> research mice have been bred to the point of near[-]uniformity in genetic terms, making for greater reproducibility.

Doesn't that also introduce a bias toward experimental treatments that happen to work well with the genetics of research mice, but not those of other mice, vs. treatments that don't work well on research mice, but that mice with different genes respond well to?

In other words, would treatment that works really well for research mice but not the majority of other mice be able to make it through mouse trials without anyone finding that out?

Are there different lines of mice to mitigate that? Is that kind of genetic fluke uncommon?