The journal Science has asked the authors of a research paper, which linked chronic fatigue syndrome to a virus, to withdraw their findings.
It has also published an editorial expressing concern that the validity of the study was "seriously in question".
The authors said they were "extremely disappointed" and that the editorial was "premature".
An expert in the UK said any link with chronic fatigue syndrome, or ME, was a myth and the decision was inevitable.
In 2009, a study at the Whittemore Peterson Institute was published in Science which showed that DNA from a mouse virus, XMRV, was present in 67% of patients with chronic fatigue syndrome, but only 4% of the general population.
Science's editor in chief, Bruce Alberts, said at least 10 studies had since failed to reproduce those results, including two studies published at the same time as his editorial.
One concluded that the mostly likely explanation for the 2009 finding was that laboratory samples were contaminated with XMRV.
The other looked at 61 patients with chronic fatigue syndrome who took part in the original study, but it found no trace of XMRV.
As a result, Science asked for the authors of the 2009 research paper to voluntarily retract their findings. They declined.
Annette Whittemore, President of the Whittemore Peterson Institute, said: "We are extremely disappointed that the editor of Science has published an 'editorial expression of concern'".
She said that other studies had not used the same experiments as the original study and that: "The authors of the Lombardi study believe that it is premature to conclude that the negative studies are accurate or change the conclusions of the original studies.
"Much of the work on this new retrovirus has yet to be performed, and we look forward to new studies which will support the results and findings described by these accomplished scientists."
Dr Jonathan Stoye, virologist at the Medical Research Council National Institute of Medical Research, said: "It comes as no great surprise, in fact it was inevitable since a series of studies failed to reproduce the original results."
"It should be made as definitive as possible that XMRV is not linked to chronic fatigue syndrome. It is a myth."
He said the implication was that the samples were contaminated, however this had not been definitively proven.
He added: "Science could have gone one step further and withdrawn it off its own bat. In football this is somewhere between a red and a yellow card."