“I am deeply saddened and disturbed:” Co-author of retracted Nature paper reveals how problems came to light
On Wednesday, we reported on a Nature retraction of a paper whose corresponding author had also had a Cell paper retracted, and had been found to have committed a “violation of academic integrity” by Utrecht University. Today, we present the back story of how those retractions came to be, from another co-author of both papers, Ben Scheres, of Wageningen University:
On April 24th 2013 I was alerted to a rumour about anomalies in figures of two papers on which I was an author – Dhonukshe et al., Cell 149:383 2012 and Kakar et al., Nature 495:529 2013. After verifying that figures had been inappropriately manipulated, I contacted the authors and journals involved and the Universities affiliated with the main authors. Shortly after this, an independent investigation of the Kakar et al. paper was initiated based on a complaint filed at Utrecht University by my colleagues Chris Ambrose and Geoff Wasteneys from Canada.
The Kakar et al. Nature 2013 paper contained several duplicated parts of confocal images in Figs. 3a-b, 4d-e, 4g-i, cut-and-pasted RT-PCR images and cut-and-pasted seedling images. The confocal images (see PDF) were particularly worrisome as they were not exact duplicates but shifted in view (3a-b; 4g-i) or both shifted in view and compressed in the vertical dimension (4d-e). These anomalies undermined the credibility of all data in the paper. Once it became clear that the responsible author had no credible explanation and could not retrieve primary data, it was obvious to me that the paper should be retracted. I communicated this opinion to Nature in my role as a co-author.
My colleagues abroad had, in addition, obtained contradictory results on the localization of PIN2 in clasp mutants as described by Kakar et al. This scientific discrepancy caused suspicion and made them scrutinize the entire manuscript. Hence they noted many of the figure anomalies that we independently uncovered, which were included in their official complaint at Utrecht University.
After being informed of the paper discrepancies by myself and the Canadian colleagues, Utrecht and Wageningen Universities installed a joint committee on Scientific Integrity in May to investigate the case. Their report was handed over in August, after which the University Boards circulated a draft decision. Several complaints were raised against this draft, so advice was sought from the National Scientific Integrity Council (LOWI). The LOWI advice was returned to the Boards and considering this advice a final decision was recently published. The conclusion was that Pankaj Dhonukshe was solely responsible for the figure manipulations. The retraction of the Kakar et al. (2013) paper has just appeared online.
My decision to support retraction of both papers is based on the proven discrepancies in their content and does not rely on scientific questions of reproducibility of experiments. Neither do they depend on a judgement about the intentionality of the figure manipulations. For our scientific field, it is however of utmost importance to find out which conclusions in the retracted papers might still hold. We are currently repeating most experiments described in the Cell (2012) paper and some of those described in the Kakar (2013) paper. As soon as we have clarity on the outcome of these experiments, we will communicate our results. I hope that this will help those that have initiated research efforts based on these papers, to whom I sincerely apologize.
There are ongoing investigations into other papers (co)authored by Dhonukshe. We have meticulously scrutinized all data in the two papers of which I am the corresponding author and Dhonukshe a coauthor: Cruz-Ramirez et al., Cell 150:1002 (2012) and Prasad et al, Current Biology 21:1123 (2011). In the former paper, Dhonukshe contributed a single control experiment (depicted in Fig. 4I-K’) of which we possessed the primary data. The primary and published data matched, but we anyway repeated the experiment and it was reproduced. In the latter paper, the final version did not contain any data provided by Pankaj Dhonukshe. In conclusion, we have no reason to suspect data in either paper and thus we fully support their conclusions.
I am deeply saddened and disturbed about all these incidents, but I am confident that with the investigations carried out we will be able to restore the integrity of the scientific literature in our field. Finally, I thank Geoff Wasteneys and Chris Ambrose for uncovering these matters, and I am grateful for the numerous messages of support that i have received from colleagues, which have helped me a lot when dealing with these difficult issues.
Like Retraction Watch? Consider supporting our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, and sign up on our homepage for an email every time there’s a new post.