1. daen on July 5, 2011 2:11 PM writes...

I could imagine using something like Google image search for this - you can now search for a specified image (so try Googling images for Western blot, and in the hover-over popup, click 'Similar') ...

2. omar on July 5, 2011 2:59 PM writes...

Repeated figures, supposedly representing different things, are exactly what got Jan-Hendrik Schon caught in his famous condensed-matter physics flameout. You can see the examples halfway down this page, where the three figures all come from Science and Nature articles:
http://www.pro-physik.de/Phy/leadArticle.do?laid=953
(site is in german, but you don't need to read it to see the evidence). The key here was that the noise, random by definition, is the same in three measurements of different materials.

3. Anonymous on July 5, 2011 3:45 PM writes...

this makes me recall a Nature paper back in April. The author did exactly the same thing by coping and pasting their own figures and claiming they are from two different experiment. http://dx.doi.org/10.1038/nature09874 See Figure 1a and 1c. Well maybe I can still accept this happens to JBC, but Nature?

4. Steve on July 5, 2011 3:56 PM writes...

Here are the author guidelines for Nature journals

http://www.nature.com/authors/policies/image.html

We don't employ any automatic checking of figures yet (we do have access to tools for plagiarism checking of text). Note that we request authors retain the original data - so that we can check for inappropriate image manipulation if we feel it is necessary.

Unfortunately - as with all cases such as this - it will still be difficult to detect a determined fraudster.

5. Smoki on July 5, 2011 4:18 PM writes...

What really bothers me with most western blots published in articles nowadays is extensive cropping of the images. It's ubiquitous, even in the very best journals. You get just a little crop-out of the band, and nothing above and below.

Makes you really think that all the authors are hiding other bands - but for what purpose, really? Anyone who has done a lot of westerns (like myself) knows that this is really not an all black-and-white method, often you see all sort of things you cannot really explain. But someone else might draw some conclusions from the additional data, which gets cropped out time and again. Perhaps reviewers are to blame, unexplained bands might really draw a lot of unwanted attention. But how is this different from cropping out everything but the part which supports your story?

6. Cellbio on July 5, 2011 5:34 PM writes...

Anyone who has done a lot of westerns (like myself) knows that this is really not an all black-and-white method.....

Well, it is if you turn up the contrast. You should try this, as it lessons the need to explain things. Helps with the specificity of your antibody too.

7. mad on July 5, 2011 7:44 PM writes...

Funny that even in fabricating data they were so lazy that not only did they not run the experiment that they did not evn bother to run a "fake" experiment to have a different western... Not knowing anything about the details my first impre4ssion from reading this was that maybe they really did run it and got the reported results but just lost the film so they stuffed in a copy just to show the "answer".

8. stephen on July 5, 2011 8:10 PM writes...

i'm sure there were scientist in the field who knew it was a fake paper within the same year that it came out. This happens often. There is a code of silence among scientists that they should not rat each other out.

9. gippgig on July 5, 2011 10:14 PM writes...

Google "open source" "image comparison" or "image search".

10. DrSnowboard on July 6, 2011 2:54 AM writes...

But they all look the same anyway...?

11. RB Woodweird on July 6, 2011 6:20 AM writes...

This just reinforced my low opinion of biochemists. Not because they are cheating, but because they are so incompetent at it. How hard would it be to write a program that generates an image of a gel with any band you want anywhere you want it?

12. Vader on July 6, 2011 11:03 AM writes...

I'm not a biologist, but could this be inadvertent? I've occasionally link the wrong image into a document and not caught my mistake at once. Could the author have inadvertently link the image twice?

I guess I'm invoking Heinlein's Law: Never attribute to malice what can be adequately explained by stupidity.

13. Paul on July 6, 2011 11:08 AM writes...

Are there any tools out there that use Benford's Law to search for faked data in papers?

14. dearieme on July 6, 2011 2:34 PM writes...

"from the group of the late Maria Diverse-Pierluissi": is it easier to catch the dead ones?

15. anon on July 6, 2011 10:27 PM writes...

One wonders what did she die of? She shouldn't have been old (became PI in 1998), judging by obituary from Mt. Sinai. Even if a person is a liar who consistently publishes fake results, a tragic accident, or a fatal disease still make loss of her life a sad occasion. But if she was being investigated for scientific fraud (just slightly over one month between her death and journal retracting papers - Mt. Sinai internal investigation must have started earlier), conceivably, she might have decided to take her own life to escape inevitable destruction of her career and reputation. If true, that brings the fierce competition in academia to new heights.

16. Innovrich on July 8, 2011 9:16 AM writes...

We all often come about academic results that appear unrepeatable. Some we've talked to other companies about and they get the same lack of result. Some labs seem to survive entirely on this unrepeatable data publication - often for decades! Until the main incentive for academics merely to publish, changes, this fakery will continue.

17. notabigshot on July 27, 2011 9:33 AM writes...

"a code of silence among scientists that they should not rat each other out. "... not really. actually it is a VERY competitive field, we all have an incentive to cut each other down. funding is a zero-sum game...

but if you think something hinky is going on, especially with a bigshot, and you're wrong - or if you simply can't (or don't have the time, or access to raw data like the journals...) prove it - then it is very risky to speak up.

science - or at least getting the funding to keep science going - is ALL about politics, and a strong old-boy network quashed whistleblowers. you can go all the way back to the first baltimore case to see that...

18. anon2 on February 16, 2012 2:49 PM writes...

anon:

I work at Sinai and heard through the grapevine that she died of an aggressive form of breast cancer.