However, the deleted data has not been lost. By searching in the Google and Amazon clouds used for storage by the SRA, the professor found some of the removed genomic sequences.
The deleted data could now help fill in a missing link for scientists puzzled by the evolution of Sars-CoV-2 in people.
Regardless of how it ultimately reached humans, experts agree that Sars-CoV-2 originated in bats. But until now, the earliest known samples from Wuhan - including those taken from the Huanan Seafood Market, originally suspected as the potential source of the outbreak - were further away from these viruses, genetically, than other Sars-CoV-2 samples obtained in other parts of China and even other parts of the world.
The deleted sequences fill in that gap, as they are closer to the bat coronaviruses, and so could represent an earlier step in the evolution of the virus. They add further evidence that the virus was already circulating in humans in Wuhan as early as autumn 2019.
Experts said this was a major development.
Professor Rasmus Neilsen, a genomics expert at the University of California, Berkeley, tweeted: "These are the most important data that we have received regarding the origins of Covid-19 for more than a year."
However, there are also other important implications, as Professor Bloom explained in a lengthy Twitter thread.
"First, [the] fact this dataset was deleted should make us sceptical that all other relevant early Wuhan sequences have been shared," he said, pointing out that China ordered many labs to destroy early samples.
Chinese scientists were also under orders to get their publications on coronavirus checked and centrally approved - an order that came one day before the publication of the pre-print linked to the deleted samples.