This repository contains scripts and data accompanying a manuscript recently submitted to medrxiv.
A novel coronavirus (SARS-CoV-2) first detected in Wuhan, China, has spread rapidly since December 2019, causing over 45,000 PCR confirmed infections and more than 1,000 fatalities (as of Feb 12, 2020). Imported cases and small transmission clusters have been reported globally. Early data suggest the virus transmits readily and a pandemic cannot be ruled out. Data from routine diagnostics show a strong and consistent seasonal variation of the four endemic coronaviruses (229E, HKU1, NL63, OC43). We use these data to explore the effect of seasonal variation in transmissibility on a potential SARS-CoV-2 pandemic. A model allowing for many subpopulations of different size with variable parameters of SARS-CoV-2 spread shows how a pandemic could unfold in 2020-2022. Simulations of different scenarios show that plausible parameters result in a peak in temperate regions of the Northern Hemisphere in winter 2020/2021. A smaller range of parameters suggests a peak in the first half of 2020 or two peaks of similar magnitude. Variation in transmission and migration rates can result in substantial variation in prevalence between regions.
While the uncertainty in parameters is large, the scenarios we explore show that transient reductions in the incidence rate might be due to a combination of seasonal variation and infection control efforts but do not necessarily mean the epidemic is contained. Seasonal forcing on SARS-CoV-2 should thus be taken into account in the further monitoring of the global transmission. The likely aggregated effect of seasonal variation, infection control measures, and transmission rate variation is a prolonged pandemic wave with lower prevalence at any given time, thereby providing a window of opportunity for better preparation of health care systems.
The directory data
contains tab delimited files with positive and negative tests for the different seasonal CoVs and a summary of the age distribution.
These files were generated using the script descriptive_graphs.py
that ingests the raw data base output and distills it into these tables.
The manuscript is written in LaTex. TEX source files and the bibliography are contained in the directory manuscript
The different figures of the manuscript are generated by scripts contained at the top-level of this repository.
descriptive_graphs.py
generates Fig 1 detailing the variation of seasonal CoV prevalence.fit_seasonal.py
simulates a large range of parameters and compares the output to seasonal CoV prevalence. This generates Fig 2 and the corresponding supplementary figures.scenarios.py
generates Fig 3a,scenarios2.py
the corresponding supplementary figure.peak_ratio.py
scans parameters and compares the height of first and second peak. These results are contained in Fig 3b and the corresponding supplementary graphscompartment_model.py
simulates 1000 coupled populations and generates Fig. 4 and 5 as well as the corresponding supplementary figures.