Wednesday, September 9, 2015

Huh, Amgen researcher let go for data manipulation

Via Retraction Watch, a Wall Street Journal article by Jonathan Rockoff on an Amgen researcher who was fired for data manipulation: 
A scientific paper that captured widespread attention because its subjects were massive grizzly bears—and hinted at a new approach to treating diabetes—was retracted Tuesday after one of the authors was said to have manipulated some of the data. 
The paper attracted news coverage around the world after its publication in August 2014 in the journal Cell Metabolism, whose cover featured the image of a grizzly bear clutching a fish between its jaws. The paper discussed how a grizzly bear’s metabolism adjusts to hibernation, and the key role of a certain fat protein. 
Biotech company Amgen Inc. was working on the bear research to get a better grip on the biology behind diseases like obesity and diabetes. 
But Amgen said it found late last year, in reviewing the computer files of one of its researchers, that some experimental data cited in the Cell Metabolism paper had been changed in a way the company said made some of the results look stronger.
[snip] 
 ...The paper had 12 authors, six of whom had worked for Amgen while conducting the research. The senior author was Kevin Corbit, who said he was dismissed by Amgen for fabricating research “on another matter.” Dr. Corbit said he fabricated that data to help a co-worker in what he said was an isolated incident and a regrettable decision.
He said he stands behind the grizzly-bear paper, and that he believes that “if independent experiments were conducted, as dictated by proper scientific discourse, the work would be reproduced.” He says his co-authors have no involvement in any dispute over the work.
Retraction Watch frequently covers retractions from academia and the legal consequences when falsification on federally funded research has been found to take place; usually, the scientist is barred from NIH funding for a period of time. This Amgen case seems to me to be a rare instance where there have been industry consequences for reproducibility issues (broadly speaking).

(I don't think this can accurately speak to differences in the quality of research between academia and industry.)

Readers, can you think of more examples where industrial scientists have been let go for retractions and the like?  

5 comments:

  1. I wouldn't be surprised if the notoriety of the research was a factor here - this story was all over the place when it came out.

    ReplyDelete
  2. We fired someone at the top of our technical ladder for this crime. We didn't publicize it, mostly in impacted internal projects and a couple young scientist's careers, who were let go for being incompetent when they really asked too many questions of the high-level scientist. My friend who lost their job had grounds to sue, but instead moved to a new job and left the baloney behind. I think the person who was fired for fabricating data has not found new work.

    ReplyDelete
  3. This doesn't happen nearly as often as it should.

    ReplyDelete
  4. Kevin Corbit also took credit for one of my grizzly bear photos that he used in his study. The photo has since been removed.

    ReplyDelete
  5. The research conducted in industry is subject to a lot of internal scrutiny because ultimately as a project moves through the pipeline to deliver a drug candidate the supporting data will have to be highly reproducible and vetted by other functional areas within the company (the sponsor) in terms of efficacy and safety. The leadership of the company then reviews the data package and its likelihood that it will show benefit in patients. A clinical trial costs millions of dollars and hence all data including raw data have to be convincing and are also subject to audits by regulators. Hence drug developers apply intense scrutiny to data generated by their scientists. I have worked in top tier industrial and academic institutions and I can say that there is a lot more scrutiny about data integrity in industry given that ultimately drugs candidates have to work in patients to become successfully marketed drugs.

    ReplyDelete

looks like Blogger doesn't work with anonymous comments from Chrome browsers at the moment - works in Microsoft Edge, or from Chrome with a Blogger account - sorry! CJ 3/21/20