Poland is a country where lots of potatoes are eaten. The good old ziemniaki can be enjoyed boiled, mashed, fried, baked, with dill or without, as dumplings, latkes, frytki, babka, or as traditional filling for pierogi. Lucky for Poland, its Academy of Science has the best potato researcher in the world. If anything should ever happen to Polish potato harvest, Professor Agnieszka Kielbowitz-Matuk at the Institute of Plant Genetics in Poznan can easily cook up a solution to all Solanum problems.
Before we move on to the art of cooking potato data, a brief look back at past troubles with research integrity in that research institution.
Past Poznan Problems
It is not even the first data fudging affair at this plant genetics institute of the Polish Academy of Sciences (IGR-PAN). There was also the case of the two Pakistani nanofabricators Gregory Franklin and Qaisar Maqbool, both funded with an EU grant called BioTalent. After the investigation, caused by my reporting of falsified data and described by one of IGR professors there as “the thunderstorm in our Institute caused by the case of photoshopped pictures done by the PhD student Qaisar Maqbool“, the student left Poland; Franklin was whitewashed and made associate professor at IGR. Three Maqbool-led papers were retracted for various kinds of misconduct, two had Franklin as last author. One fraudulent preprint was deleted by the authors. Maqbool is now doing PhD in Italy, his new PI announced vigilance.
Then, there was also a strange case of the IGP graduate Maciej Majka‘s undercooked raw data. It was about this paper:
M. Majka, M. Kwiatek, J. Belter, H. Wiśniewska Characterization of morphology and resistance to Blumeria graminis of winter triticale monosomic addition lines with chromosome 2D of Aegilops tauschii Plant Cell Reports (2016) doi: 10.1007/s00299-016-2023-x
A PubPeer user commented 4 years ago, in 2017: “In Figure4, lanes 1,2 and 3 and duplicated.” It sure looks like this. Majka replied in January 2020:
“You suggested that some lines are duplicated and I agree that these lines may be a bit confusing because of the similarities between them, but it is not true. To prove it, I have attached, the original photo of the agarose gel which part was used to prepare the Figure 4. There is no scientific advantage in manipulating this figure. What is more, this doubt does not affect the main results as well as conclusions of this paper.“
Problem is, the gel image Majka supplied is sure very similar, but… it doesn’t show the same gel as the figure shows, this was also confirmed by Elisabeth Bik and other experts. Here an analysis by the PubPeer user Oeneis Buddha:
Majka eventually conceded, but still insisted the image would fit:
“Indeed, I was not precise enough. The photo described as ‘raw’ comes from the same gel, but it was taken after a longer time and under a different exposure. Our gel doc system should be upgraded, it is hard to high tune it so well.
All in all, I was not clear; the attached photo is not identical. Unfortunately, the original photo is not stored in my files, what I deeply regret. Nevertheless, the photo of the gel that I have uploaded originates from the same experiment. […]
Thanks for the comments. This is my first important paper and the whole story teaches me to store all the results for a long time and formulate clear explanations. A bitter lesson, but it will help to improve my skills“.
The former IGP deputy director Malgorzata Jedryczka however seemed to have felt an investigation was needed:
“you have some serious accussations towards Dr Maciej Majka and I do not feel competent and decisive to discuss this case. My role as the Deputy Director for Research finished at the end of 2019 and there is a new Director and his Deputy since then. […] This is why I am redirecting your mail to Prof. Pawel Krajewski – the current Director of the Institute.“
The case was closed. But remember this name: Pawel Krajewski, IGP director and bioinformatics professor, coordinating several EU projects. He will play a role in this story’s main feature, potatoes à la Kielbowicz.
Ziemniaki à la Kiełbowicz
The potato researcher Agnieszka Kiełbowicz-Matuk studied in Poznan, did her PhD there, and returned as professor after two short postdoc stints in Germany and France, one year each. From 2017 to 2020, she managed a €200k nationally-funded research project together with Krajewski, who is now IGP-PAN director. This biography also mentions:
“Since 2010, she has been a member of the Scientific Council of IGR PAN in Poznań, where in 2015-2019 she was the Secretary and member of the Presidium of the Scientific Council. At that time, she was also a member of the Council of Young Scientists of IGR PAN. In 2010-2015, she was a member of the Management Board of the Poznań Branch of the Polish Genetic Society.
In this high function at IGR-PAN, Kielbowicz was likely also responsible for research integrity. Possibly also for whitewashing Franklin in 2019. So now let’s have a look what kind of science Kielbowicz herself has published:
Agnieszka Kiełbowicz-Matuk, Jagoda Czarnecka, Ewa Banachowicz, Pascal Rey, Tadeusz Rorat Solanum tuberosum ZPR1 encodes a light-regulated nuclear DNA-binding protein adjusting the circadian expression of StBBX24 to light cycle Plant Cell & Environment (2017) doi: 10.1111/pce.12875
As you see, the gel bands already featured in an earlier Kielbowicz paper:
Agnieszka Kielbowicz-Matuk, Pascal Rey, Tadeusz Rorat Interplay between circadian rhythm, time of the day and osmotic stress constraints in the regulation of the expression of a Solanum Double B-box gene Annals of Botany (2014) doi: 10.1093/aob/mct303
Total train wreck, these gel bands are all fake, let’s call a spud a spud. These and the following two papers have regular co-authors: Kielbowicz’s mentor at IGR-PAN, Tadeusz Rorat, and Pascal Rey, tenured staff scientist at French CEA, whom Kielbowicz must have met during her brief postdoc stint there.
Let’s have a look at the other two papers, which are a bit older.
Agnieszka Kiełbowicz-Matuk, Pascal Rey, Tadeusz Rorat The abundance of a single domain cyclophilin in Solanaceae is regulated as a function of organ type and high temperature and not by other environmental constraints Physiologia Plantarum (2007) doi: 10.1111/j.1399-3054.2007.00968.x
I don’t really have to explain what happened here, right? Pick an f-word. And finally, this:
Agnieszka Kielbowicz-Matuk, Pascal Rey, Tadeusz Rorat The organ-dependent abundance of a Solanum lipid transfer protein is up-regulated upon osmotic constraints and associated with cold acclimation ability Journal of Experimental Botany (2008) doi: 10.1093/jxb/ern088
Already half a year ago, Kielbowicz-Matur replied on PubPeer regarding the latter and the 2014 paper from above:
“I will try to find all the materials and attach them again. Each of the experiments has been repeated 3 times, so I hope that they have survived with any films after so many years. The photos were strongly brightened. The pictures were also narrowing to sharpen the products on the gel. Please let me know by when I should attach the documentation and if there is any technical repetition I can find.“
Despite the invitation from the Pubpeer sleuth Aneurus Inconstans to post raw data, Professor Kielbowicz-Matur was too busy. Does the raw data exist, at all? After all, everyone knows potatoes must never be consumed raw, is this why they first cook their data there?
And now we finally return to the IGR-PAN director Pawel Krajewski, who explained to me what’s wrong with the Kielbowicz papers. Namely nothing at all:
“The authors of the articles are in contact on this matter with the publishers of the relevant journals. No accusation of “fraud” was formulated by the publishers.
In my opinion, the sentence “… bands … are much more similar than expected” is meaningless and pseudoscientific as it is not supported by any results of research on distributions of “simililarities” of bands based on their populations and appropriate statistical hypothesis testing procedures about their “expected” values. Please note that without such evidence, this sentence must be itself called “fradulent” if used to support any (scientific) conclusion, for example on the quality of some research inference.“
That is actually even funnier given the fact that Krajewski is bioinformatician and hence expert in digital technology. I thanked the director for his reply, which in my view rivals the science fiction by Stanislaw Lem because it has even less to do with reality. I also asked Krajewski to confirm that he and his institute indeed never asked Kielbowicz to provide any raw data for the contested figures. Krajewski replied:
“Thank you for your interest in integrity of research at our institute. I will be able to serve you when anything new is known.“
The only thing Krajewski seems to know is who the real fraudsters are: myself and the PubPeer sleuths. What a guy.
As you may have seen Krajewski posted a comment under this article, and he and I exchanged several emails where he stated same:
“So far the Journal of Experimental Botany informed the authors that the checks were perfomed by experts, nothing wrong was proved, and no further actions will be taken. Their answer is also related to the material published in Physiologia Plantarum. Also, they admit that, due to circumstances that the authors described, they accept the explanation concerning availability of orginal data.“
I contacted the editors of both journals, but received no reply. But a reader did receive a reply, which the reader kindly shared with me. The full statement by John Lunn, emeritus senior researcher at Max Planck Institute in Golm, Germany, is here. Now Lunn’s key quotes, explained:
“we set up an investigating committee comprising myself, three other senior scientific editors, the managing editor and another member of staff with expertise in the image analysis software (Photoshop) that we currently use to examine all figures before publication. Our first step was a request to the authors to provide the original gel images for the figures in question. Their response was that these were no longer available, as the lab where the work was carried out had disbanded about four years ago following the retirement of Prof. Rorat. […] in the past it was a regrettable but common practice to destroy lab records when a group disbanded,”
Meaning: Raw data disappeared completely.
“There was unanimous agreement that there were marked similarities between some of the bands in Figures 2-7 in Kiełbowicz-Matuk et al. (2008), and also similarities with some of the bands in Figure 3 in Kiełbowicz-Matuk et al. (2007). However, given the relatively poor quality of the scanned images, and the need to resize some of the images to allow them to be overlaid for comparison, five out of the six members of the committee judged that the case for duplication of bands in the figures was not proven beyond reasonable doubt. Based on this majority verdict, the final conclusion of the committee was that there was not a strong enough case to justify a retraction of the Kiełbowicz-Matuk et al. (2008) paper.”
Meaning: No denial that the bands are likely duplicated, while their main argument against similarity is the “need to resize” (sic!)!!! Thus, stretching images to hide similarities is nowadays proof that the images are not duplicated. JXB may have had a Photoshop expert on board, but here my own take on Figure 3.
Now Lunn explains how he and his JXB colleagues decided to reject the evidence of their own eyes:
“As a professional courtesy, we informed the Editor-in-Chief of Physiologia Plantarum (Prof. Vaughan Hurry) that we were investigating the allegations that some of the data published in Kiełbowicz-Matuk et al. (2007) Physiologia Plantarum 131; 387-398 had been duplicated in the Kiełbowicz-Matuk et al. (2008) paper in JXB. As noted above, one of the findings of our investigating committee was that, while there were similarities between some of the data in the 2007 Physiologia Plantarum paper with those in the 2008 JXB paper, the committee could not conclude beyond reasonable doubt that the data were identical and therefore duplicated. The JXB investigating committee examined only the question of data duplication with respect to the 2007 Physiologia Plantarum paper. It was not within their remit or capability to examine and comment on the validity of the data in the 2007 Physiologia Plantarum paper. It should be noted that even if the allegations of data duplication in the 2008 JXB had been substantiated, this would have had no bearing on the validity of the data in the 2007 Physiologia Plantarum paper, which was published before the JXB paper.“
Meaning: we created a straw man debate of alleged band similarities between Kiełbowicz-Matuk et al. (2007) Physiologia Plantarum and Kiełbowicz-Matuk et al. (2008) JXB, which was never reported on PubPeer or anywhere else. Having debunked something we ourselves made up, we decided that because our own false allegations were wrong, the actually reported and undeniable gel band duplications inside each of the two papers become irrelevant.
“We have not concluded anything at this point, regardless of what is written online. The initial discussion only relates to the potential overlap between the two journals, not the 2007 data in PPL. […] I need to be able to make a fact-based investigation, or rather have the host institute do so.“
Problem is, Krajewski told me several times he leaves the investigation to the journals while his IGR-PAN plans none.
If you are interested to support my work, you can leave here a small tip of $5. Or several of small tips, just increase the amount as you like (2x=€10; 5x=€25). Your generous patronage of my journalism will be most appreciated!