Inadvertent errors during figure preparation and foolish we

Pubpeer has been purging the literature successfully. While some of the posts can be genuine human errors, most of them seem intentional manipulation of the data. For example, if the same western blot is copied as it is, it can happen during copy-pasting of figures, wherein the clipboard was not flushed properly. This is a benefit of doubt that many authors have cherished when spotted. Imagine another scenario, wherein the duplicated images were distinctly compressed/stretched, contrast–tuned, flipped or captured with different exposure times of the same blot to make them look different. The most common phrase the authors cite is “inadvertently duplicated during figure preparation”. Chutzpah !  These are clear cases of data manipulation and should ideally follow retraction of articles, not the erratum, no matter whether it alters the conclusion of the article (another common response of authors) or not. The reason is the loss of credibility. These are “within-article” manipulations, which might just be the tip of an iceberg. If authors do not shy to manipulate the same image in the very same article, cross-article copying of figures or including manipulated unpublished images stored in the gel-docs might be widespread. So, why should one trust the other figures of the article?

The problem of duplicated western blots is the outcome of journals’ flawed policies that lacked the requirement of original complete blot images to be uploaded as supplementary data. Interestingly, this has been selectively asked only for genomics/high-throughput datasets because that adds to citation of the paper and boosts journal impact factor. The good news is that some journals are taking the matter seriously for the new submissions at-least. Before publishing they are now scrutinising the figures thoroughly. It would be great if they can do the same with the published articles too, leaving aside their conflict of interests and brand reputation. If not, hail Pubpeer.