The proof they touched up
In May a science sleuth named Sholto David noticed something wrong in a picture that should have proved an antibody worked. One band in a western blot had been flipped and pasted into other lanes, like a signature repeated. He posted it. Soon a professor in Munich and a postdoc at Northwestern were looking too. By August the tally passed eighteen thousand questionable images across the big suppliers' catalogues. Some showed duplicated bands, making an antibody look like it binds where it does not. Some showed cleaned-up backgrounds, hiding where it binds by accident.
The company said the edits were cosmetic, made for clarity, and the products are fine. Maybe they are. But the point of the picture was to be proof, and proof that has been prettied stops being proof. Researchers trusted those pictures so they wouldn't have to check everything themselves.
What stays with me is who found it. A handful of people who looked carefully, for free, because looking is what they do. No algorithm found it first. No committee. Just people who refused to be told it was fine, and kept it in the light.
Careful attention is a kind of love. The world runs on the people who give it.