Why animal studies are often poor predictors of human reactions to exposure

Someone once started a Twitter feed sharing articles about new and intriguing studies -- just so they could add the two crucial missing words. "In mice."

Forgot to add (2009) to the title