11

We are about to submit a revised manuscript for publication in a prestigious journal. This is a team work and one of my colleague performed most of the molecular biology experiments, including many electrophoreses.

I just noticed that on one of these electrophoreses, if play around with the contrast etc., some bands are surrounded by a rectangle with a different background as if the image had been assembled.

When pointed this weird detail (without making accusations) and asked where that came from, he said something along the lines of "wrong image, my mistake", and sent me a different gel later on (without this weird signal).

Now, I can easily imagine how one may accidentally mislabel an image, but I have a really hard time imagining how one can accidentally copy-paste signal from one image to another. I am now concerned about the validity of all the work that was performed by this teammate. Not only that experiment, but also graphs etc.; while I have no further evidence of weird results, I am now suspicious of anything that was produced by this person.

The deadline for sending our revisions is in a few weeks, and the experiments performed by this teammate are both 1) too tedious to be repeated within the next few days, and 2) too important to be removed from the paper.

I am afraid that if I bring this up with our PI (who is good friends with this colleague), and if my colleague finds a way to talk himself out of this, I would be seen as the "team breaker" and have a really bad time for the rest of my graduate studies in this laboratory.

How should I approach this, and with who?

Improve this question
  • 9
    Just to add some context for people not familiar with molecular biology... it is usually okay to adjust contrast, brightness, etc. of your images, but copying and pasting a band to another gel is an absolute no-no and there is literally no justification for that under any circumstance. – Drecate May 14 '19 at 2:59
  • 1
    @Drecate As someone without the background, I was again confused by your comment because I guess there is a distinction between band and lane that I was unaware of. I've seen people copy and paste lanes (for comparison), but I guess the band is what's dishonest – aidan.plenert.macdonald May 14 '19 at 15:21
  • @Drecate Definitely not from the field, but since you seem to be: I have a recollection of a relatively famous (?) molecular biology (?) paper being retracted for falsifying results, and the description key-words were similar to here ("adjusting contrast", "artefacts appear around some bands", "cutting and pasting bands together"). Was there indeed such a paper, retracted for reasons similar to what OP suspects, either recently enough to make a splash, or famous enough to be remembered (and overheard by computer scientists)? – penelope May 14 '19 at 16:14
  • @aidan.plenert.macdonald Yes the lanes and the bands are two quite different things... Since bands on different lanes migrate more or less independently, it is okay to crop out the lanes for side by side comparisons. – Drecate May 14 '19 at 20:41
  • @penelope Sorry I don't really keep close track of this kind of things... but what you described sounds imminently plausible. And just to be clear, when I said "adjusting contract", I meant the contrast of the whole image, not a rectangular subsection of it. – Drecate May 14 '19 at 20:43
12

Talk to your PI and/or colleague. It might be wise to start with something less confrontational than the "I think [Colleague] may have forged data", however. Maybe "hey, I still want to understand what was going on with that figure. Perhaps other figures might be affected too? What do you think?".

Best case scenario, there is a convincing answer that can calm your nerves. There are a whole slew of other worse possible scenarios that can follow from this discussion, but they're all better than finding out after publication that your suspicions were correct. As for the revision deadline, it can likely be extended if need be. It'd be up to the editor, but I imagine few prestigious journals would say no to a reasonable extension request to double check something important.

Improve this answer

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.