#8. Gel image alterations – what’s acceptable?
This post is part of a weekly series about image screening. All published posts can be found on this page.
When conducting our routine figure screen, our production and editorial team carefully examines all gel images for resolution, but also for any other type of alteration. Our rule is that “no specific feature within an image may be enhanced, obscured, moved, removed, or introduced.” In particular, we watch for the following alterations:
- Splicing lanes out of a gel is permitted if consistent throughout panels, disclosed in the figure legend and represented in the figure by a clearly visible line. We believe, however, that all samples blotted for a protein of interest must come from the same membrane/gel as those blotted for a loading control, so inconsistent splicing between loading control and protein of interest is not accepted, and we will ask you to revise the figure if we observe this type of inconsistency.
- Duplication of panels and deletion/addition of bands constitute misrepresentations of the original data and are not appropriate. (We’ll get into what we would do if we found such issues in a JCB submission in our next post.) Examples are shown here.
- Background cleaning or brightness/contrast adjustments (examples here) cannot be done to the extent that all background information has been lost or that specific elements are obscured or enhanced. For instance, although non-specific bands might not be relevant to you or to the paper’s conclusions, altering them constitutes a misrepresentation of the primary data and goes against our policy. It provides the false impression of a different (often cleaner) background. Importantly, non-specific bands or signal might not mean anything in the context of your investigation, but they might be biologically meaningful to another researcher, now or in the future.
Overall, these rules are largely based on good scientific behavior and common sense: you would want to see the blot as first obtained by the researcher doing the experiment in order to accurately interpret the results. We simply want to afford our readers the same opportunity. And while some adjustments are absolutely acceptable and do not alter the interpretation of the data, our screening is here to make sure that we only publish figures that you can trust are an accurate depiction of the primary results. In our next installment, we will discuss a related and very important aspect of proper figure-making: keeping the original data as first acquired for a long, long time. More next week.