杏吧原创

TWO months ago, 2005 looked like going down as the year when therapeutic cloning took a major step closer to reality. Now it looks more likely to be remembered for one of the biggest scientific scandals of recent times, a scandal that has left cloning research in disarray.

In May, Science published a landmark paper by a team led by Woo Suk Hwang of Seoul National University, South Korea, which described the creation of 11 embryonic stem cell (ESC) lines, each cloned from a different person. The paper laid the foundations for the idea of treating sick people with cells that are genetically identical to their own.

Then, last month, Hwang admitted that some of the eggs for an earlier study came from women on his team 鈥� an ethical lapse he had previously denied. Since then, trouble has snowballed. Hwang admitted 鈥渋rretrievable mistakes鈥� in photographs in the Science paper. In one case an image was enlarged, and in another it was distorted, making it look like a different cell. Last week, Hwang and his co-authors called for the paper to be retracted because data in it 鈥渃ould not be trusted鈥�. Infighting broke out between the authors, and several other papers have now been questioned or retracted (see 鈥淐loning crisis goes from bad to worse鈥�).

This shambles is inflicting enormous damage. The retracted paper reported a relatively high success rate for creating cloned ESCs. That claim is now in doubt. Questions are also being asked about a 2004 paper by Hwang鈥檚 group which described the first cloned human ESC. If that is retracted, all published evidence of success in human therapeutic cloning will be wiped out. Opponents of ESC research, who make a point of stressing any trace of misrepresentation or hype, will seize on this new ammunition.

Worst of all is the damage done to science itself. Science runs on trust. Governments give researchers money on the understanding that they will use it fairly, and honestly report their results. Peer reviewers assume that what they are judging is a fair account of what happened; they are not yet charged with policing dubious data. Without trust, the whole scientific project will collapse.

Some will argue that this is an overreaction, and that such grim predictions should not be based on an isolated case. But it is not an isolated case. Only three years ago, Hendrik Sch枚n was sacked for fabricating evidence in 17 physics papers 鈥� some in Science and Nature. This year, MIT sacked the high-flying biologist Luk Van Parijs after he admitted fabricating data. Doubts remain over at least three of the 40 papers he published in the past eight years.

Aside from these high-profile cases, a certain amount of low-level bad behaviour seems to be endemic. Earlier this year, Brian Martinson and colleagues from Minneapolis published a survey of more than 3000 researchers funded by the US National Institutes of Health (Nature, vol 435, p 727). They found that 10 per cent or more admitted to withholding details of methodology or results, inappropriately assigning authorship credit, and dropping observations or data points based on a gut feeling that they were inaccurate.

鈥�2005 looks likely to be remembered for one of the biggest scientific scandals of recent times鈥�

So what is to be done? At the very least, research ethics need a higher priority both in the education of young scientists and within research institutions. Also, Martinson found that 1 in 8 of his respondents admitted to overlooking other researchers鈥� flawed data or questionable interpretations of data. It is time for scientists to become more active in challenging such instances 鈥� as a group of young Korean researchers has done in the Hwang case.

Peer review also needs tightening up. To have spotted the mistakes in Hwang鈥檚 retracted paper may have required an unfeasibly high level of scrutiny. But journal editors can demand more supporting evidence, especially for papers that make extraordinary claims. It appears astonishing, for instance, that Nature accepted Hwang鈥檚 paper on Snuppy, the first cloned dog, without seeing original DNA data.

Should researchers who transgress be penalised? The idea of a deterrent runs counter to the collaborative spirit that still flourishes in the scientific world. But have we reached a point where serious discouragement against bad behaviour is essential? The world鈥檚 science academies need to address these issues. If they do not, it is a safe bet that their paymasters will. And nobody wants 2006 to go down as the year when governments and the public lost faith in science.