Researchers at UC San Diego School of Medicine have found that several of the most dangerous viruses to strike human populations in recent decades, including SARS-CoV-2, did so without first evolving special genetic adaptations in animal hosts. The finding, published in Cell, challenges a widely held assumption in virology: that animal viruses need a period of evolutionary “warm-up” before they can efficiently infect people. If correct, the result means the threat window for the next pandemic may be far shorter than many scientists have assumed, and current surveillance strategies built around detecting early adaptation signals in animals could be looking for warnings that never arrive.
What the Researchers Actually Tested
The study, titled “Dynamics of natural selection preceding human viral epidemics and pandemics,” used phylogenetic and genome-wide selection tools, including a method called RELAX, to measure whether natural selection intensified on the evolutionary branch leading directly to human outbreaks. In plain terms, the team asked whether viruses showed signs of being “shaped” for human infection right before they jumped species. They applied this framework across a broad panel of outbreak viruses responsible for major epidemics, drawing on genomic data from SARS-CoV-2, the 2009 H1N1 swine-origin influenza pandemic, the 2014 Ebola outbreak in Guinea that ignited the West Africa epidemic, the 2022 multi-country mpox outbreak, and the Marburg hemorrhagic fever outbreak in Angola.
To test the longstanding assumption that adaptation in animals is a prerequisite for human spread, the authors compared selection pressures on viral lineages immediately before spillover with those on background lineages circulating in reservoir hosts. According to a technical summary of the work, they specifically evaluated whether branches leading into human epidemics showed the intensified positive selection that would indicate a period of fine-tuning for human infection.
The answer, across most of these cases, was no. The study’s own highlights state bluntly that adaptation before emergence is not a necessary precursor to outbreaks of novel zoonotic viruses. That sentence carries significant weight. It means the viruses behind some of the deadliest epidemics of the 21st century were, in evolutionary terms, already capable of infecting humans the moment they crossed over from their animal reservoirs, rather than gradually acquiring that capacity in a detectable way.
Why the Prevailing View Assumed Otherwise
For decades, the dominant model of pandemic emergence worked roughly like this: a virus circulating in bats, birds, pigs, or primates would accumulate mutations that gradually improved its ability to bind human cell receptors, replicate in human tissues, or evade human immune defenses. Surveillance programs were designed accordingly, scanning animal virus genomes for molecular red flags, specific mutations in receptor-binding domains or polymerase genes, that might signal a virus was “getting closer” to making the jump.
This logic shaped how public health agencies allocated resources and how researchers prioritized which animal viruses to watch. The 2009 H1N1 pandemic, for instance, was later traced to a virus with a complex history in swine before it emerged in humans, a narrative that seemed to fit the gradual-adaptation model. But the UC San Diego team’s analysis suggests that even in that case, the selection pressures acting on the virus before spillover did not look meaningfully different from what was happening in the broader swine viral population. The virus did not need to “practice” on pigs in a way that left a clear evolutionary signature.
The assumption of a detectable warm-up period also dovetailed with how scientists thought about receptor usage and host range. Many classic examples of cross-species jumps, from avian influenza adapting to mammals to simian immunodeficiency viruses giving rise to HIV, appeared to involve stepwise changes in key viral proteins. That history encouraged the belief that, with enough genomic surveillance, researchers could spot dangerous viruses as they were evolving toward humans.
The Outbreaks That Showed No Warm-Up
The study’s cross-virus comparison is what gives the finding its force. The 2014 Ebola emergence in Guinea, first characterized in a preliminary genetic report, killed thousands across West Africa. Genomic analysis of the Marburg virus from the Angola outbreak, documented through time-series sequencing across locations, provided another data point. The 2022 mpox outbreak, whose phylogenomic profile was mapped in a detailed analysis of viral lineages, rounded out the comparison.
In none of these cases did the UC San Diego team detect the kind of intensified positive selection on pre-spillover branches that the standard model would predict. The viruses appear to have jumped abruptly, not after a detectable ramp-up in adaptive evolution. This pattern held across RNA viruses (like SARS-CoV-2 and Ebola) and DNA viruses (like mpox), and across viruses with very different reservoir hosts and transmission routes. A news release from the UC San Diego research team underscores that this lack of pre-adaptation was consistent across the major pandemic and epidemic viruses they examined.
Independent coverage of the Cell paper has echoed that conclusion. Reporting from a public health news service noted that, contrary to prevailing belief, an evolutionary warm-up in animals was not needed for recent pandemic viruses, including SARS-CoV-2, to spread efficiently among people. That convergence between the primary study and external assessments strengthens the case that the observed pattern is not an artifact of a particular analytical pipeline.
What This Means for Pandemic Preparedness
If viruses do not reliably telegraph their intentions through pre-spillover adaptation, the practical consequences for outbreak prevention are serious. Genomic surveillance programs that rely on spotting “pre-adapted” animal viruses may be searching for a signal that does not consistently exist. That does not make genomic surveillance useless, but it does suggest that other indicators, particularly ecological and behavioral ones like changes in human-animal contact patterns, wildlife trade routes, and habitat disruption, may deserve equal or greater weight in early warning systems.
The study also highlights the importance of building and maintaining large, well-annotated viral genome databases. Tools such as RELAX and related methods depend on dense sampling across hosts and time, something enabled by repositories like the National Center for Biotechnology Information and other international sequence archives. Better baseline data on how viruses evolve in their natural reservoirs will make it easier to recognize truly unusual patterns when they do occur, even if those patterns are not a universal prelude to spillover.
The finding adds a new dimension to the ongoing debate over the origins of SARS-CoV-2. The Cell study’s highlights specifically reference selection signatures on SARS-CoV-2, and a recent news feature in Nature situates the evolutionary evidence as one line among several, including epidemiology, wildlife trade analysis, and laboratory records. The absence of detectable pre-adaptation does not by itself resolve the origins question, but it removes one argument sometimes used to claim the virus must have been artificially manipulated: the idea that a “perfectly adapted” virus could not have arisen naturally without a stepwise trail of mutations.
Gaps and Cautions
Despite its broad scope, the UC San Diego study has important limitations. Evolutionary analyses are only as powerful as the data they draw on, and for many animal reservoirs, genome sampling remains sparse or biased toward certain regions and time periods. It is possible that some adaptive events occurred along unsampled branches or in understudied host populations, leaving no trace in the available sequences. The absence of evidence for pre-spillover adaptation is not the same as proof that such adaptation never happened.
Selection tests like RELAX also detect only certain kinds of evolutionary change. They are well suited to picking up shifts in the overall intensity of positive or purifying selection across many sites, but they may miss more subtle forms of adaptation, such as changes in regulatory regions, epistasis among mutations, or host-range expansions driven by a small number of critical amino acid substitutions. A virus could, in principle, become better suited to humans through a handful of key mutations that do not produce a strong genome-wide signal.
Another caveat is that the study focuses on the period immediately preceding recognized human outbreaks. For pathogens like influenza, which circulate in multiple animal hosts and undergo frequent reassortment, important evolutionary steps may occur long before the lineages that are eventually sampled in humans. The lack of intensified selection right before spillover does not rule out a more extended history of adaptation in intermediate hosts or ecological niches that remain poorly characterized.
Still, the overall message is difficult to ignore. When multiple, unrelated viruses with different genomes, reservoirs, and transmission modes all show no clear sign of pre-spillover fine-tuning, it becomes harder to rely on the comforting idea that pandemics announce themselves in advance through a detectable evolutionary drumbeat. Instead, the emerging picture is one in which many animal viruses already possess the intrinsic capacity to infect humans, and the decisive factor is often opportunity rather than gradual genetic preparation.
For policymakers and scientists alike, that shift in perspective argues for a more precautionary approach to human–animal interfaces. Reducing risky contact, improving biosafety in settings where people handle wild or farmed animals, and rapidly scaling up response once a novel human cluster is detected may offer more realistic protection than hoping that genomic surveillance alone will provide an early, unambiguous warning. As the UC San Diego researchers and outside commentators emphasize, understanding viral evolution remains essential, but it must be integrated with ecology, behavior, and public health infrastructure if it is to help avert the next pandemic.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
