In a new study co-authored by Temple biologist Sergei Pond, researchers found that SARS-CoV-2, the virus responsible for the COVID-19 pandemic, showed no signs of unnatural evolution. They also found that many viruses can spread to humans with less adaptation than was previously thought.
Sergei Pond has spent two decades researching viral evolution. His lab specializes in developing statistical models and computational tools for analyzing large data sets related to pathogens.
Photo by Ryan S. Brandenberg
In the aftermath of the COVID-19 outbreak, there was much speculation over how the virus called SARS-CoV-2 came to infect humans. Some suggested it evolved just as viruses normally do—originating in animal populations and then jumping to humans after a period of mutating and adapting. Others believed that humans manipulated the virus in a lab.
But new research co-authored by Sergei Pond, professor of biology in Temple’s College of Science and Technology, found that SARS-CoV-2 evolved normally and showed no signs of human manipulation. At least, not the kind of human manipulation that scientists have seen before.
“There was nothing remarkable about SARS-CoV-2’s evolution from what we could tell,” said Pond, who has spent two decades researching viral evolution.
Published in the journal Cell on March 6, 2026, “Dynamics of natural selection preceding human viral epidemics and pandemics,” offers a new understanding of how viruses make the jump from animals to humans.
Pond and the article’s co-authors analyzed seven viruses, including SARS-CoV-2, Ebola and influenza A, looking for clues about how they evolved.
Scientists have long believed that viruses require a significant period of adaptation in animal hosts, sometimes multiple animal hosts, before making the jump to humans. But this latest research suggests that many viruses can spread to humans with less adaptation than was previously thought, and the finding may help scientists better identify and prepare for viral threats.
“SARS-CoV-2 demonstrated pretty dramatically that whatever mechanisms we had in place for predicting and managing pandemics were inadequate,” Pond said. “Our research suggests we don’t necessarily have to wait for a virus to do something extraordinary for it to be a threat.”
Pond describes his work as evolutionary forensics. Scientists can’t recreate the crime, so to speak, of a virus’ original transmission to humans. By the time humans discover a virus spreading between them, it is usually years, sometimes decades, since the original transmission occurred. Instead, scientists must look for clues in the evidence that is left behind.
Pond’s lab specializes in this kind of work by collecting genetic sequences from infected humans and animals and analyzing them using computational methods.
Genetic sequences are loaded with clues about how a virus mutates and evolves. Pond uses genetic sequences gathered from infected animals to establish what could be considered the natural way that a virus evolves. Comparing those sequences with sequences collected from infected humans allows Pond to determine whether a virus evolved naturally in humans.
In the case of SARS-CoV-2, Pond found that the virus evolved in humans similarly to how it evolved in bats.
Pond also compared the seven viruses’ genetic sequences with those of viruses known to have been altered through human manipulation.
For example, in 2011, a group of researchers from the Netherlands conducted an experiment involving H5N1, a deadly bird flu strain that hadn’t yet infected humans. The researchers’ goal was simple: They wanted to evolve the virus in a lab to become airborne transmissible and gain the ability to infect humans. They were successful.
“Our model has sequences from that particular experiment, where we know the virus was driven to change by humans,” Pond said.
His model also has the sequence for a strain of the Spanish flu that accidentally leaked from a Siberian lab in the 1970s, as well as sequences from a handful of other scenarios where human manipulation occurred, like sequences involving vaccine trials.
If SARS-CoV-2’s evolution involved human manipulation, like in the case of H5N1 or the 1970s leak of the Spanish flu, or even in the case of a vaccine trial, Pond’s computational model would have caught it.
This doesn’t rule out human manipulation entirely, and Pond explained that his model can only compare against a small selection of instances where human manipulation was involved. But it does add to the evidence that SARS-CoV-2 evolved as one would expect.
The researchers did find that many viruses are well-prepared to jump from animals to humans with minimal adaptation and oftentimes without the need for an intermediary animal host, offering a new understanding of how viruses evolve.
“Every single significant pathogen that you’ve heard of, like HIV, coronaviruses, influenza or measles—they all originated in animals, but it is not necessarily obvious how it makes its way into a different host,” Pond said. “We found that the vast majority of these viruses don’t seem to need an intermediate host. A lot of them are quite good at just doing direct to human transmission, or transmission with minimal adaptation.”
Humans still have much to learn about viruses and how they evolve, but this research might help scientists better predict and prepare for the emergence of new diseases.
“This is just one brick in our foundation of tools that we can use to understand and classify viruses by how dangerous they are,” Pond said. “Our research shows that we can focus on studying viruses that already possess the keys to human infection to help us better understand which animal viruses pose a real threat.”