I am surprised by the reaction that it is ok that papers get retracted, this is how science works. Well, no. Science works by all co-authors, especially the lead author, checking all the details of experiments and insisting, when necessary, to perform experiments/analyses again and again. Yet, because of pressures to publish and the trend to have labs with tens of people, it is nearly impossible to check results of your peers. Another issue is the expertise and difficulty in understanding each other's results. I am a faculty member and I am very worried of publishing results that turn out to be unreproducible. This is why everybody in my lab, including myself, must show thorough analysis of results, including all controls, and repeat the analysis/experiments several times with alternative assumptions/methods. Yes, this makes us "slow" lab with publishing relatively slowly, and many people cannot work in my lab because of this "scrutiny". But the alternative is not acceptable, I think. So, it is not ok that a Nobel prize winner retracts the paper because of lack of appropriate oversight of experiments done in her lab. What if that paper was used to secure multi-million dollar grant? Should the grant be "returned"?