Ten days ago, we reported on the dismissal of Zhiguo Wang, a Montreal Heart Institute researcher who had already retracted two papers because of image manipulation. At the time, an official said the institute had requested three more retractions, but when we asked which three papers, we were told:
As written in the press release, the MHI has requested the retraction of three additional scientific articles. We will not be able to confirm the name of the scientific articles and/or publications until confirmation of the retractions.
The first of those three has now appeared, in the Journal of Cell Science, for the 2007 paper, “The muscle-specific microRNAs miR-1 and miR-133 produce opposing effects on apoptosis by targeting HSP60, HSP70 and caspase-9 in cardiomyocytes.” According to the retraction notice — which is unfortunately behind a paywall (see update at end):
After it was brought to our attention by the Journal of Cell Science, careful examination of the above paper published in the journal in 2007 highlighted some errors we made related to the re-use of our previously published western blot bands in parts A, B and C of Figure 3, and duplication of western blot bands in Figures 3A and 4A. The misuse and re-use of western blot bands violated the editorial policy of Journal of Cell Science, and so we must retract this article.
The errors originated in Dr Zhiguo Wang’s laboratory at the Montreal Heart Institute. Dr Wang takes full responsibility and apologises to the editors and readership of Journal of Cell Science for any inconvenience caused and any negative impact this might have on the journal.
We had a good hunch that this paper would be retracted, thanks to a number of Retraction Watch readers who, as we’ve come to expect, ripped apart the images in Wang’s papers brilliantly. Based on those analyses, we asked five journals who’d published his papers whether they were planning to retract them.
The Journal of Cell Science said they couldn’t comment until a retraction appeared, but suggested we look at the next issue, which has obviously now been published. We never heard back from anyone at the Journal of Cellular Physiology, which published “Overexpression HERG K+ channel gene mediates cell-growth signals on activation of oncoproteins SP1 and NF-κB and inactivation of tumor suppressor Nkx3.1.”
Here are the other three responses:
PLoS One, which published “Transcriptional and Post-Transcriptional Mechanisms for Oncogenic Overexpression of Ether À Go-Go K+ Channel:”
PLoS ONE is already aware of the concerns regarding one of the figures in this article. I can tell you that the situation is currently being investigated; however, due to our policies on confidentiality, I am afraid that I am unable to provide you with any additional details.
Nature Medicine, which published “The muscle-specific microRNA miR-1 regulates cardiac arrhythmogenic potential by targeting GJA1 and KCNJ2:”
We are aware of the allegations of misconduct about this paper. We therefore got in touch with the authors for an explanation. While we waited for their reply, we became aware of an independent investigation into these allegations. The committee at the Montreal Heart Institute in charge of the investigation is presided by Prof. Eric Thorin, who informed us that they will issue a report on the case by mid-September for follow-up action. Until the report is made public, there is nothing more to add about this incident.
EMBO Journal, which published “miR-605 joins p53 network to form a p53:miR-605:Mdm2 positive feedback loop in response to stress:”
Thank you for your inquiry. All I can say at this time is that we are in discussions with the author and the journal PLoS One.
We’ll of course continue to follow up.
Update, 9:45 p.m. Eastern, 9/21/11: Journal of Cell Science executive editor Sharon Ahmad tells us that the paywall in front of her journal’s retraction notice was an error on their part, and has been fixed.
OK, it’s trivial, but in the title and first paragraph, it should probably be “Zhiguo,” not “Zhighuo.”
Fixed — thanks.
I guess that Western blots are indeed truly frustrating affairs. It seems that manipulation of this particular experiment is the most common, or perhaps, most easily detectable, of all figure frauds encountered.
The discussion of bands you point to is some pretty nifty sleuthing – well done commenters.
I can recall the first time the lab I worked in got access to Photoshop. Just to see if it could be done, a colleague of mine used the clone tool to “create” a mutation in an autoradiograph of a DNA sequencing reaction (that’s the old school way of sequencing folks, before fluorescent dyes and capillary sequencers).
It looked perfect. I remember thinking at the time, “oh, this could be bad”.
Disclaimer: neither he nor I ever used this technique for anything important, although I think he created a $200 bill for fun as well.
Western blots themselves are really not that difficult. Of course, generating the result you want out of the WB is a different story. Anyone who is familiar with the technique knows that there are many ways in which WBs can be manipulated and most of these are way before Photoshop even comes in to play. It would be almost impossible to spot. The only reason this guy got caught is because he started using the same figures in multiple papers. Then people started paying more attention and the rest is history, or soon will be.
Shameless self-promotion:
I’m writing case studies related to the teaching of the Responsible Conduct or Research. (Retraction Watch is an invaluable resource for me.) On one case study I pose the question,”Which is the more reliable type of data, the microarray image of the cellular expression or the riboprobe?” Maybe I should re-word the question to “more difficulty to falsify”.
Website where this occurs:
http://www.uab.edu/ethicscenter/educational-materials/rcr-materials/image-manipulation/image-manipulation-case-study
I welcome feedback from this community.
Reliability of data used to mean a different thing, not resistance to fraud. Obviously, there is nothing that cannot be falsified in scientific experiment or its interpretation and presentation. Second, don’t forget that these falsifications are carried out by scientists and they can invent any new fraud that nobody else can even think of. So, I wouldn’t measure that side of reliability.
Does anyone know of any formal analysis of the rates of multiple retractions of an author(s) verses a singular retraction by an author(s) where FFP and not honest scientific error is the reason for retraction?
As of today the paper does not link to a retraction
Yes, it does. But it surely requires very close attention. On the right hand side you have the following links:
Summary
Figures Only
» Full Text
Full Text (PDF)
A retraction has been published
Supplementary Material
More fabricated evidences have been found for this dismissed Montreal cardiology researcher Zhiguo Wang. He has at least 9 articles with fabricated data as the Chinese Blog indicated: http://blog.sciencenet.cn/home.php?mod=space&uid=51597&do=blog&id=492774#comment
Gao et al. （2006） Mol Pharmacol 70, 1621
Lin et al., （2007）J Cell Physiol 212, 137.
Xiao et al., （2007） J Biol Chem, 282, 12363-12367.
Xu et al., （2007）J Cell Sci 120 (17), 3045-3052.
Yang et al., （2007） Nature Medicine 13, 489.
Luo et al., （2008）J Biol Chem 283 (29),20045-20052.
Lin et al., （2009）Cell Physiol Biochem 23, 317-326.
Lin et al., （2011）PloS One 6(5): e20362
Xiao et al., （2011）EMBO J 30, 524.
Three of them have been retracted as you guys described. Others have not.
It’s really terrible and shameless