Researchers developing a virus spread model for Sweden

2 April 2020

Researchers are hopeful that they will soon be able to present figures for how the coronavirus is likely spread in Sweden, depending on the measures implemented.

Researchers at Uppsala University are working on a model of how the new coronavirus is likely to spread in Sweden in order to provide policy-makers with a better basis on which to inform the public of the need to maintain social distancing. The model will show the effects of various levels of measures on the spread of infection.

Jasmine Gardner, post-doctoral researcher at
the Department of chemistry.

Under normal circumstances, post-doctoral researcher Jasmine Gardner of the Department of Chemistry can be found studying proteins; however, for the past two weeks she has put her own research aside. Instead of studying the behaviour of atoms, she is currently building a model calculating the spread of infection among the inhabitants of Sweden.

The model is loaded with data on the Swedish population; for example, age, occupation and geographical location. This comprehensive geographic and demographic data is supplemented with cultural parameters retrieved by the researchers from studies of countries similar to Sweden; for example, how many contacts every individual might be expected to have each day and how we travel.

In order to feed the correct data into the model, Gardner and her colleagues are collaborating with epidemiologists with knowledge of which factors affect the health of a population.

“It has been very interesting to work with researchers from different fields in this project. The various perspectives have been much appreciated,” says Jasmine Gardner.

Demands a great deal of computing power

A great deal of computing power is demanded in order to perform calculations based in excess of 10 million people, with so much data pertaining to each and every one of them. The researchers have therefore be granted access to several supercomputers, including at UPPMAX in Uppsala, HPC2N in Umeå, NSC in Linköping and PDC in Stockholm.

Lynn Kamerlin, professor at the Department
of chemistry. Photo: Mikael Wallerstedt

The model uses data concerning the spread of the virus in China and the researchers are hopeful that they will soon be able to present figures on how the coronavirus is likely to spread in Sweden, depending on the measures put in place. They are preparing calculations based on five different scenarios, from no measures whatsoever up to closing schools and enforcing a large degree of isolation.

The research is being led by Uppsala researchers Professor Lynn Kamerlin of the Department of Chemistry and Wallenberg Scholar Peter Kasson, senior lecturer at the Department of Cell and Molecular Biology, together with Nele Brusselaers, associate professor of clinical epidemiology at Karolinska Institutet. Dozens of researchers at Uppsala University have been involved in the project as well as researchers at KI, KTH and the University of Antwerp.

The results can provide guidance

Professor Tove Fall of Uppsala University is one of the epidemiologists contributing to the model.

What benefits can policy-makers and the public gain from the results of these calculations?
“The results can provide guidance in understanding how various measures are likely to affect infection and demands on healthcare. They can also illustrate to the public the importance of following recommendations regarding social distancing.”

How well will the results of the model reflect reality?
“Right now we are missing some important data, such as how many people are currently infected in Sweden and how the recommendations are affecting people’s contacts with one another. In this regard we can make assumptions based on the available data. Another important parameter that remains uncertain is the extent to which those without symptoms are infectious. Naturally, the more we learn about the virus and the spread of infection from ongoing studies, the better the forecast we will be able to make. The forecasts will therefore be able to provide guidance regarding the effects of various measures and an estimate of healthcare requirements, but there are major uncertainties.”

Last modified: 2021-02-14