Laboratory mice are not the same creatures you'll find living in your wall or a farm field. Instead, they are highly specialized. Whether born of high-tech genetic manipulation or good, old-fashioned selective breeding, each type of mouse—and there are thousands in the catalogs—is designed to test specific types of questions. Use the wrong mouse, and your data will be worthless. Or, at least, your conclusions will be incorrect.
That's what happened recently, when the authors of a 2006 paper in the Journal of Immunology retracted their work after belatedly realizing that it was based on studies done with the wrong mouse. They'd meant to buy a mouse that lacked a specific gene. Instead, thanks to a simple typo, they'd ended up with mice that lacked that gene—and a key chemical receptor in its cells, which changed the outcome of the research.
(Via Alexandra Witze)