Bookmark and Share
[Sine Larsen]

Most of us know that letters that deal with unpleasant matters tend to be more difficult to write, and the present letter is such an example. 2010 will be the remembered as the year when crystallography lost its scientific innocence. The thorough checking of the publications in the IUCr journals had previously been considered as a protection against scientific fraud, but the editorials published in the January issues of Sections D and E of Acta Crystallographica revealed that a number of structures published in the IUCr journals were based on falsified data. The revelation of this extent of scientific fraud in crystallography was as big a shock to me as it was to all crystallographers. On the international scene the retraction of articles received much attention and was the topic of editorials in several well known international journals that also publish crystal structures.

After a couple of months digesting these unpleasant facts, it is now time to consider how similar situations can be avoided in the future and to see if this bad experience can be turned into something constructive. There are examples of fraud in most other scientific disciplines, but I would like to emphasize that those in Acta E were found by Ton Spek during continuing development of his software for checking crystal structures (PLATON). Very few scientific areas are blessed with scientists like Ton Spek who has devoted his scientific career to the development of software for the analysis and checking of other people's results. I was therefore very happy to learn that he will receive the Trueblood award at the upcoming ACA meeting in Chicago in July. With the new software, and its continuing development, and the help of our vigilant Co-editors I feel confident that we are doing all we can to minimize the risk of fraudulent papers appearing in IUCr journals. But what about the crystal structures published in other journals, many of which do not require the deposition of structure factors that have provided the basis for the determination of the structure? The analysis of structure factors is an essential prerequisite for the detection of problem structures. Therefore it should be in the interests of all journals that publish crystal structures that they require deposition of structure factors. It will be an important task for the IUCr in the coming months to convince other publishers that if they wish to avoid publication of fraudulent structures the deposition of structure factors is necessary. The extensive checking procedures of the IUCr are already in use by a number of other publishers; it is my hope that more will realize that checking of crystal structures is essential prior to publication and that they will use the IUCr facilities.

So far I have mainly commented on the fraud that has taken place with structures of smaller molecules. Macromolecular crystallography is a highly competitive field and also contains examples of retractions of papers. Many journals that publish macromolecular structures limit publication of the experimental details that form the basis for the deposition of the structure in the Protein Data Bank. It is therefore extremely valuable that since January 2008 structure factors are now a mandatory part of a PDB deposition.

Looking to the future, one way to avoid scientific fraud would be to deposit the raw data recorded by the detector. Worldwide there is an increasing awareness of the need to deposit the raw data, but in order for these data to be useful and susceptible for analysis it is necessary to have suitable standards for the associated metadata. The IUCr could play a leading role in the development of standards for metadata, a process that will benefit from the experience of the R&D staff in Chester who are responsible for the development and implementation of the IUCr software. Since more than 80% of the crystal structures deposited in the Protein Data Bank are based on diffraction data measured at synchrotron facilities, these sites must also be involved in the process that can lead to the storage of raw diffraction data in a format suitable for subsequent checking. It is my hope that the recent bad experiences with fraudulent structures may in the future ensure that all crystal structures will be thoroughly checked through the analysis of the raw data and that the training of young scientists will also include the ethical aspects. I would like to conclude this difficult letter by thanking the Editor-in-Chief, the Section Editors, the Co-editors and the Chester staff for their diligence.

Sine Larsen (Sine@kemi.ku.dk)

28 April 2010