Most pandemic viruses jumped to humans without evolving for it first

There is a persistent assumption in virology that pandemic viruses must undergo a period of evolutionary fine-tuning in an animal host before they can infect and spread among humans. A new study published in Cell challenges that assumption with data from six viral families and more than two decades of genomic records. The finding is both reassuring and unsettling: most pandemic viruses did not need special adaptations to jump to humans. They were already capable when they made the leap.

Testing the adaptation hypothesis across six virus families

Joel Wertheim, a professor of medicine at UC San Diego, and his team analyzed viral genomes from outbreaks caused by influenza A, Ebola, Marburg, mpox, SARS-CoV, and SARS-CoV-2. For each virus, they focused on the evolutionary period immediately before human outbreaks began, the window where any substantial pre-spillover adaptation should leave a detectable genetic imprint.

The method relies on measuring the ratio of different types of mutations across entire viral genomes. Mutations that change protein function (nonsynonymous) versus those that do not (synonymous) provide a signal of whether natural selection is acting more strongly than expected. If a virus were adapting to a new host before spillover, the selection pressure should intensify during that transition period.

Across all the viruses analyzed, the researchers found a strikingly consistent pattern: selection pressures before zoonotic emergence were indistinguishable from those during routine circulation in animal reservoirs. No evolutionary ramp-up. No fine-tuning. The viruses that caused pandemics looked, genetically speaking, like any other version of themselves circulating in their animal hosts.

What this means for SARS-CoV-2 origins

The finding has direct relevance to the ongoing debate about COVID-19 origins. Wertheim's team found no evidence that SARS-CoV-2 was shaped by selection in a laboratory or through prolonged evolution in an intermediate host before its emergence. That absence of a pre-adaptation signal is exactly what the researchers observed for every other natural zoonotic event in their dataset.

This does not prove a natural origin. But it demonstrates that the genomic evidence is fully consistent with natural spillover and inconsistent with the specific claim that the virus was extensively manipulated or passaged in laboratory conditions.

The 1977 flu: a different kind of outlier

One virus in the dataset did not fit the pattern. The H1N1 influenza A virus that re-emerged in 1977 showed two anomalies. First, it was genetically almost identical to H1N1 strains from the 1950s, with far less divergence than 20 years of natural evolution should have produced. Second, it carried a selection signature consistent with viruses that had been passaged through cell cultures or laboratory animals.

This finding adds molecular evidence to a long-standing suspicion: the 1977 flu pandemic was likely sparked by a laboratory strain, possibly in the context of a failed vaccine trial. The virus had characteristics that real-world evolution does not produce but that laboratory handling does.

Establishing a genomic baseline for outbreak forensics

The practical value of the study extends beyond settling historical questions. By defining what "normal" zoonotic emergence looks like at the genomic level, the framework provides a reference point for interpreting future outbreaks.

If a new virus emerged with the selection signature of laboratory passage, the method could flag it. If it looked like every other natural spillover event, the method would indicate that too. The researchers validated their approach against known laboratory-adapted viruses and live-attenuated vaccines, which produced clear and reproducible signatures distinct from natural transmission.

Exposure, not adaptation, is the bottleneck

The broader implication reframes how we think about pandemic risk. If most pandemic viruses do not require rare, finely tuned adaptations to jump species, then the bottleneck is not viral evolution. It is human exposure. The more contact humans have with diverse animal viruses, through wildlife trade, habitat encroachment, farming practices, or other activities, the more opportunities exist for spillover.

This shifts the emphasis for pandemic preparedness away from trying to predict which mutations might make a virus dangerous and toward surveillance, prevention, and reducing the frequency of human-animal viral contact.

The study does not claim that laboratory accidents never happen. It claims that if a virus had been extensively passaged in a lab, the evidence would be visible in its genome. For nearly every pandemic virus studied, that evidence was absent.