A man hangs an Ebola awareness banner in the Kigonze camp in Bunia, in the eastern Democratic Republic of the Congo, on May 28, 2026. An Ebola hemorrhagic fever outbreak is spreading rapidly in the east of the Democratic Republic of Congo, one of the poorest countries in the world. Nearly one million people have been displaced by conflict in Ituri province, the epicenter of the outbreak and one of the most troubled areas of the DRC, where armed groups, rebels linked to the Islamic State, and community militias regularly carry out massacres. (Photo by Glody MURHABAZI / AFP via Getty Images)
AFP via Getty Images
The first genomes from the Ebola outbreak in eastern Congo have been published in Nature Medicine. It was already known that the virus responsible for the outbreak is a variant of a viral species known as Bundibugyo. The new analysis shows that it is only distantly related to the Bundibugyo viruses that have caused outbreaks in humans previously, and therefore almost certainly the result of recent spillover from an animal reservoir rather than recurrence of an earlier, human-adapted strain.
Genome sequences were produced by the Institut National de Recherche Biomédicale in Kinshasa and Uganda’s Central Public Health Laboratories in Kampala from 22 specimens. Of these, 21 were clean enough to use. These sequences form a single cluster distinct from the 2007 Bundibugyo epidemic in western Uganda and the 2012 epidemic in Isiro.
The Genomes Show Months Of Human Spread
The genomes also show that the outbreak has been spreading from person to person for several months. Twenty-five mutations separate the 2026 sequences from one another, and nine bear the mark of ADAR, a human enzyme that attacks the double-stranded RNA the virus forms while copying itself. ADAR rewrites adenosine bases as inosine, which the virus then copies as guanosine, so each edit becomes a permanent A-to-G change in the genomes of its descendants. Because the enzyme acts only on replicating viruses, a cluster of these edits is a record of sustained copying in human hosts. On one branch of the tree, four fall within a stretch of 27 nucleotides.
Dating The Jump Will Take Older Genomes
Dating the origin of the outbreak is a harder problem. Only twenty-two genomes, the authors write, “precluded a robust estimation of the time to the most recent common ancestor.” Every sequenced specimen was collected on May 2 or later, months into transmission. Scientists calibrate a molecular clock from the mutations a virus accumulates across an epidemic’s whole history, and this sample comes from the middle of one.
Other evidence points to a start months earlier. A field reconstruction in Mongbwalu, reported by Kai Kupferschmidt in Science, counted more than 500 cases before Congo declared the outbreak on May 15 and put the earliest deaths in January. This is the reason I argued last week that every published estimate of this outbreak’s size begins too late.
WHO has come partway. Its regional director for Africa, Mohamed Yakub Janabi, told reporters on August 10 that sequencing puts the start in February, and explained that early cases had been mistaken for malaria and typhoid. The details of that sequencing effort have not been published, and it may rest on more genomes than the 22 in this paper, which sit late in an epidemic now past 4,000 confirmed cases.