The NPOF has now taken a decision on the matter handed over by the Karolinska Institutet. The council convicts four researchers of misconduct by falsifying figures that do not show what they claim to show. These are images of so-called Western blot analyses, which are used to detect and identify proteins.

According to the committee, one of the researchers acted intentionally, while the other three were grossly negligent.

All deny their involvement in fraud. They claim, among other things, that it is all very much a matter of the notifier having second opinions as to which techniques are most appropriate and how best to interpret the data. One of those convicted, an award-winning diabetes researcher, admits that some mistakes were made, but he does not believe that affected the conclusions reached.

The Council also obtained expert opinion. The decision stipulates the following:

» The expert believes that there is no basis for all suspicions of forgery and fabrication. It concludes that there are some errors in the articles, but they are small and had no effect on the results, or their effect was very small.

But the committee is therefore making a different assessment. According to the Council, the two numbers found in one of the articles in question constitute forgery.

According to the committee, it is not plausible that Western blot panels that should have appeared in one of the figures were mistakenly replaced with the results of previous measurements. The conclusion instead is that the first author must have intentionally changed the numbers.

In another article, eight numbers were forged, according to the committee. The first author admits that the spots were confused in two shapes and that the protein band is the wrong orientation in another shape. Otherwise, the authors question the whole matter.