Szyf and Rabbani: old gels evil, new genomics cool

Canadian universities have a peculiar attitude to research ethics. First of all, University of Toronto remains convinced of Gideon Koren‘s innocence, the investigation into his patient abuse and research papers was buried, while the real baddies remain the whistleblowers like Nancy Olivieri. Which probably means there is absolutely no point to even mention the diabetes researcher and bearer of the Order of Ontario, Herbert Gaisano, significant co-author of currently 9 badly rigged papers on PubPeer. His University of Toronto will not care.

The University of British Columbia in Vancouver recruited former Toronto scientist Josef Penninger from Austria, to lead their Life Sciences Institute. All troubles about his past Canadian papers forgotten and forgiven, Penninger is presently curing COVID-19. And the University of Montreal is fully behind their biomaterials fabricator May Griffith: even the misappropriated research from Linköping, Sweden, is now apparently intellectual property of this Quebec university. The Montreal rector even tried to intimidate me in a letter, where he announced everything I report will be regarded as slanderous lies. So there is no point for me to write about any biomedical cheaters at that Canadian university either, I guess.

So let’s try McGill University, also in Montreal. The heroes of this story are its professors Moshe Szyf, from Israel, and Shafaat Rabbani from Pakistan. A dream team to crack epigenetics of cancer, both big names in the area of DNA methylation research.

As it happens, I can only discuss some of their older papers here, because as Clare Francis, who posted most of the PubPeer evidence about Szyf’s and Rabbani’s papers, commented: “it becomes “genomical” so you can’t readily interpret the data“. Indeed, the data in their newer papers is just bars, tables and diagrams. No fabricated pictures of gels or cells there, which, dear reader, means that science suddenly became honest and 100% reproducible once it got rid of those cumbersome images and switched to genomics and other big data-omics, which not even experts with access to raw material can always wade through.

Who needs those gel images anyway, where every Tom, Dick and Harriet could find Photoshopped falsifications. Hooray to the brave new world of genomics, proteomics, transcriptomics, secretomics, epigenomics and wank-omics with their pristine bars, graphs and numbers. No Photoshop needed there to get your paper in a high-ranking journal.

This is why there is no reason whatsoever to mistrust the newer papers by Szyf and Rabbani, unless maybe considering their older papers? Do leopards change their spots when they go digital? For starters, this joint fabrication rots on PubPeer for 4 years already:

Nicholas Shukeir, Pouya Pakneshan, Gaoping Chen, Moshe Szyf, Shafaat A. Rabbani Alteration of the methylation status of tumor-promoting genes decreases prostate cancer cell invasiveness and tumorigenesis in vitro and in vivo Cancer Research (2006) doi: 10.1158/0008-5472.can-06-1954

Bunch of copy-pasted bands in the pictures of fake western blots and RT-PCR gels. Nothing happened since that was posted in January 2017, at least nothing in the public domain. But to be fair, you can’t really correct or even explain that, no?

From around the same period, this interesting study on how to cure cancer via DNA methylation and data fabrication pathways:

Pouya Pakneshan, Moshe Szyf, Robin Farias-Eisner, Shafaat A. Rabbani Reversal of the hypomethylation status of urokinase (uPA) promoter blocks breast cancer growth and metastasis Journal of Biological Chemistry (2004) doi: 10.1074/jbc.m401669200

You might be shocked, dear reader, but I personally am not even sure anymore if the once so tough journal JBC will retract that, especially what with their collaboration with Elsevier and their recently settled lawsuit over a retraction with Raju Reddy.

Another paper published in parallel by this same team, very similar in its gel artistry:

Pouya Pakneshan, Moshe Szyf , Shafaat A Rabbani Methylation and inhibition of expression of uPA by the RAS oncogene: divergence of growth control and invasion in breast cancer cells Carcinogenesis (2004) doi: 10.1093/carcin/bgi009

The first author of both of these masterpieces, Pouya Pakneshan, is currently acting as Principal Global Scientific Director at Roche in San Francisco, which is probably great news for this pharma giant’s drug discovery pipeline (Roche must compete with Sanofi, I guess). Photoshop could obviously cure cancer in vitro, surely this attitude to experimental research also translates into successful clinical trials and medicine market. Here another one of Pakneshan’s papers with her PhD mentor Rabbani:

Pouya Pakneshan, Bernard Têtu, Shafaat A. Rabbani Demethylation of urokinase promoter as a prognostic marker in patients with breast carcinoma Clinical Cancer Research (2004) doi: 10.1158/1078-0432.ccr-03-0545 

Apparently, a fabricated RT-PCR gel again, good thing the Photoshop-prone technology is extinct now, these days you get some plotted colour bars from the digital Real-Time PCR where you must simply trust the authors. As Cheshire commented on PubPeer also in this case:

“Figure 3A appears to be drawn from Figure 3-3 in the 2004 PhD thesis of the first author.“

I hope that thesis won some award. Here one more from Pakneshan’s PhD period with Shabbani, featuring Szyf:

Yongjing Guo, Pouya Pakneshan, Julienne Gladu, Andrew Slack, Moshe Szyf, Shafaat A. Rabbani Regulation of DNA methylation in human breast cancer. Effect on the urokinase-type plasminogen activator gene production and tumor invasion Journal of Biological Chemistry (2002) doi: 10.1074/jbc.m201864200

Those gels, eh? Good that technology moved away from images, trust is restored because you won’t be able to find anything anymore. In this regard, this final one also, please (there is more by Pakneshan on PubPeer!):

Pouya Pakneshan, Rosie Hongmei Xing, Shafaat A. Rabbani Methylation status of uPA promoter as a molecular mechanism regulating prostate cancer invasion and growth in vitro and in vivo FASEB journal (2003) doi: 10.1096/fj.02-0973com

Beautiful artwork. Will McGill University investigate Pakneshan’s PhD thesis and the role of her mentor Rabbani? Surprise coming at the end!

But first, it’s not like you can blame the student for the following Rabbani paper:

There are two authors, and the second one, Julienne Gladou is not even a scientist. No prizes for guessing who fabricated the Figure 1B picture in that publication.

Maybe Szyf fell victim to malfeasances in Rabbani’s lab? Is he maybe just too trusting? Here the outcome of Szyf’s other collaborative effort, with the lab of the neurologist Michael Meaney, also at McGill University:

I. C. G. Weaver, A. C. D’Alessio, Shelley E Brown, I. C. Hellstrom, Sergiy Dymov, Shakti Sharma, Moshe Szyf, Michael J Meaney The transcription factor nerve growth factor-inducible protein a mediates epigenetic programming: altering epigenetic marks by immediate-early genes The Journal of neuroscience (2007) doi: 10.1523/jneurosci.4164-06.2007

Why are these gels copy-pasted to stand in for similar, but sufficiently differently designed experiments? Here another such collaboration, with similar authors:

Ian C G Weaver, Frances A Champagne, Shelley E Brown, Sergiy Dymov, Shakti Sharma, Michael J Meaney, Moshe Szyf Reversal of maternal programming of stress responses in adult offspring through methyl supplementation: altering epigenetic marking later in life The Journal of neuroscience (2005) doi: 10.1523/jneurosci.3652-05.2005

That problem with falsified and recycled gels was flagged by Elisabeth Bik, who referenced another collaborative paper by Szyf and Meaney labs, again with a very similar authors list, Weaver et al Nature Neuroscience 2004 (Update 10.03.2021: McGill University and Dalhousie University investigated Weaver et al 2004 in 2016-2017 and now decreed it’s exempt from all future investigations). It was criticised on PubPeer with allegations of irreproducibility, maybe these fake gels provide a possible explanation?

The first author and Meaney’s mentee Ian Weaver is now professor at Dalhousie University, also in Canada. So how can we rescue Szyf from his friends? But wait, this is purely from Szyf’s own lab:

These bands sure look copy-pasted, but the authors removed all background signal, not really helpful. Also this paper might be problematic:

Something weird happened here. Some bands do look strangely similar….

But now , the surprise I promised. This Rabbani paper has a star of research fraud Anil Potti as coauthor, but was Potti really to blame for this one?

Gaoping Chen, Kanishka Sircar, Armen Aprikian, Anil Potti, David Goltzman, Shafaat A. Rabbani Expression of RANKL/RANK/OPG in primary and metastatic human prostate cancer as markers of disease stage and functional regulation Cancer (2006) doi: 10.1002/cncr.21978

Fake images, fake gels, what does one do with that one? Retract swiftly, sack the perpetrator? No, this is Canada, dear reader!

That joke of a paper was in fact corrected by the publisher Wiley in September 2018, the notice declared:

“An investigation was conducted by a McGill University Ad Hoc Committee, and the first author acknowledged to the committee that he had made a mistake when selecting from hundreds of photographs to assemble figure 2. The committee concluded that the image manipulation resulted from unintentional human error and did not affect the scientific validity or soundness of the data and findings reported in the publication at issue. In addition, such findings have since been independently replicated in other studies.”

Student, meet bus. A certain biomedical researcher Gaoping Chen now works at the neighbouring University in Montreal, often it pays off to take the fall for someone else.

Why did I bother writing about another Canadian university, silly me….

Update 22.01.2021

Well, here proof that the shenanigans continued even after Szyf and Rabbani moved on from image data to genomics. Here a rather new paper:

David Cheishvili, Surabhi Parashar, Niaz Mahmood, Ani Arakelian, Richard Kremer, David Goltzman, Moshe Szyf, Shafaat A Rabbani Identification of an Epigenetic Signature of Osteoporosis in Blood DNA of Postmenopausal Women J Bone Miner Res. (2018) doi: 10.1002/jbmr.3527. 

Time for these two to go into a science menopause and just leave.

Update 4.02.2021

I notified the university on 22.01.2021. On 1 February, McGill Deputy Research Integrity Officer David Ragsdale wrote to me that:

“merely providing links to PubPeer or other websites does not suffice as a formal complaint, since the sites do not specify a Respondent and can be continuously updated by anyone.”

I was asked to assemble all PubPeer evidence into a complaint form, which I refused because it’s not my job. Ragsdale stopped replying to emails, it seems the notification was not admitted and there will be no investigation at McGill.

Update to this update: Few hours after I shared the above quote via a tweet with McGill University tagged, Ragsdale wrote to me:

“We take the allegations seriously and are conducting an inquiry based on the information provided in the links in your previous emails.”

Update 10.03.2021

The article was updated above to include the information provided by the McGill Deputy Research Integrity Officer David Ragsdale, who wrote to me that a list of 8 Szyf and Rabbani papers has been investigated already in 2008, 3 more were investigated by McGill and Dalousie universities in 2016-2017. I double-checked with him several times, and indeed, Ragsdale confirmed that even though the evidence for all these 11 papers was posted on PubPer for the very first time in 2021, this evidence is not admissible anymore, because the papers were already investigated, and nothing wrong was found. Ragsdale instructed me:

“I have carefully reviewed the 14 papers cited in your blog post. For 11 of the 14 papers, the issues raised have already been assessed by McGill University, Dalhousie University, CIHR, and/or the journals. These papers will not be re-adjudicated, as per Canadian Tri-Agency RCR policy (https://rcr.ethics.gc.ca/eng/interpretations_allegation.html):

“The second element for an allegation to be considered responsible is that the allegation must be novel and, to the best of the complainant’s knowledge, never previously investigated.” 

This leaves three papers:

Campbell, Bovenzi, Szyf (2004) Carcinogenesis, 25:499-507, Figure 1

Torrisani et al, (2007) Molecular and Cellular Biology, 27:395-410, Figure 2E

Weaver et a., J Neurosci (2007) 27:1756-1768, Figures 4 and 5.

As part of McGill and RCR policy, a misconduct inquiry is initiated by a good faith complaint, made in confidence to the RIO (https://rcr.ethics.gc.ca/eng/interpretations_good-faith-bonne-foi.html). If you are concerned that figures in the above papers were inappropriately manipulated, then what I request from you is a written complaint explaining, for each of the figures, why you think the evidence you present suggests the figures were falsified. This would be the document to which the Respondents would provide a rebuttal. 

Note that a good faith allegation is expected to maintain confidentiality, as per RCR policy”

After some back-and forth, Ragsdale then agreed to add three more papers which were not previously investigated.

Ragsdale then invited me to write the investigative report for him, so his McGill colleagues can whitewash 6 more papers:

“According to the McGill Regulations on the Investigation of Allegations of Research Misconduct( https://www.mcgill.ca/secretariat/files/secretariat/research-misconduct-regulations-concerning-investigation-of_1.pdf)  the review of allegations of research misconduct is triggered by a complaint.  The complaint has to spell out the allegation, (for example, is it fabrication, misappropriation of IP rights, etc.).  The complaint should provide evidence of the alleged misconduct, and indicate how that evidence suggests misconduct.  It should also spell out who is accused of misconduct.  To be clear, comments such as “this figure looks strange” or “this is much more similar than expected” are not particularly helpful.  The allegations should explain what is being alleged (duplication, reproduction of the same figure, etc.) what elements of context are relevant to the analysis, what similarities are expected (or not), what manipulations may have occurred (if that is the allegation) on what basis the similarities pointed to suggest misconduct, etc. “

I told Ragsdale I charge €200 per hour and never heard back. So I don’t submit anything, publish this communication and breach confidentiality. Which means: the McGill University will not be investigating anything now.

But the fake data is public, so form your own opinion of Szyf’s and Rabbani’s research. And of McGill’s attitude to ethics.

---