My take on the ScienceIntegrityDigest article "Cassava Sciences: Of Stocks and Blots". I have experience with image processing and "photoshopping". Many of the images in the article certainly look suspect, as the author points out. Tell tale signs of editing are, changes in background texture, color, clarity, or the appearance of lines and rectangles. For example, rectangles are caused by selecting part of a picture, with the most common selection type (a rectangle), copying it from one image, and pasting it in another image. Usually the background of the two images will be slightly different, which causes your eye to "see" a rectangle around the copied part. Similarly, the lines are also an indication of a larger copy and paste. The articles example of these issues are pretty obvious.
The article's Figure 8 has a harder to detect issue. It took me a while to see it. But the dark horizontal brownish "bar" at the bottom of the blue boxes, has a unique shape in both pictures. At first I was not sure there was an issue, because the spots above the "bar" were not visible in both pictures. But upon further review, some of the spots that are plainly visible in the lower box "anti-Ab42" are in the same position in the upper box "anti-NFT". Many of the spots from the lower box are not obviously visible in the upper box. It seems they have been deleted and/or reduced in contrast. However, there are still SOME SPOTS visible above and to the right of the "bar", though not all are visible. The confirmation of the issue, is that you can see a FEW spots that are in identical positions on both. This is to unlikely to happen randomly. Analogous to no people having exactly the same fingerprint. The issue then is what happened to the other spots that have been apparently deleted and or reduced in contrast?
I have no experience with Western blots at all, I viewed the article from an image processing perspective. The author did a good job commenting on the blots and the pictures. I wish she had not included the "Other papers with image concerns", because it was not clear how much that related to Sava or not. She definitely knew what she was talking about.
That said, possible explanations for some of the issues might be that the preparer was trying to emphasize some spot, so they increased contrast, brightness, etc. around certain spots. I do not know if this is common, or accepted practice with Western blots. But if I owned Sava shares, I would want to find this out asap. Because if not, somebody has a lot of splainin to do.
Thanks to all who posted the link to the article.
