 has the potential to grow more quickly in regions dominated by the 20A and 20C clades, including most of South and North America.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.