True enough—but get the genetic diversity incorporated into rodent models—not just fully inbred C57BL/6J and Fischer 344.

And if the experimental design calls for replicable genometypes (cases and controls) then just use sets of diverse types of F1 hybrids.