On Jan. 12, 2020, Chinese investigators combing a market for clues about the outbreak of a mysterious new illness in the city of Wuhan swabbed a cart. It was the kind typically used for transporting animal cages, and it came back positive for the coronavirus.

Three years later, a team of international experts has sifted through the genetic contents of that swab, which were quietly uploaded to an international database and made public only this year. In a report released on Monday night, the scientists described in detail for the first time evidence from the swab that they say strengthens the case that illegally traded wild animals ignited the coronavirus pandemic.

Chinese researchers who had originally uploaded the raw data had it removed from the database after they were contacted by the international team. Now administrators of the database itself have cut off access to the international scientists for what they said were rules violations, raising questions about the database’s own role in the tug of war over access to data that could shed light on the origins of a virus that has killed seven million people.

Along with genetic signatures of the coronavirus, the swab from the cart contained more than 4,500 lengthy fragments of genetic material from raccoon dogs, the report said. It had none from humans. Some Covid-positive swabs taken from other objects and surfaces at the market, the report said, also had more genetic material from animals than from humans.

Finding genetic footprints from animals in the same place as genetic material from the virus does not prove that the animals themselves were infected. But some scientists who reviewed the report said that the dominance of genetic material from animals — and especially raccoon dogs — suggested that species known to be able to spread the coronavirus were indeed carrying infections at the market in late 2019.

That scenario, they said, was consistent with the virus spilling into humans from market animals and touching off the pandemic, a set of circumstances similar to the one that gave rise to the first SARS outbreak in China two decades earlier.

“You look at them and say those are probably infected animals,” Theodora Hatziioannou, a virologist at The Rockefeller University in New York who was not involved in the research, said of the latest findings. “If it was a human shedding the virus, one would expect to find human DNA there, too.”

The swabs could yet hold more clues about where the virus in the samples had come from. The report said, for instance, that there was evidence of particular genes that could suggest the material had come from a raccoon dog’s upper respiratory tract.

Even if an animal had been infected, however, it would not be clear that it had spread the virus to people. Someone infected with the virus could have gotten a market animal sick. And only by swabbing animals directly could scientists prove whether they had been carrying the virus, a step that was precluded by the market being cleared of animals soon after the outbreak began.

The report has been the subject of intense speculation since the international experts presented their findings to the World Health Organization last week and then raced to compile their analyses. At the same time, the findings set off a battle for access to the genetic sequences at their heart.

Chinese scientists had initially uploaded the raw sequences to a global database some time after publishing a study describing them last year. But once the international experts discovered the data in early March and alerted Chinese researchers to what they had found, the data was taken offline.

Last week, the W.H.O. rebuked China for hiding such crucial information from the rest of the world for three years. Now the nonprofit organization based in Munich that runs the database, called GISAID, has come under scrutiny for its role in controlling access to the data.

In the new report, the international team of scientists said that GISAID had “deviated from its stated mission” in allowing the Chinese researchers to withhold the data for so long.

The database administrators responded to the report on Tuesday by cutting off the team members’ access to their online accounts and saying that they had violated its rules by getting out ahead of the Chinese scientists and posting their own analysis. The scientists said that they hewed to GISAID’s database-access agreement in downloading and studying the sequences, and noted that they had made multiple offers to work with the Chinese scientists.

“The ramifications of cutting off access to this group of authors are huge,” said Michael Worobey, an evolutionary biologist at the University of Arizona and co-author of the new report, noting that GISAID also jeopardized work by team members related to coronavirus variants and flu preparedness. “They’re making false accusations.”

The international team homed in on raccoon dogs — fluffy mammals related to foxes and sold for meat and fur — because of how much of the animals’ genetic material was found in the key swab from the cart and because they are known to spread the virus. They said their findings were consistent with that animal harboring the virus, which originated in bats, and passing it to humans at the market.

“This isn’t an infected animal,” said Joel Wertheim, a virologist at the University of California, San Diego and a co-author of the report, referring to the new genetic data. “But this is the closest you can get without having the animal in front of you.”

The report, though, also offered the most concrete evidence to date of other animals susceptible to the virus being sold at the market, noted Kristian Andersen, a virologist at the Scripps Research Institute in La Jolla, Calif., and a co-author of the report. Genetic material from those animals — like the masked palm civet, a small Asian mammal that was implicated in the SARS outbreak two decades ago — was also found in swabs that were positive for the coronavirus.

“It’s literally Disney Land for zoonotic transfer,” said Joseph DeRisi, a professor of biochemistry at the University of California, San Francisco, and co-president of the Chan Zuckerberg Biohub, referring to the range of animals documented in the report.

A number of other swabs at the market found large quantities of human genetic material — an indication, the report said, of certain virus samples likely being shed by infected people. Many of the earliest known Covid patients worked or shopped at the market.

Still other positive swabs, the report said, were dominated by genetic material from animals that are not believed to be susceptible to the virus. A sample taken from a fish packaging surface, for example, contained a lot of fish genetic material. That virus was likely to have been deposited by a person, scientists said, illustrating that substantial amounts of animal genetic material did not necessarily mean that animals had produced the virus there.

Citing those findings, some scientists said that the kinds of swabs analyzed in the report simply could not offer conclusive proof of an infected animal.

“The report does contain useful information,” Sergei Pond, a virologist at Temple University in Philadelphia, said. But, he added, “Does it tell you anything about which animal was infected? It really doesn’t.”

Dr. David Relman, a microbiologist at Stanford, said that it was difficult to assess the findings without more details about how the Chinese investigators had collected and analyzed their swabs. The Chinese paper last year, he noted, described using a kit to filter out human genetic material and make the tests more sensitive to the virus.

But a number of swabs from the Chinese researchers described in the international team’s report contained substantial amounts of human genetic material, suggesting that those filtering methods had not depleted the samples of human material. Dr. DeRisi, who specializes in the type of analysis described in the report, said that such kits were generally ineffective at removing the genetic signatures of people.

Frederic Bushman, a microbiologist at the University of Pennsylvania who also specializes in sequencing techniques, agreed that the report’s methods were sound.

“I think the simplest explanation is that it’s an infected raccoon dog,” he said. “I don’t think it’s absolute proof.”