Tuesday, November 09, 2010

It's wrong, but I like the way you think

A high school classmate (he's in the picture here) writes regarding wrong but flashy scientific results.

Incidentally, I also saw the Ioannidis article you posted, and I think the guy is exactly correct. And it's not just for medical trial-type science - it's all sorts of flashy hot science. One instance that I found interesting because I've watched it develop for several years is David Sinclair's work with resveratrol. This is a compound found in red wine that's supposed to have all sorts of fantastic effects in a bunch of organisms, including making them live longer. One of the things I find interesting is that this is an area where big business is making a big bet on what I believe they know is bogus science. Resveratrol is supposed to activate an enzyme called a sirtuin, and that's supposed to elicit these nice effects. The problem is that the data showing that it activates the enzyme is an artifact of the assay. My boss gave a talk at GlaxoSmithKline explaining all of this, showing the data, publishing the data (with another group publishing the same finding at the same time). Those big companies have top notch chemists and they understood all of this. And after all of this, GSK lays out $720 million for David Sinclair's resveratrol company. I used to think that once big money becomes involved, they would never do something like this. Now I think that they actually may have made a good financial decision. They're laying out less than 10% of their market capitalization, they get a bunch of hype with Sinclair on Oprah and people talking about red wine and this new wonder drug, they will end up taking a bit of a hit once it all unravels, but who cares - by then they're on to something new!

A Nobel laureate down the hall from me recently retracted articles in Nature, Science and PNAS because the underlying data were bogus, and a friend of mine who was in that lab is writing a book on how that fraud was perpetrated. I think this stuff is remarkably common and that Ioannidis is right in his analysis.

No comments:

Post a Comment