Genetic Drift Versus Climate Region Spreading Dynamics of COVID-19

Background: The current propagation models of COVID-19 are poorly consistent with existing epidemiological data and with evidence that the SARS-CoV-2 genome is mutating, for potential aggressive evolution of the disease. Objectives: We looked for fundamental variables that were missing from current analyses. Among them were regional climate heterogeneity, viral evolution processes versus founder effects, and large-scale virus containment measures. Methods: We challenged regional versus genetic evolution models of COVID-19 at a whole-population level, over 168,089 laboratory-confirmed SARS-CoV-2 infection cases in Italy, Spain, and Scandinavia at early time-points of the pandemic. Diffusion data in Germany, France, and the United Kingdom provided a validation dataset of 210,239 additional cases. Results: Mean doubling time of COVID-19 cases was 6.63 days in Northern versus 5.38 days in Southern Italy. Spain extended this trend of faster diffusion in Southern Europe, with a doubling time of 4.2 days. Slower doubling times were observed in Sweden (9.4 days), Finland (10.8 days), and Norway (12.95 days). COVID-19 doubling time in Germany (7.0 days), France (7.5 days), and the United Kingdom (7.2 days) supported the North/South gradient model. Clusters of SARS-CoV-2 mutations upon sequential diffusion were not found to clearly correlate with regional distribution dynamics. Conclusion: Acquisition of mutations upon SARS-CoV-2 spreading failed to explain regional diffusion heterogeneity at early pandemic times. Our findings indicate that COVID-19 transmission rates are rather associated with a sharp North/South climate gradient, with faster spreading in Southern regions. Thus, warmer climate conditions may not limit SARS-CoV-2 infectivity. Very cold regions may be better spared by recurrent courses of SARS-CoV-2 infection.

Accumulation of genomic mutations of SARS-CoV-2 over time around the globe, at the early time points of the pandemic. Each data-point is represented as a bead, whereby each bead corresponds to a specific set of virus mutations (mutation haplotype) (nextstrain.org/ncov/).
(bottom left) Country color codes. Beads and bead sequences were color-coded, according to the country where the virus sample was isolated from. The 'beads-on-a-string' plots link successions of viral mutations, i.e. mutation haplotypes that acquired additional mutations over time. Phylogeny trees for such mutations are presented, that draw distinct evolutionary branches of SARS-CoV-2 over calendar dates, which are indicated at the bottom of each graph. Larger beads indicate mutations identified in the indicated country. UK: bright yellow; Australia: blue; Canada: red. Rectangles enclose mutations evolutionary branches by country.

Figure S3. Genomic mutations of SARS-CoV-2 in Europe.
Accumulation of genomic mutations of SARS-CoV-2 over time in Europe, at the early time points of the pandemic. Each data-point is represented as a bead, whereby each bead corresponds to a specific set of virus mutations (mutation haplotype). Beads and bead sequences are color-coded, according to the country where the virus sample was isolated from (upper left). The 'beads-on-a-string' plots link successions of viral mutations, i.e. mutation haplotypes that acquired additional mutations over time.
Phylogeny trees for such mutations are presented, that draw distinct evolutionary branches of SARS-CoV-2 over calendar dates, which are indicated at the bottom of each graph (nextstrain.org/ncov/europe?branchLabel=aa).

Figure S4. Genomic mutations of SARS-CoV-2 -Spain.
The accumulation of genomic mutations of SARS-CoV-2 over time in Spain, at the early time points of the pandemic, is shown as strings of dark-yellow large dots. Each data-point is represented as a bead, whereby each bead corresponds to a specific set of virus mutations (mutation haplotype). Beads and bead sequences are color-coded, according to the country where the virus sample was isolated from (upper left). The 'beads-on-a-string' plots link successions of viral mutations, i.e. mutation haplotypes that acquired additional mutations over time. Phylogeny trees for such mutations are           *: Cumulative case incidence in the indicated municipalities is plotted versus time (days; calendar dates from March 1 st to April 7 th , 2020). For comparison purposes, all graphs were normalized versus the highest number of cases per province. The doubling times of COVID-19 cases were computed on non-normalized, absolute numbers of infection cases. Graphs are in alphabetical order by province name.

Figure S16. Progression of COVID-19 over Sweden*.
*: Cumulative case incidence in the indicated municipalities is plotted versus time (days; calendar dates from March 1 st to March 23 rd , 2020). For comparison purposes, all graphs were normalized versus the highest number of cases per province. The doubling times of COVID-19 cases were computed on non-normalized, absolute numbers of infection cases. Graphs are in alphabetical order by province name.