30 June 2020

Coronavirus family tree can reveal new transmission chains

Mathematics

Researchers at the University of Copenhagen have used unconventional methods to identify coronavirus transmission chains. An advanced family tree of the virus’ development can provide an exceptionally accurate overview of how the virus was carried into Denmark and its subsequent spread domestically. The researchers hope that their tool will prove useful for contact tracing as new outbreaks flare up locally.

Coronavirus - modelfoto
More than 700 publicly available Danish and 30,000 international sequences of the coronavirus genome underlie the analysis. Photo: Getty

Imagine a large international family with many members and intricate branches. To better understand the family, one would construct a family tree to trace the heritage of each member back to their ancestors.

Along these lines, researchers at the University of Copenhagen’s Department of Mathematical Sciences have constructed a family tree of the Danish coronavirus in a preliminary study, now available as preprint. The researchers assembled the tree using a computer program that allows them to decode information about transmission chains using a customized visualization.

Unsurprisingly, the family tree confirms that a large proportion of the coronavirus sequences observed in Denmark originated in the Austrian ski resort of Ischgl. More surprisingly, the family tree demonstrates that some of the transmission to Denmark originated elsewhere, and that Danish infections probably spread to Iceland and Sweden, among other countries. 

"Based on more than 700 publicly available Danish and 30,000 international sequences of the coronavirus genome, we have traced Danish transmission chains back to viral ancestors in Austria, Italy — and even the Netherlands. We have also been able to trace transmission routes around Denmark and on to other countries," says Professor Matthias Christandl, one of the study’s authors.

Ability to verify contact tracing

By means of the advanced genome analysis of published coronavirus sequences, the researchers' method can identify whether transmission chains are related vis-à-vis contact tracing, or whether there are mutations in the family tree that contact tracing is unable to account for.

"Through traditional contact tracing, one might conclude that person A infected person B. However, if upon closer inspection, one finds various mutations of viral genomes, they will know this not to be the case. To halt transmission, they will need to keep hunting for the person who infected person B. As such, we hope that our tool can be used to complement traditional contact tracing," says Professor Matthias Christandl.

At the ready to investigate new outbreaks

Coronavirus consists of a strand of 30,000 nucleotides — the building blocks of its RNA. As the virus spreads from person to person, changes or mutations occur to the strand of nucleotides.

By identifying the changes, researchers can see which of the mutations have a common heritage. This allows them to determine the origin of individual mutations, based upon Danish and international coronavirus samples.

"There has never been so much genetic information about a pandemic available online. It is nearly a revolution in itself, one that has made it possible for us to develop this tool," explains Professor Christandl, adding:

"While our computer programs are publicly available on github, studying an outbreak requires manual work and expertise. But we are eager to work with anyone interested in studying an outbreak."

Facts:

  • More than 700 publicly available Danish and 30,000 international sequences of the coronavirus genome underlie the analysis.
  • The researchers developed a computer program that makes it easier to visualize changes in the genome and provide an overview of the transmission chain.
  • For smaller data sets, the researchers' method can analyse a corona genome within hours.
  • Research is being conducted at the Centre of Excellence for the Mathematics of Quantum Theory (QMATH), which is funded by the VILLUM FOUNDATION
  • The study is a 'working paper', and yet to be peer-reviewed.